Auditing STV
0 views
Skip to first unread message
SomeTh...@aol.com
Oct 26, 2009, 6:20:39 PM10/26/09
to Election...@googlegroups.com, CalifElecti...@yahoogroups.com, voting...@yahoogroups.com
The principle addressee of this email is the national Election Integrity group.
There has been boundless discussion of whether or not IRV is
in principle good or bad, or whether or not people like it. I would like
to suggest that contributors now focus the disussion on the
topic of IRV & EI.
That said, I have been participating in a check of a small STV election.
STV is multi-seat IRV. People vote for several candidates to fill
several seats.
The case in point is the election to the citizens board of the KPFA
radio station in Berkeley, CA. There are 2 groups in the election.
The one I have data for involves 174 volunteer staff members voting
for 3 staff seats on the board. The vote by registered listeners is
in the thousands, and I do not have data for that yet.
Much data is posted here:
http://www.4shared.com/dir/22095198/b1d9cdca/STAFFBALLOTS.html
Ballot # 0069 looks like this:
The most useful of the files is 2009KPFAstaff.in, which I have appended to the end
of this email.
The good news is that they published a lot of data, including ballot images.
The bad news is that the ballot images are hand-copied ballots, all of them.
My guess is that the primary reason for this is to make sure the brand-new
scanning software reads the ballots clearly. They might have also
posted images of the original ballots for verification. There are exist issues
of privacy and secrecy that are not fully clear to me. Being able to look at
the ballots is an enormous plus in terms of retracing how the votes were
counted. In this case, I am not able to say that the ballot images are
accurate with what I know now.
This is an STV election http://en.wikipedia.org/wiki/Single_transferable_vote
using fractional surplus votes. This variation also allows voters to cast
three 1st-choice votes, as ballot # 115 did. In this case, the votes
were split into thirds, 0.33333 votes for each candidate.
I was able to retrace how the counting software dealt with each ballot
and fractionalized and refractionalized each ballot/votes as it
worked through the rounds. The data in KPFA Ballot Allocation Report.txt,
KPFA Final Piles Report.txt and KPFA Detail Report.txt was very good
and made this possible. No special computer skills needed.
That was all very good. But then I imagined extrapolating this to
over 100,000 votes. Retracing by hand would be so cumbersome
as to be nearly impossible. A full hand recount is also not doable.
You would have to keep track of all the fractional votes, which
keep on changing as they go through the rounds. Clerical errors
would occur, and trying to reconcile many steps and fractional
votes by hand would be a nightmare.
The best way to check the counting of a large STV election with
fractional votes is to publish data like that in 2009KPFAstaff.in,
and let anybody run it through their own tabulating progam.
If 10 open-source programs all came out with exactly the same
results, I would have confidence in the counting. Fortunately,
such a program is not hard to write. You have to make sure
that the rules are clear and followed exactly.
BTW: The German Constitutional Court would not allow checks
of counts by computer program. It requires computer skills.
But KPFA is not Germany, and they do not have government
resources. They do have choices they can make in the future,
so long as they understand the costs and benefits.
Which leads me to ask, what other ways exist for multi-seat
elections for a small civic organization with no parties?
I am asking for resources that detail how it is done. I would
like to be able to present KPFA with other options. I will
ask that you send the procedures directly to me, and
not to this list, as such a discussion would be straying
off the EI topic, and other readers are getting annoyed with
the long IRV debate in general. My email address is
SomeThoughts at Aol dot Com.
I would like to make a couple of things clear. (1) The verification
problems involve the use of fractional votes in large elections.
It is extremely difficult to do that by hand. There exist
procedures for STV elections without fractional votes.
I don't know the details.
(2) IRV does not use fractional votes. The first round of
an IRV election seems to be spot-checkable, just like in a
plurality election. The "margin of victory" is measured 1st-place
candidate vs 50%, not 1st-place candidate vs 2nd-place
candidate, like in a plurality election. I am still studying
the question of how to check further rounds. It may be
that a full hand recount is necessary. There are some
other ideas, but I won't comment on them now.
