Milwaukee primary voter turnout data inflated including 145% turnout in some wards

[Ralph Shnelvar <ralphs@xxxxxxxxx>]

Whether it is government or private companies that do the programming and/or
administer elections is, to me, irrelevant.

What we should all agree on is that we need total transparency ... and that
means that the gold standard should be paper ballots counted by hand.

The following story from Milwaukee shows why.  

As a computer programmer myself, Jan Kok's guess as to what the programming
bug probably is sounds true.

Ralph Shnelvar



On Thu, 14 Sep 2006 12:55:47 -0600, "Jan Kok" <@gmail.com> wrote:

I bet you guys will love this!

It's hard to say what's really going on here, but there is one programming
bug that has occurred in the past in voting software in other places (
http://www.votersunite.org/MB2.pdf page 7) and might be happening in
Milwaukee: In C, assembly, and some other programming languages, if you
declare a variable to be a signed "short" (16 bits) integer, things go
haywire if the number stored goes over 32767. 32767 + 1 = -32768 (note minus
sign). 32766 + 3 = -32767, etc. If you don't see the minus sign, adding more
votes makes the total seem to decrease: -32767 + 1 = -32766.

Cheers,
- Jan

---------- Forwarded message ----------
From: JOcall7868@xxxxxxx <JOcall7868@xxxxxxx>
Date: Sep 14, 2006 9:07 AM
Subject: [electionwatch2006] Milwaukee primary voter turnout data inflated
including 145% turnout in some wards
To: electionwatch2006@xxxxxxxxxxxxxxx


  [image: The Milwaukee Journal Sentinel Online] <http://www.jsonline.com/>
www.jsonline.com |



City ballots don't add up Totals are much larger than tallies for
candidates; state wants explanation By DAVE UMHOEFER and DERRICK NUNNALLY
dumhoefer@xxxxxxxxxxxxxxxxxxx *Posted: Sept. 13, 2006* A day after the City
of Milwaukee reported a primary election turnout above 80,000 - more than a
quarter of the city's registered voters - a Journal Sentinel analysis found
that the number might be inflated by tens of thousands.
Voter turnout figures in nearly two-thirds of the city's 314 wards are
suspect, and state officials advised the city late Wednesday to recalculate
its numbers. The city missed a 4 p.m. deadline to turn in polling lists and
voter information to the Milwaukee County Election Commission.
It was unclear whether the suspect figures signal problems with individual
candidate totals.
By the city's calculation, only about half the ballots cast in Tuesday's
primary actually included votes in the hottest races - those for sheriff and
attorney general. For example, the city reported 78,801 ballots cast in the
attorney general race in primaries for the two major parties, but vote
totals for the Democratic and Republican candidates combined amounted to
only 40,971. By that count, 37,830 ballots did not include a vote in the
race - a number that political observers regard as obviously flawed.
In the same manner, the city reported 65,581 ballots cast in the Democratic
primary, but only 35,182 votes recorded for the party's sheriff candidates.
State Elections Board chief Kevin Kennedy said Wednesday it appeared a
computer programming glitch likely caused inaccurate totals. But the city
should painstakingly re-examine its totals, using polling-place records, to
be sure no outcome was affected, he said.
Noting that the city is the rare municipality that programs its own
vote-tallying software, Kennedy said: "I'm wondering if the city's vote
totals have been inflated for years."
City Election Commission executive director Susan Edman said Wednesday that
the Journal Sentinel's questions alerted her to the big gap in vote totals.
She also suspected a programming error by the city was to blame.
Edman expressed confidence that the results of any individual race were
unaffected. She said city and county election officials would meet today to
decide how to verify the results.
Turnouts higher than 100% In every election, some of the total ballots cast
don't contain votes for individual races, either, because voters chose not
to weigh in on which candidate they liked or because a mistake invalidated
that part of an individual's ballot. Other voters may have marked only the
party-preference blank before turning in their ballots, mistakenly thinking
they had cast votes that counted for all their party's candidates - a
practice that works in general elections, but not primaries.
Under any circumstance, the divide between ballots cast and votes that count
is typically much smaller than the city's reported totals for Tuesday's
primary. For example, in the 2002 Democratic primary for sheriff won by
David A. Clarke Jr., 84% of total ballots cast included votes that counted
in the sheriff's race.
The problems the newspaper found with the city's canvass Tuesday involved
multiple wards at the same polling places. At virtually all those locations,
the total number of voters was identical across the board in every ward,
according to the city's count - a red flag that the totals were bogus.
The wards did show varied numbers of people registered to vote, which
resulted in four wards coming in with turnouts inexplicably higher than
100%. The highest was at one Sholes Middle School ward, which showed 746
ballots cast and just 513 registered voters - a turnout of 145.4%.
More than 60,000 votes were recorded as cast in the 198 wards where the
duplicated figures appeared. It was not clear Wednesday what the correct
figures were, although Kennedy said his educated guess was that a computer
miscalculation had simply doubled or tripled turnouts at locations that
housed two or three wards.
"We need to understand what is going on here," Kennedy said. "The city needs
to do this (review) in a systematic way."
But Kennedy said voters should not conclude that thousands of votes are
missing or unaccounted for in Milwaukee. His initial review suggested a
computer error involved only total votes, not individual candidate votes.
The inflated vote totals had surprised several political campaigns, which
had expected that a run-of-the-mill primary on a bad-weather Tuesday might
bring a turnout below 15%. But the dramatic gap between total votes and the
ballots cast in key individual races left observers at a loss.
"That's absolutely impossible," said Mary Clare Fagan, a longtime political
consultant who worked for Democratic sheriff candidate Vince Bobot, on
learning how the city's reported totals compared to the turnout for that
race. Bobot lost to Clarke, the incumbent, with the sheriff getting his
biggest margin in the city of Milwaukee.
In the district attorney's race, 32,659 votes were cast for the two
Democratic candidates - less than half the number of votes the city reported
on the Democratic slate overall. The winning candidate in that race, John
Chisholm, said he was surprised to hear of the problem considering the
recent history of accounting snafus in city elections.
"That's just unbelievable, isn't it?" Chisholm said. "Given the problems
they've had, you'd think they would make sure nothing like this happens."
The 2004 election in Milwaukee was plagued by problems with registration and
voting.
Kennedy said statewide turnout Tuesday appeared to be around 16%, and
Milwaukee is always significantly lower than the rest of the state, he said.

