12/15 Agenda and Controlling Contests
John Dziurlaj
Agenda for 12/15:
- Define the following concepts
- Controlling Contest
- Non-Party Activation Contest
- Activation Option
- Party Activation Contest
- Straight Party Contest
- Controlled Contest
- Indirect Selection
- Direct Selection
- Contest Eligibility/Activation
This document (below) is to serve as a description of controlling/controlled contests and guide discussion within the CDF research group.
Talk to you tomorrow (Wednesday) at Noon ET!
John
---
There are some “contests” appearing on certain ballots that do not contain candidate or ballot measure options. Instead they affect how (and sometimes if) other contests on the ballot are processed. Such contests are called controlling contests and the contests they control are called controlled contests. Each controlling contest controls one or more controlled contests, with controlled contests almost always appearing subsequent to their controlling contest in ballot order.
What is controlled depends on the particular kind of controlling contest, as:
- Non-Party Activation Contest
- Party Activation Contest
- Straight Party Contest
When a non-party activation contest (e.g. a recall) is used, the state of its controlled contests can be configured to have a “default controlled state”, the value active/eligible meaning the controlled contests may be voted on without making a selection in the controlling contest, inactive/ineligible, meaning that the controlled contests cannot be voted on unless an “activation option” selection is made in the controlling contest. In Cast Vote Record (CVR) terms, when the controlled contest is inactive/ineligible, indications can still be made (e.g. via marks on a paper ballot), but votes in inactive contests cannot be allocated to their corresponding contest option accumulator. A non-party activation contest can have one or more associated options, one of which must be an activation option. A selection for the activation option causes all controlled contests to the switch their state (e.g. a controlling contest with DefaultControlledState=ineligible with a selected activation option would flip all controlled contests to “eligible”).
Party activation contests can be seen as a subset of regular “party contests” (e.g. a party list), for which all contest options are parties. However, the impact of selecting a party is different: to make ineligible all contests associated with parties other than the one selected by the voter. A fuller treatment is in the attached Voting Methods excerpt (attached).
Last is straight party voting which is probably the most complicated of the bunch. A straight party can have one or more controlled contests (each of which containing partisan candidates). A valid selection in the straight party contest will result in indirect selections being made across the associated controlled contests. There are rules regarding how to handle direct selections in a controlled contest, which are fully treated in the Voting Methods excerpt (attached).
The impact of a selection in each controlling contest varies by contest method. I’ve attached a spreadsheet describing the three contest types and the impact of 1) selections in the controlling contest and 2) selections in the controlled contest(s).
John McCarthy
Thanks for the agenda and helpful attachments John.
In order to make the spreadsheet more easily readable on a single page, I did a bit of reformatting. I've attached a copy of the reformatted single page spreadsheet in case it might be helpful for others.
Did the Voting Methods Working Group provide any definitions for the terms you've outlined? If so that might provide a helpful starting point for (y)our work.
Here's a question I'll pose ahead of our meeting...
How would this framework be used to define California-style
Judicial Retention Elections on the ballot? If I recall correctly,
for each individual judge, voters are asked to simply vote "yes"
or "no" to retain that judge or not (or something like that). See
https://ballotpedia.org/Retention_election
Would this be treated as a special kind of recall election
(except that the questions is whether to retain rather than
whether to recall)?
See you tomorrow...
John
--
To unsubscribe from this group, send email to cdf-ballot-sty...@list.nist.gov
View this group at https://list.nist.gov/cdf-ballot-styles
---
To unsubscribe from this group and stop receiving emails from it, send an email to cdf-ballot-sty...@list.nist.gov.
| John McCarthy Volunteer Advisor (he/him) |
| jo...@verifiedvoting.org• 510.666.5309 |
verifiedvoting.org
    |
 |
.
Deutsch, Herb
Get Outlook for Android
Sent: Tuesday, December 14, 2021 9:52:59 PM
To: John Dziurlaj <jo...@turnout.rocks>; cdf-ball...@list.nist.gov <cdf-ball...@list.nist.gov>
Subject: Re: cdf-ballot-styles 12/15 Agenda and Controlling Contests
CONFIDENTIALITY NOTICE:
This e-mail transmission and any documents, files or previous e-mail messages attached to it may contain information that is confidential, protected by the attorney/client or other privileges, and may constitute non-public information. It is intended to be conveyed only to the designated recipient(s) named above.
Any unauthorized use, reproduction, forwarding, distribution or other dissemination of this transmission is strictly prohibited and may be unlawful. If you are not an intended recipient of this email transmission, please notify the sender by return e-mail and permanently delete any record of this transmission. Your cooperation is appreciated.
To find out more Click Here.
John Dziurlaj
Hello John,
Election Results Reporting (which Ballot Definition is using as a strawman) supports a RetentionContest, which allows ballot measure style contest options to be used for voters to determine whether an existing officeholder (modeled as a candidate) should be retained. I do not believe they behave like RetentionContests behave a controlling/controlled contest pair (would replacements appear on the same ballot?), but if anyone else has information on this please let me know!
John Dziurłaj
Carl Hage
> Here's a question I'll pose ahead of our meeting...
> How would this framework be used to define California-style Judicial
> Retention Elections on the ballot? If I recall correctly, for each
> individual judge, voters are asked to simply vote "yes" or "no" to
> retain that judge or not (or something like that). See
redundant questions or comments.