======================
2009KPFAstaff.in
A line like
88) 2= 9, 1, 4
means, Ballot 88, gave 1st choice to candidates 2 & 9.
Second choice to candidate 1. Third choice to candidate 4.
.TITLE KPFA LSB Staff Election
.TITLE October 18, 2009
.ELECT 3
.CANDIDATE 1, "Emiliano Echeverria"
.CANDIDATE 2, "Anthony Fest"
.CANDIDATE 3, "David Gans"
.CANDIDATE 4, "Renee Yang Geesler"
.CANDIDATE 5, "Lewis O'Sawyer Jr."
.CANDIDATE 6, "Clay Leander"
.CANDIDATE 7, "Antonio Ortiz"
.CANDIDATE 8, "Laura Prives"
.CANDIDATE 9, "Write-In"
.EXCLUDE-CANDIDATE 6
.CONTEST KPFA LSB Staff
.STATISTICS-ON
.THRESHOLD DROOP
1) 3, 6, 8
2) 3, 8, 7, 6
3) 3, 6, 8, 7, 5, 2, 4, 1
4) 3, 8, 6, 4, 7
5) 3, 6, 8, 7
6) 3, 1, 6, 8
7) 3, 8, 6, 7
8) 3, 8, 7, 6
9) 7, 8, 3, 6
10) 7, 8, 6, 3
11) 4, 2, 1
12) 4, 2, 1
13) 4, 2, 1
14) 4, 7, 6, 1, 5
15) 4, 1, 2, 6
16) 4, 1, 2, 5
17) 3, 8, 7, 6
18) 3, 8, 6, 7
19) 3, 8, 2, 4, 5, 6, 7, 1
20) 3, 8, 6
21) 4, 2, 1, 6
22) 4, 2, 1
23) 4, 2, 1
24) 4, 1, 7, 8
25) 4, 1, 2, 7
26) 4, 6, 2
27) 4, 1, 2, 6
28) 4, 7, 8, 4, 7, 8, 4, 7, 8
29) 4, 1, 8
30) 8, 7, 6, 3
31) 4, 5, 1, 2
32) 1, 2, 4
33) 1, 2, 4, 5
34) 2, 1, 4
35) 6, 5, 4, 7, 1, 2, 3, 8
36) 2, 4, 1, 5
37) 2, 4, 5, 8, 6, 7
38) 4, 2, 1, 6, 7
39) 4, 1, 7, 6, 8
40) 4, 2, 1, 8, 5, 3, 7, 6
41) 4, 2, 1
42) 2, 1, 4, 9, 6
43) 2, 4, 1
44) 1, 4, 2
45) 1, 2, 4
46) 1, 2, 4, 5
47) 1, 4, 2
48) 1, 4, 2
49) 1, 2, 4
50) 1, 4, 2, 5, 6, 7, 3, 8
51) 1, 2, 4
52) 3, 7, 8, 6
53) 3, 6, 7, 8
54) 3, 7, 8, 6
55) 3, 6, 7, 4, 2, 5, 8, 1
56) 3, 7, 8, 6
57) 4, 1, 2
58) 6, 3, 1
59) 3, 4
60) 3, 1, 6
61) 9, 2, 5
62) 1
63) 1, 2, 6
64) 4, 1, 2
65) 4, 2, 1= 5
66) 5, 2, 4, 1, 7, 3, 6, 8
67) 2, 4, 1, 5
68) 2, 4
69) 2, 6, 3, 7, 8, 1= 4, 5, 1
70) 8, 7, 6, 4, 3
71) 8, 7, 6, 3
72) 8, 3, 6, 7
73) 8, 7, 3, 6, 1, 4
74) 8, 7, 4, 3, 6
75) 8, 6, 7, 3
76) 8, 3, 7, 6, 1, 4, 5, 2
77) 8, 7, 3, 6
78) 8, 7, 3, 6, 4, 1
79) 8, 7, 6
80) 2, 4, 1, 5
81) 2, 4, 1, 5
82) 2, 4, 1
83) 2, 1, 8, 4, 7, 3, 6
84) 2, 1, 4
85) 2, 4, 1, 5
86) 2, 1, 4, 5
87) 2, 4, 1
88) 2= 9, 1, 4
89) 2, 4, 1
90) 2, 7, 1, 5
91) 2, 4, 1, 7
92) 2, 6, 4
93) 2, 4, 1
94) 2, 1, 4
95) 2, 4, 1, 5
96) 2, 1, 4
97) 2, 1, 4