In the 198 wards found with suspect totals in the Journal Sentinel's
analysis, the voter-turnout rate amounted to 33.9%, by city figures. This
includes two wards where a total of 650 ballots were counted even though the
number of registered voters in each was listed as zero - an obvious error.
The 116 wards for which the Journal Sentinel found no sign of miscalculation
had a cumulative voter turnout rate of 16.2% - more in line with the rest of
the state.
Kennedy emphasized the importance of the city's getting the figure
corrected.
"I need to know by law how many voters there were," Kennedy said.

*Original Story URL:*
http://www.jsonline.com/story/index.aspx?id=497705



 ------------------------------
*Check out the new
AOL*<http://pr.atwola.com/promoclk/1615326657x4311227241x4298082137/aol?redir=http%3A%2F%2Fwww%2Eaol%2Ecom%2Fnewaol>.
Most comprehensive set of free safety and security tools, free access to
millions of high-quality videos from across the web, free AOL Mail and more.
 __._,_.___

 Please give us your town state and zip code when you post a message. Also
please avoid personal attacks and policy pronouncements and try to stick to
what you have heard or read in your neighborhood or in a publication
regarding opinion in this election.




 SPONSORED LINKS
  Campaign and
election<http://groups.yahoo.com/gads;_ylc=X3oDMTJkb3R2OGdqBF9TAzk3NDc2NTkwBF9wAzEEZ3JwSWQDMTE5Mjc3NDIEZ3Jwc3BJZAMxNjAwMDYwMzc2BHNlYwNzbG1vZARzdGltZQMxMTU4MjQ4MTA0?t=ms&k=Campaign+and+election&w1=Campaign+and+election&c=1&s=27&g=0&.sig=9TBSY9MlLDIQBEZ579fgfg>
 Your email settings: Individual Email|Traditional
Change settings via the
Web<http://groups.yahoo.com/group/electionwatch2006/join;_ylc=X3oDMTJnMjgxOWx1BF9TAzk3MzU5NzE0BGdycElkAzExOTI3NzQyBGdycHNwSWQDMTYwMDA2MDM3NgRzZWMDZnRyBHNsawNzdG5ncwRzdGltZQMxMTU4MjQ4MTA0>(Yahoo!
ID required)
Change settings via email: Switch delivery to Daily
Digest<electionwatch2006-digest@xxxxxxxxxxxxxxx?subject=Email+Delivery:+Digest>|
Switch
to Fully
Featured<electionwatch2006-fullfeatured@xxxxxxxxxxxxxxx?subject=Change+Delivery+Format:+Fully+Featured>
 Visit Your Group
<http://groups.yahoo.com/group/electionwatch2006;_ylc=X3oDMTJlZ3B1ZXF0BF9TAzk3MzU5NzE0BGdycElkAzExOTI3NzQyBGdycHNwSWQDMTYwMDA2MDM3NgRzZWMDZnRyBHNsawNocGYEc3RpbWUDMTE1ODI0ODEwNA-->|
Yahoo!
Groups Terms of Use <http://docs.yahoo.com/info/terms/> | Unsubscribe
<electionwatch2006-unsubscribe@xxxxxxxxxxxxxxx?subject=Unsubscribe>

__,_._,___