Utah's New Election Audit and Recount Procedures Found Lacking by Utah's Desert Greens Party and Utah Count Votes
Desert Greens/Green Party of Utah
www.GPUT.org
CONTACT: Kathy Dopp, kathy@utahcountvote.org, 435-658-4657
Thursday, October 26, 2006
"Election Policy" Adopted October 17, 2006 by the Office of the
Lt. Governor of Utah
Utah's Desert Greens Party, the Utah affiliate of the Green Party of
theUnited States (gp.org), and Utah Count Votes, an election integrity
group, oppose Utah's new election procedures for vote count audits and
recounts. After a careful review of Utah's new internal audit procedures for the upcoming November 7, 2006 general mid-term election,
Utah Count Votes and the Desert Greens Party feel that Utah's newly adopted election policies weaken Utah's election processes and fail to
ensure Utah's election integrity.
Their concerns include:
The uncertainty of using volatile invisible electronic ballots that
computer experts have shown are open to vote fraud and glitches makes
manual audits important to give the public confidence in election results. Public comment was not solicited for
Utah's new audit and recount procedures and no qualified mathematicians, degreed
statisticians, or expert auditors were advisors to the Lt. Governor to
develop these newly adopted election policies.
Major problems they identified with Utah's new audit and recount
procedures include:
1. Not Verifiable by the Public
No auditable report of election results (counts on each voting
machine for each race or issue) is publicly released prior to selecting
which counts will be audited, so the audit could be manipulated (by adjusting election results numbers to match
aggregated vote totals, despite not matching detailed machine counts).
The counts to audit may be randomly selected prior to poll
closure and announcement of election results, and therefore the audit could be manipulated (by manipulating machine vote counts to alter
election outcomes that are known will not be audited).
Not publicly observable (the public is not permitted to observe
either the random selection of the counts, or the manual audits).
No public access to records needed to verify audits.
Results of the audit are not required to be publicly released
2. Not Transparent
Random selection process not is not publicly observable. Random
selection process is not transparent random selection (Uses easily manipulated software and no measures for verifying the selection
software are included ? could be hacked to avoid certain voting machines.)
Random selection method is not transparent.
Audit procedures are not required to be publicly observable.
The public and even the candidates themselves do not have access to observe any of the procedures (neither the audit, nor the random
selections).
Not verifiable by the public (see above)
Results of the audit are not required to be publicly released
3. The Audit Results Could be Easily Undetectably Manipulated
Auditors are not randomly selected and are not randomly
assigned to specific counts
Auditors are not independent ? election officials are the auditors.
Election Officials are free to manipulate the audit results
without observation
Lt. Governor's Office in particular and, depending on who the
Lt. Governor's office tells which machines are to be audited during the
election, others, could be free to manipulate un-audited vote counts not selected for audit prior to the end of the
election (due to possibly selecting which vote counts will be audited prior to the end
of the election by the Lt. Governor, and before the announcement of election results and without any public auditable report that the
public could use to verify the audit.)
Early voting machines and voting machines used in multi-precinct polling locations have a higher probability of being
selected. Absentee ballots and paper optical scan ballots (which are trivially easy to tamper with) have a much lower
probability of being selected for audit, so could be targeted for vote fraud.
4. Not Independent ? i.e. not conducted by independent auditors
5. Mathematically Insufficient for Ensuring Election Outcome Integrity
1% is insufficient in any close contests or contests with a
small number of vote counts
Differing probabilities of selecting machines ?
lower probability of selecting machines w/ single precincts
lower probability of selecting optical scan counts with
multiple precincts ? procedure slightly unclear ? but unit sizes varied
Probability of detecting outcome-altering miscount is
hopelessly low in recounts required under Utah law for races with margins under 1%
Early voting machines and voting machines used in multi-precinct polling locations have a higher probability of being
selected. Absentee ballots and paper optical scan ballots have a much
lower probability of being selected for audit, so miscounts in these types would be more likely to be missed.
6. Subverts the Intent of the Legislature and the Public regarding
Recounts.
In races where margins between candidates are less than 1%,
these procedures call for a simple audit rather than an entire hand recount.
A recount is merely a re-tabulation of DRE memory cards (which
we saw in Cuyahoga County, were off in 10% of the total votes counted where 72% of the total vote counts were off by at least one vote; and
here in Utah, they denied access to do any investigation of our primary
election and did not warn voters to use absentee ballots like in Cuyahoga and
elsewhere).
7. Lack of Consequences - No Teeth
No provision or process is provided for expanding the audit
when discrepancies are found, and
No provision for correcting election results when discrepancies
are found
Utah's Desert Greens Party and Utah Count Votes believe that the Utah's
Election Policy could be improved by taking the following steps:
1. Audits should be conducted by independent auditors.
2. The public should be allowed to observe all audit procedures,
including selection and manual vote counts.
3. A sufficient number of vote counts should be audited calculated
to assure that any outcome-altering vote miscount would be detected.
4. Any recount should have a 100% manual audit, as per the public
and the Utah Legislature's intentions.
5. Diebold optical scan (OS) system counts should be audited
separately from touch-screen (TS) counts.
6. Random assignment of auditors and auditors to counts as per the
Brennan Center's recommendations should be used.
7. Recounts and audits should be publicly verifiable which
necessitates that a public report of votes counted on each machine
should be publicly released prior to randomly selecting the counts to
be audited.
Utah's new audit procedures are an internal audit, conducted without
permitting public observation or verification, by the same election officials who count the votes and handle the ballots.
Is there any other industry that permits industry insiders to conduct
its own audits, yet determines who controls such large budgets?
Since Utah's new audit procedures allow the Lt. Governor's office to
select the voting machines that will be audited during Election Day, before the election is over, this information could be used by
unscrupulous individuals within the Lt. Governor's office to manipulate
the vote counts on unselected voting machines.
Utah Count Votes and the Utah's Desert Greens Party believe that
auditing of the voting machines should be publicly observable and transparent, and that the determination as to which machines are
audited be made after polls are closed and a detailed auditable report
of each machines vote counts is publicly released.
Kathy Dopp, Founder and President of Utah-based National Election Data
Archive, and Joycelynn Straight of Utah Clear have developed an alternative election audit proposal and are seeking sponsors in the
Utah legislature for independent election audit legislation that meets
the transparency, verifiability, and sufficiency conditions.
Utah's election officials reiterated in their October 17, 2006
"Election Policy" that their plan to keep secret all the detailed
election records, including the two electronic and two paper records of
Diebold vote counts and then destroy them after 22months - never to look at them. Without access to these records, Utah election officials
will not be able to do their job of ensuring that Utah election results
are accurate and auditors will not have access to all the detailed election records needed
to conduct a thorough investigative audit. Utah Count Votes has an appeal scheduled on November 9th at 1:30 p.m.
before the Utah State Archives and Records Service to try to gain public access to these election records needed to evaluate Utah
election integrity.
This detailed response by the Utah's Desert Greens Party and Utah Count
Votes to the Election Policies implemented on October 17, 2006 by the Office of the Utah Lt. Governor can be found at
http://utahcountvotes.org/ltgov/Response2LtGov-Audit-Recount.pdf