98) 2, 6, 3, 7, 8
99) 2, 1, 4
100) 2, 6, 1
101) 2, 1, 4
102) 2, 1, 4
103) 2, 1, 4, 3
104) 2, 1, 4
105) 2, 3, 5, 4
106) 2, 1, 4
107) 2, 1, 4
108) 2, 1, 4, 5
109) 2, 4, 1
110) 2, 1, 4
111) 8, 7, 3, 6, 4
112) 4, 7, 6
113) 4, 1, 7
114) 6, 4, 7
115) 1= 4= 7
116) 6, 7, 8, 4, 2, 3, 1
117) 6, 7, 8
118) 6, 7, 8, 3
119) 6, 7, 3, 4= 8, 1, 5, 2
120) 7, 6, 8, 3
121) 7, 8, 6
122) 7, 8, 6, 3
123) 7, 8, 6, 3
124) 7, 8, 6, 3
125) 7, 6, 8, 3
126) 7, 1, 8, 2
127) 7, 8, 6, 3
128) 7, 6, 8, 3
129) 3, 8, 7, 6, 1, 4
130) 4, 8, 1
131) 6, 8, 3, 7
132) 6, 8, 7
133) 6, 8, 2, 3, 4
134) 6, 8, 7
135) 6, 8, 4
136) 8, 6, 3
137) 8, 3, 7, 6
138) 8, 6, 3
139) 8, 6, 3, 7
140) 8, 7, 3
141) 8, 7, 6
142) 8, 7, 6
143) 6, 7, 8, 3
144) 4, 2, 1, 6, 8, 5, 7
145) 8, 3, 7, 6
146) 7, 8, 6, 3, 4
147) 7, 8, 6, 3
148) 4, 5, 2, 6, 8
149) 1, 2, 4, 5, 6, 8, 7, 3
150) 2, 1, 4
151) 2, 1, 4
152) 2, 1, 4
153) 5, 2, 1
154) 4= 6, 8, 1, 2, 3, 5, 7
155) 5= 6= 7
156) 1, 2, 4
157) 3, 6, 8, 7
158) 7, 8, 6
159) 7, 3, 8
160) 2, 4, 1
161) 2, 1, 4
162) 2, 1, 4, 5
163) 8, 1, 5, 7
164) 3, 8, 7, 6
165) 4, 1, 2
166) 4, 1, 7
167) 4, 2, 1
168) 4, 2, 7
169) 6, 4, 2
170) 2, 1, 4
171) 4, 1, 2
172) 7, 8, 6, 3
173) 7, 6, 3
174) 8, 3, 6, 7, 4, 5, 1, 2
There has been boundless discussion of whether or not IRV is
in principle good or bad, or whether or not people like it. I would like
to suggest that contributors now focus the disussion on the
topic of IRV & EI.
That said, I have been participating in a check of a small STV election.
STV is multi-seat IRV. People vote for several candidates to fill
several seats.
The case in point is the election to the citizens board of the KPFA
radio station in Berkeley, CA. There are 2 groups in the election.
The one I have data for involves 174 volunteer staff members voting
for 3 staff seats on the board. The vote by registered listeners is
in the thousands, and I do not have data for that yet.
Much data is posted here:
http://www.4shared.com/dir/22095198/b1d9cdca/STAFFBALLOTS.html
Ballot # 0069 looks like this:
The most useful of the files is 2009KPFAstaff.in, which I have appended to the end
of this email.