Besides "Retention" (keep an incumbent yes/no) there is "Approval"
voting for CA judges, almost the same as retention, but the candidate
was nominated (I am not sure how) and voters confirm, analogous to
Senate confirmation of judges. Both are similar to recall contest
(yes/no) but there is no linked replacement contest, and also the
interpretation of yes/no is inverted. (A yes vote in a recall means the
incumbent is not retained, vs retention, a yes vote means keep the
incumbent).
A retention/approval/recall contest is similar to a ballot measure (has
yes/no selections) except there is a candidate or incumbent with a name
and elected office. We want to represent these explicitly as data in a
CDF rather than expect an automated system to understand the words in
the ballot question.
There are 3 kinds of usage in the extended ballot definitions discussed
here:
- Contest definitions used to construct ballots and ballot layouts
(paper or machine).
- Information for tabulators that take a ballot instance with
marks/selections (or CVR) and compute vote totals. For this usage,
control contests need to be defined so a contest (for
non-party-activation) or selections (for party-activation) can be
deactivated, or in the case of the straight party model, undervote (or
blank vote) can be counted with the party selection default.
- Information for results reporting. This would enable automated
interpretation of the results for generating public results reporting.
For ballot measures and approval/recall we need to indicate the meaning
of the selections as approve/disapprove since the words in the selection
title are not always yes/no. For these we also have an approval required
to pass, in California is >50%, 55%, or 2/3 depending on the measure type.
A recall replacement contest has the control contest that is the recall
question, but instead of activating a selection for counting, the
control contest indicates if the replacement contest result is active
(applies). So a usual case is that anyone can vote for replacements (no
matter which selection was made in the recall), but the winner of the
recall replacement is elected only if the recall succeeds.
An automated voter information system generally cannot read English, so
if the CDFs for recall or retention do not identify the incumbent being
removed/retained, then manual intervention is required to supply the
missing info, or write patterns to interpret the English.
The control contest model seems OK to me. I presume this would be
represented by a value in the controlled contest that references the
controlling contest, plus some sort of attribute that indicates how to
interpret the controlling contest results or for electronic voting, the
selection displays. A single controlling contest link and attribute
seems simpler and more extensible than trying to make a contest subclass
for each controlling contest type. Perhaps there would be multiple
attributes, e.g. indicating how an overvote in a straight party
controlled contest is interpreted. (We could of course just use
attribute values to indicate combinations, e.g.
"straight-party-overvote" means apply the party selection to undervote
or overvote, while "straight-party" means apply a party selection for
undervote only.
----
Here is another complex contest relationship that applies to results
reporting that I have seen. For some contests (ballot measures or board
of directors election) for a special district, eligible voters are
residents in the district and also people who own land within the
district. There is a separate contest for nonresident landowners,
tabulated separately, and the results of the contest combines a contest
for resident voters and the contest for landowners. The contest
definition data exported from an EMS is sometimes confusing due to this
distinction. Sometimes there is a separate district definition to
represent non-resident landowners with voter registration. (Nonresident
landowner ballots have been vote by mail in all cases I've seen.)
I this example, we don't have controlling contests, but there are
separate subtotals. I am not sure, but there might be some ballot
measures with separate subtotals, e.g. for a school district transfer
area or city annexation, there could be separate voting for residents of
the transfer/annexation area, and residents of the school district or
city joined. Also measures for merger of 2 districts might have separate
votes and interpretation, e.g. the transfer is approved only if approved
in both contests.
For results reporting in districts that span counties (election admins),
each election admin usually only gives subtotals within the county (or
whatever area) they administer. There is no indication that the result
is a subtotal and no linkage or definition of any contest in other
election admins. It's up to every newspaper, TV station, web site, etc.
to identify and correlate the cross-county subtotals.
The approval required definitions and combined contest subtotal
definitions will affect risk limiting audits. So for a 2/3 approval
required, the number of ballots requiring audit depend on the difference
between yes/no and the 2/3 of total votes.
Deutsch, Herb
Illinois also has Judges Retention referenda on the ballot but they are all standalone contests and are treated like referenda. There is no special contest type for them for tabulation purposes. Judges are elected on a different election’s ballots like other contests.
From: cdf-ball...@list.nist.gov <cdf-ball...@list.nist.gov>
On Behalf Of Carl Hage
Sent: Thursday, December 16, 2021 7:49 PM
To: cdf-ball...@list.nist.gov
Subject: Re: cdf-ballot-styles 12/15 Agenda and Controlling Contests
On 12/14/21 7:52 PM, John McCarthy wrote:
--
To unsubscribe from this group, send email to
cdf-ballot-sty...@list.nist.gov
View this group at
https://list.nist.gov/cdf-ballot-styles
---
To unsubscribe from this group and stop receiving emails from it, send an email to
cdf-ballot-sty...@list.nist.gov.
John Dziurlaj
From: cdf-ball...@list.nist.gov <cdf-ball...@list.nist.gov> On Behalf Of Carl Hage
Sent: Thursday, December 16, 2021 8:49 PM
To: cdf-ball...@list.nist.gov
Subject: Re: cdf-ballot-styles 12/15 Agenda and Controlling Contests