[EM] Legacy IRV limitations

[Richard, the VoteFair guy]

On 12/19/2023 7:03 PM, Greg Dennis wrote:
 > The word Australia doesn't appear anywhere in that document. Whatever
 > the rationale for using that treatment of overvotes, there is no
 > evidence that backwards-compatibility with Australia elections was a 
factor.

I don't have an encyclopedia-like reference that explicitly states there 
is a linkage between Australia and Portand (Oregon) in how overvotes are 
counted.

Instead, here are some facts that show this linkage ...

RCVRC = Ranked Choice Voting Resource Center

Long ago the original leader of the FairVote organization, Rob Richie, 
co-wrote a book that, if I recall correctly, promoted what we now call 
IRV and STV as used in Australia.  Here's the Amazon link to that book:

https://www.amazon.com/gp/product/0807044237

I don't know the degree of, and nature of, collaboration between 
FairVote and RCVRC.  They promote similar legaleze wordings for IRV. 
And those wordings always refer to not counting overvotes.

Remember that RCVRC was created long after the FairVote organization.

Remember the timeline for Australia.  They have been using ranked choice 
voting for more than 100 years (if I recall correctly), and the switch 
to computer counting occurred bit by bit, with the "holdouts" switching 
not that many years ago.  This is added evidence that the counting 
conventions did not formally (and probably informally) change when they 
switched to computer counting.  In other words they had no reason to 
start counting overvotes differently as each jurisdiction adopted 
computer counting.

During the Portland City Council meeting in which testimony was given 
for, and against, the specific wording for ranked choice voting (which 
in principle had already been approved by Portland voters), the Portland 
(and related) election officials made it clear they were relying on 
RCVRC for RCV details.  In fact, a person at RCVRC attended the meeting 
via video, and he answered some questions about RCV.

(Remember that Portland has adopted STV for city council elections and 
IRV for mayoral elections, with both together being called "ranked 
choice voting.")

One of the topics that was discussed was the option for how to count 
overvotes.  The only two options available (according to the 
election-system vendors) were to skip over them -- which was chosen a 
few weeks earlier based on advice from RCVRC -- or else to dismiss them 
when the counting reached an overvote -- which would have been a better 
choice in Oregon where everyone marks ballot ovals at home with a pen.

So I can't give you a definitive adoption path for overvotes being 
counted in Portland the same as in Australia.  Yet the fingerprints are 
quite clear.  They go from Australia to FairVote to RCVRC to Portland.

The word "legacy" in this forum thread refers to what I have referred to 
as backwards compatibility, but I'm using this term loosely.  I'm not 
claiming it's like what happens when software changes are carefully tracked.

Richard Fobes
The VoteFair guy


On 12/19/2023 7:03 PM, Greg Dennis wrote:
>       >>   Ranked choice voting software must be backwards compatible with
>       >>   prior elections in Australia.
>       > I do not know the source of this claim and have never heard or read
>       > anyone from FairVote state this. If you have a source, I'd be
>     glad to
>       > read it.
> 
> 
>     Here's the wording that was adopted for Portand (Oregon) using ranked
>     choice voting:
> 
> 
>     https://efiles.portlandoregon.gov/Record/16114624/File/Document
>     <https://efiles.portlandoregon.gov/Record/16114624/File/Document>
> 
> 
> The word Australia doesn't appear anywhere in that document. Whatever 
> the rationale for using that treatment of overvotes, there is no 
> evidence that backwards-compatibility with Australia elections was a factor.
> 
> - Greg