The good news is that they published a lot of data, including ballot images.
The bad news is that the ballot images are hand-copied ballots, all of them.
My guess is that the primary reason for this is to make sure the brand-new
scanning software reads the ballots clearly. They might have also
posted images of the original ballots for verification. There are exist issues
of privacy and secrecy that are not fully clear to me. Being able to look at
the ballots is an enormous plus in terms of retracing how the votes were
counted. In this case, I am not able to say that the ballot images are
accurate with what I know now.
This is an STV election http://en.wikipedia.org/wiki/Single_transferable_vote
using fractional surplus votes. This variation also allows voters to cast
three 1st-choice votes, as ballot # 115 did. In this case, the votes
were split into thirds, 0.33333 votes for each candidate.
I was able to retrace how the counting software dealt with each ballot
and fractionalized and refractionalized each ballot/votes as it
worked through the rounds. The data in KPFA Ballot Allocation Report.txt,
KPFA Final Piles Report.txt and KPFA Detail Report.txt was very good
and made this possible. No special computer skills needed.
That was all very good. But then I imagined extrapolating this to
over 100,000 votes. Retracing by hand would be so cumbersome
as to be nearly impossible. A full hand recount is also not doable.
You would have to keep track of all the fractional votes, which
keep on changing as they go through the rounds. Clerical errors
would occur, and trying to reconcile many steps and fractional
votes by hand would be a nightmare.
The best way to check the counting of a large STV election with
fractional votes is to publish data like that in 2009KPFAstaff.in,
and let anybody run it through their own tabulating progam.
If 10 open-source programs all came out with exactly the same
results, I would have confidence in the counting. Fortunately,
such a program is not hard to write. You have to make sure
that the rules are clear and followed exactly.
BTW: The German Constitutional Court would not allow checks
of counts by computer program. It requires computer skills.
But KPFA is not Germany, and they do not have government
resources. They do have choices they can make in the future,
so long as they understand the costs and benefits.
Which leads me to ask, what other ways exist for multi-seat
elections for a small civic organization with no parties?
I am asking for resources that detail how it is done. I would
like to be able to present KPFA with other options. I will
ask that you send the procedures directly to me, and
not to this list, as such a discussion would be straying
off the EI topic, and other readers are getting annoyed with
the long IRV debate in general. My email address is
SomeThoughts at Aol dot Com.
I would like to make a couple of things clear. (1) The verification
problems involve the use of fractional votes in large elections.
It is extremely difficult to do that by hand. There exist
procedures for STV elections without fractional votes.
I don't know the details.
(2) IRV does not use fractional votes. The first round of
an IRV election seems to be spot-checkable, just like in a
plurality election. The "margin of victory" is measured 1st-place
candidate vs 50%, not 1st-place candidate vs 2nd-place
candidate, like in a plurality election. I am still studying
the question of how to check further rounds. It may be
that a full hand recount is necessary. There are some
other ideas, but I won't comment on them now.
======================
2009KPFAstaff.in
A line like
88) 2= 9, 1, 4
means, Ballot 88, gave 1st choice to candidates 2 & 9.
Second choice to candidate 1. Third choice to candidate 4.
.TITLE KPFA LSB Staff Election
.TITLE October 18, 2009
.ELECT 3
.CANDIDATE 1, "Emiliano Echeverria"
.CANDIDATE 2, "Anthony Fest"
.CANDIDATE 3, "David Gans"
.CANDIDATE 4, "Renee Yang Geesler"
.CANDIDATE 5, "Lewis O'Sawyer Jr."
.CANDIDATE 6, "Clay Leander"
.CANDIDATE 7, "Antonio Ortiz"
.CANDIDATE 8, "Laura Prives"
.CANDIDATE 9, "Write-In"
.EXCLUDE-CANDIDATE 6
.CONTEST KPFA LSB Staff
.STATISTICS-ON
.THRESHOLD DROOP
1) 3, 6, 8
2) 3, 8, 7, 6
3) 3, 6, 8, 7, 5, 2, 4, 1
4) 3, 8, 6, 4, 7
5) 3, 6, 8, 7
6) 3, 1, 6, 8
7) 3, 8, 6, 7
8) 3, 8, 7, 6
9) 7, 8, 3, 6
10) 7, 8, 6, 3
11) 4, 2, 1
12) 4, 2, 1
13) 4, 2, 1
14) 4, 7, 6, 1, 5
15) 4, 1, 2, 6
16) 4, 1, 2, 5
17) 3, 8, 7, 6
18) 3, 8, 6, 7
19) 3, 8, 2, 4, 5, 6, 7, 1
20) 3, 8, 6
21) 4, 2, 1, 6
22) 4, 2, 1
23) 4, 2, 1
24) 4, 1, 7, 8
25) 4, 1, 2, 7
26) 4, 6, 2
27) 4, 1, 2, 6
28) 4, 7, 8, 4, 7, 8, 4, 7, 8
29) 4, 1, 8
30) 8, 7, 6, 3
31) 4, 5, 1, 2
32) 1, 2, 4
33) 1, 2, 4, 5
34) 2, 1, 4
35) 6, 5, 4, 7, 1, 2, 3, 8
36) 2, 4, 1, 5
37) 2, 4, 5, 8, 6, 7
38) 4, 2, 1, 6, 7
39) 4, 1, 7, 6, 8
40) 4, 2, 1, 8, 5, 3, 7, 6
41) 4, 2, 1
42) 2, 1, 4, 9, 6
43) 2, 4, 1
44) 1, 4, 2
45) 1, 2, 4
46) 1, 2, 4, 5
47) 1, 4, 2
48) 1, 4, 2
49) 1, 2, 4
50) 1, 4, 2, 5, 6, 7, 3, 8
51) 1, 2, 4
52) 3, 7, 8, 6
53) 3, 6, 7, 8
54) 3, 7, 8, 6
55) 3, 6, 7, 4, 2, 5, 8, 1
56) 3, 7, 8, 6
57) 4, 1, 2
58) 6, 3, 1
59) 3, 4
60) 3, 1, 6
61) 9, 2, 5
62) 1
63) 1, 2, 6
64) 4, 1, 2
65) 4, 2, 1= 5
66) 5, 2, 4, 1, 7, 3, 6, 8
67) 2, 4, 1, 5
68) 2, 4
69) 2, 6, 3, 7, 8, 1= 4, 5, 1
70) 8, 7, 6, 4, 3
71) 8, 7, 6, 3
72) 8, 3, 6, 7
73) 8, 7, 3, 6, 1, 4
74) 8, 7, 4, 3, 6
75) 8, 6, 7, 3
76) 8, 3, 7, 6, 1, 4, 5, 2
77) 8, 7, 3, 6
78) 8, 7, 3, 6, 4, 1
79) 8, 7, 6
80) 2, 4, 1, 5
81) 2, 4, 1, 5
82) 2, 4, 1
83) 2, 1, 8, 4, 7, 3, 6
84) 2, 1, 4
85) 2, 4, 1, 5
86) 2, 1, 4, 5
87) 2, 4, 1
88) 2= 9, 1, 4
89) 2, 4, 1
90) 2, 7, 1, 5
91) 2, 4, 1, 7
92) 2, 6, 4
93) 2, 4, 1
94) 2, 1, 4
95) 2, 4, 1, 5
96) 2, 1, 4
97) 2, 1, 4
98) 2, 6, 3, 7, 8
99) 2, 1, 4
100) 2, 6, 1
101) 2, 1, 4
102) 2, 1, 4
103) 2, 1, 4, 3
104) 2, 1, 4
105) 2, 3, 5, 4
106) 2, 1, 4
107) 2, 1, 4
108) 2, 1, 4, 5
109) 2, 4, 1
110) 2, 1, 4
111) 8, 7, 3, 6, 4
112) 4, 7, 6
113) 4, 1, 7
114) 6, 4, 7
115) 1= 4= 7
116) 6, 7, 8, 4, 2, 3, 1
117) 6, 7, 8
118) 6, 7, 8, 3
119) 6, 7, 3, 4= 8, 1, 5, 2
120) 7, 6, 8, 3
121) 7, 8, 6
122) 7, 8, 6, 3
123) 7, 8, 6, 3
124) 7, 8, 6, 3
125) 7, 6, 8, 3
126) 7, 1, 8, 2
127) 7, 8, 6, 3
128) 7, 6, 8, 3
129) 3, 8, 7, 6, 1, 4
130) 4, 8, 1
131) 6, 8, 3, 7
132) 6, 8, 7
133) 6, 8, 2, 3, 4
134) 6, 8, 7
135) 6, 8, 4
136) 8, 6, 3
137) 8, 3, 7, 6
138) 8, 6, 3
139) 8, 6, 3, 7
140) 8, 7, 3
141) 8, 7, 6
142) 8, 7, 6
143) 6, 7, 8, 3
144) 4, 2, 1, 6, 8, 5, 7
145) 8, 3, 7, 6
146) 7, 8, 6, 3, 4
147) 7, 8, 6, 3
148) 4, 5, 2, 6, 8
149) 1, 2, 4, 5, 6, 8, 7, 3
150) 2, 1, 4
151) 2, 1, 4
152) 2, 1, 4
153) 5, 2, 1
154) 4= 6, 8, 1, 2, 3, 5, 7
155) 5= 6= 7
156) 1, 2, 4
157) 3, 6, 8, 7
158) 7, 8, 6
159) 7, 3, 8
160) 2, 4, 1
161) 2, 1, 4
162) 2, 1, 4, 5
163) 8, 1, 5, 7
164) 3, 8, 7, 6
165) 4, 1, 2
166) 4, 1, 7
167) 4, 2, 1
168) 4, 2, 7
169) 6, 4, 2
170) 2, 1, 4
171) 4, 1, 2
172) 7, 8, 6, 3
173) 7, 6, 3
174) 8, 3, 6, 7, 4, 5, 1, 2
Kathy Dopp
Oct 27, 2009, 10:46:21 AM10/27/09
to Election Integrity
Jim, you asked:
"Which leads me to ask, what other ways exist for multi-seat
elections for a small civic organization with no parties?
I am asking for resources that detail how it is done. I would
like to be able to present KPFA with other options. I will
ask that you send the procedures directly to me, and
not to this list, as such a discussion would be straying
off the EI topic, and other readers are getting annoyed with
the long IRV debate in general. My email address is
SomeThoughts at Aol dot Com."
There are other far better options for multi-seat elections (any
method is better than the grossly inequitable, nonmonotonic, complex,
virtually manually inauditable IRV/STV methods) that are also precinct
summable and so easier to audit. If you are interested in learning
from actual election methods experts who study such systems, I would
contact Warren Smith and ask to be added to his election methods email
list where lots of persons will answer your inquiries. "Warren Smith"
<warre...@gmail.com>.
Also simple systems that combine at-large with single member districts
using fair compact districting methods work fairly well in municipal
elections, although some new alternative methods can facilitate
proportional representation based on shared beliefs rather than shared
residency, better.
Cheers,
Kathy
On Oct 26, 6:20 pm, SomeThoug...@aol.com wrote:
> The principle addressee of this email is the national Election Integrity
> group.
> There has been boundless discussion of whether or not IRV is
> in principle good or bad, or whether or not people like it. I would like
> to suggest that contributors now focus the disussion on the
> topic of IRV & EI.
>
> That said, I have been participating in a check of a small STV election.
> STV is multi-seat IRV. People vote for several candidates to fill
> several seats.
>
> The case in point is the election to the citizens board of the KPFA
> radio station in Berkeley, CA. There are 2 groups in the election.
> The one I have data for involves 174 volunteer staff members voting
> for 3 staff seats on the board. The vote by registered listeners is
> in the thousands, and I do not have data for that yet.
>
> Much data is posted here:http://www.4shared.com/dir/22095198/b1d9cdca/STAFFBALLOTS.html
>
> Ballot # 0069 looks like this:
>
> Untitled01
> 32KViewDownload
"Which leads me to ask, what other ways exist for multi-seat
elections for a small civic organization with no parties?
I am asking for resources that detail how it is done. I would
like to be able to present KPFA with other options. I will
ask that you send the procedures directly to me, and
not to this list, as such a discussion would be straying
off the EI topic, and other readers are getting annoyed with
the long IRV debate in general. My email address is
SomeThoughts at Aol dot Com."
method is better than the grossly inequitable, nonmonotonic, complex,
virtually manually inauditable IRV/STV methods) that are also precinct
summable and so easier to audit. If you are interested in learning
from actual election methods experts who study such systems, I would
contact Warren Smith and ask to be added to his election methods email
list where lots of persons will answer your inquiries. "Warren Smith"
<warre...@gmail.com>.
Also simple systems that combine at-large with single member districts
using fair compact districting methods work fairly well in municipal
elections, although some new alternative methods can facilitate
proportional representation based on shared beliefs rather than shared
residency, better.
Cheers,
Kathy
On Oct 26, 6:20 pm, SomeThoug...@aol.com wrote:
> The principle addressee of this email is the national Election Integrity
> group.
> There has been boundless discussion of whether or not IRV is
> in principle good or bad, or whether or not people like it. I would like
> to suggest that contributors now focus the disussion on the
> topic of IRV & EI.
>
> That said, I have been participating in a check of a small STV election.
> STV is multi-seat IRV. People vote for several candidates to fill
> several seats.
>
> The case in point is the election to the citizens board of the KPFA
> radio station in Berkeley, CA. There are 2 groups in the election.
> The one I have data for involves 174 volunteer staff members voting
> for 3 staff seats on the board. The vote by registered listeners is
> in the thousands, and I do not have data for that yet.
>
> Much data is posted here:http://www.4shared.com/dir/22095198/b1d9cdca/STAFFBALLOTS.html
>
> Ballot # 0069 looks like this:
>
> The most useful of the files is 2009KPFAstaff.in, which I have appended to
> the end
> of this email.
>
> The good news is that they published a lot of data, including ballot
> images.
>
> The bad news is that the ballot images are hand-copied ballots, all of
> them.
> My guess is that the primary reason for this is to make sure the brand-new
> scanning software reads the ballots clearly. They might have also
> posted images of the original ballots for verification. There are exist
> issues
> of privacy and secrecy that are not fully clear to me. Being able to look
> at
> the ballots is an enormous plus in terms of retracing how the votes were
> counted. In this case, I am not able to say that the ballot images are
> accurate with what I know now.
>
> This is an STV electionhttp://en.wikipedia.org/wiki/Single_transferable_vote
> the end
> of this email.
>
> The good news is that they published a lot of data, including ballot
> images.
>
> The bad news is that the ballot images are hand-copied ballots, all of
> them.
> My guess is that the primary reason for this is to make sure the brand-new
> scanning software reads the ballots clearly. They might have also
> posted images of the original ballots for verification. There are exist
> issues
> of privacy and secrecy that are not fully clear to me. Being able to look
> at
> the ballots is an enormous plus in terms of retracing how the votes were
> counted. In this case, I am not able to say that the ballot images are
> accurate with what I know now.
>
> 32KViewDownload
Kathy Dopp
Oct 28, 2009, 1:26:53 PM10/28/09
to Election Integrity
I hate to reply to my own message but I forgot to mention several good
proportional representation systems that are in use in other countries
including:
1. parliamentary systems, and
2. party list systems which I particularly like.
Each party (but it could be each top-listed candidate or any group)
lists candidates from top to bottom, where the top candidate is the
first to get elected. If the party list receives more votes, then a
2nd and subsequent candidate on that party list may get elected, etc.
So that each party list gets a number of candidates elected that is
proportional to its vote share. With a party list system (could be
called something different), the voter can also vote to change the
order of the candidates on the party list and raise the chances of a
candidate getting elected.
Kathy
On Oct 27, 10:46 am, Kathy Dopp <kathy.d...@gmail.com> wrote:
> Jim, you asked:
>
> "Which leads me to ask, what other ways exist for multi-seat
> elections for a small civic organization with no parties?
> I am asking for resources that detail how it is done. I would
> like to be able to present KPFA with other options. I will
> ask that you send the procedures directly to me, and
> not to this list, as such a discussion would be straying
> off the EI topic, and other readers are getting annoyed with
> the long IRV debate in general. My email address is
> SomeThoughts at Aol dot Com."
>
> There are other far better options for multi-seat elections (any
> method is better than the grossly inequitable, nonmonotonic, complex,
> virtually manually inauditable IRV/STV methods) that are also precinct
> summable and so easier to audit. If you are interested in learning
> from actual election methods experts who study such systems, I would
> contact Warren Smith and ask to be added to his election methods email
> list where lots of persons will answer your inquiries. "Warren Smith"
> <warren....@gmail.com>.
> ...
>
> read more »
proportional representation systems that are in use in other countries
including:
1. parliamentary systems, and
2. party list systems which I particularly like.
Each party (but it could be each top-listed candidate or any group)
lists candidates from top to bottom, where the top candidate is the
first to get elected. If the party list receives more votes, then a
2nd and subsequent candidate on that party list may get elected, etc.
So that each party list gets a number of candidates elected that is
proportional to its vote share. With a party list system (could be
called something different), the voter can also vote to change the
order of the candidates on the party list and raise the chances of a
candidate getting elected.
Kathy
On Oct 27, 10:46 am, Kathy Dopp <kathy.d...@gmail.com> wrote:
> Jim, you asked:
>
> "Which leads me to ask, what other ways exist for multi-seat
> elections for a small civic organization with no parties?
> I am asking for resources that detail how it is done. I would
> like to be able to present KPFA with other options. I will
> ask that you send the procedures directly to me, and
> not to this list, as such a discussion would be straying
> off the EI topic, and other readers are getting annoyed with
> the long IRV debate in general. My email address is
> SomeThoughts at Aol dot Com."
>
> There are other far better options for multi-seat elections (any
> method is better than the grossly inequitable, nonmonotonic, complex,
> virtually manually inauditable IRV/STV methods) that are also precinct
> summable and so easier to audit. If you are interested in learning
> from actual election methods experts who study such systems, I would
> contact Warren Smith and ask to be added to his election methods email
> list where lots of persons will answer your inquiries. "Warren Smith"
>
> read more »
Bob Richard
Oct 28, 2009, 2:47:13 PM10/28/09
to Election Integrity
Kathy Dopp and I appear to be
in general agreement on the value of proportional representation.
However, whether a jurisdiction has a parliamentary system (versus
direct election of the executive branch) has nothing at all to do with
whether it has proportional representation in legislative elections
(versus winner-take-all). The U.K. and Canada (and many other
countries) have a parliamentary form of government combined with
winner-take-all. Ms. Dopp isn't alone in her confusion on this. When I
advocate for PR I often hear people ask "you mean like they have in
Britain?" and/or "but wouldn't you have to have a parliamentary system
to do that?"
Also, some party list systems allow voters to pick individual candidates from within a party list (called "open list"), while others do not (called "closed list").
--Bob Richard
Also, some party list systems allow voters to pick individual candidates from within a party list (called "open list"), while others do not (called "closed list").
--Bob Richard
-- Bob Richard Executive Vice President Californians for Electoral Reform P.O. Box 235 Kentfield, CA 94914-0235 415-256-9393 http://www.cfer.org
-- Bob Richard Executive Vice President Californians for Electoral Reform P.O. Box 235 Kentfield, CA 94914-0235 415-256-9393 http://www.cfer.org
Reply all
Reply to author
Forward
0 new messages