North Carolina Election Audits
Random Manual Post Election Audits
Procedures for auditing North Carolina's elections, random selection of precincts to audit, average manpower, hours and costs per election.
~ Audits are one of the most important means for ensuring the accuracy of election outcomes, and for allowing observers to verify that accuracy.~
Pam Smith, President of VerifiedVoting.org
Post Election Audits were mandated by state legislators in August 2005. North Carolina became one of about a dozen states that conducts post election audits in late 2005. Following the loss of an estimated 4,400 votes on paperless machines in Carteret County in November 2004, the state reformed its election laws. In addition to requiring a voter verified paper ballot, the law requires random, manual hand to eye audits to check that the voting machines are counting correctly.
Target precincts are determined by a statistician, not a flat percent. Pamela Smith, President of the national organization Verified Voting, referred to North Carolina's audits in her testimony before the Committee on House Administration’s Elections Subcommittee on March 15, 2007. She noted that:
North Carolina’s provision is unique in that it does not pre-determine a percentage of precincts, but uses a statistician to determine the appropriate quantity for each election....Audits serve to confirm the accuracy of the vote count, which in turn gives voters more confidence in the integrity of the outcome, as many election officials attest:
In North Carolina, Moore County Election Director Glenda Clendenin described the audit as simple, saying it was “no more than a clerical exercise” and that the purpose of the hand count was “to get voter confidence back and make sure it’s right.”7 A Boone County paper described the hand-to-eye count as a “test of accuracy of the machines.”8
Contests to be audited
SECTION 5.(a) G.S. 163‑182.1(b) reads as rewritten:
(1) Provide for a sample hand‑to‑eye count of the paper ballots or paper records of a statewide ballot item in every county. The presidential ballot item shall be the subject of the sampling in a presidential election. If there is no statewide ballot item, the State Board shall provide a process for selecting district or local ballot items to adequately sample the electorate.
Random selection for audits chosen after consultation with a statistician
"The size of the sample of each category shall be chosen to produce a statistically significant result and shall be chosen after consultation with a statistician."
The NC SBOE hired Dr. William Kalsbeek, to direct them in this new territory of "statistically significant". William D. Kalsbeek Ph.D. is a professor in the department of Biostatistics and director of the Survey Research Unit at the UNC School of Public Health. Kalsbeek is a fellow of the American Statistical Association. Dr. Kalbseek's job was to develop a "statistically significant" audit sample for our state. The May 2006 primary was a sort of the Beta Test to use to set parameters for future audits.
Cost and labor for Audits. In North Carolina’s first audit after passage of their new law, when a single race was examined on ballots in 260 precincts, the average cost was $65 per precinct .
Optical scan ballots found easier to audit compared to touch screen paper "trails". North Carolina’s first audit (of the May 2006 primary) showed that paper optical scan ballots can be audited two times faster than the "paper trails" printed on rolls for the touch-screen voting machines. Approximately 8 + paper ballots can be counted per minute, compared to an estimated one or two "paper trail" records per minute.
Don Wright, General Counsel for the North Carolina State Board of Elections advises -
"The number of persons used for the General Election was the same as for the primary.
As you see we followed the recommendation of Dr. Kalsbeek and increased the
number of samples to be counted. The sample counts showed no major problems.
...
Audit counts in North Carolina are:
1. Statistically meaningful
2. Are reviewed by a statistician for concerns and needed improvements
3. Are done promptly without delaying certification
4. Are done at a reasonable cost
5. Keep the confidential voted ballots within the exclusive control of the county boards of elections
6. And are non-partisan "
As a participant in the May and November 2006 audits in Forsyth County NC, I observed that it took about 1 1/2 hours each time, to perform the audits, including set up time.
NC's election audit - for the Nov 2008 General Election
An Assessment of the Recount and the Certification of the Election Result for the November 2008 Election Post-election audit results North Carolina audited the vote counts in 273 “precincts/places” under the methodological direction of Dr. William D. Kalsbeek, Professor of Biostatistics at the University of North Carolina. Kalsbeek has coauthored a report on the results
Obama won by about 14,000 votes out of over 4.3 million ballots cast –
'the statistical probability that Obama in fact defeated McCain in the North Carolina election is higher than 99.9 percent'
...
The 95% confidence interval for estimated total actual vote for the Election Day
winner and loser: Since manual recount is only available in selected precincts, the
total actual vote count is unknown. However, we can estimate this vote count and
provide a 95% confidence interval for this actual total vote count. If the reported
total vote count is within this 95% confidence interval, we then can claim the
reported total vote count is correct and can represent the actual total vote count. (p.11)
NC's first election audit - for the May 2006 primary
The first post election audit was conducted shortly after the May 2, 2006 primary election. The audit was not random because the targets were selected prior to the election. This was due to an unintentional gap in original legislation that failed to specify that audit targets must be selected after the election and at a public meeting. (That gap was corrected in other legislation one year later).
On April 28th, 4 days before the NC Primary, counties received an email from the State Board of Elections -
"The statewide ballot item to be counted for the May 2, 2006, Primary will be the office of the Wainwright Seat on the N.C. Supreme Court. The sample count will be conducted after primary election day, and will be part of the canvass procedure."
The Random Selection of Your Samples to be Recounted
"At their April 13, 2006 meeting, the State Board of Elections, after consultation with a UNC-CH Statistics PhD., voted that two precincts be selected, at random by the statistician, from each county for the required sample. The statistics provided to the statistician for random selection consisted of all precincts, one-stop sites, and also included mail-in absentees for each county."
Each county needed at least 6-8 people to participate in the audit:
"There will be two bi-partisan teams of three or four for each precinct that represents an equal number from each political party. The members of each team will rotate their responsibilities as caller and tallier. There will be one caller and two talliers. Teams of four will have an observer for the caller.. Observers may witness the vote counting, but cannot interfere.
The NC State Board of Elections provided County Election Directors a list of the precincts that were to be "audited" for the May 2006 primary.
Here is Kalsbeek's first report on North Carolina's May 2006 NC audits, describing his audit protocol and a review of our primary election.
When Warren County NC, a touchscreen county - had huge discrepancies in the May 2006 Pirmary, Kalsbeek saw it -addressed it as an anomaly, but also noted that the county audited the wrong precincts anyway.
Differences in the paper and digital counts of the ballots can be explained by election official error, malfunction of machines, programming errors, ballots mis marked by voters and not read by machines, etc
Here are our own comments about the audit in May 2006 - NC's May 2nd Primary - Uh, Regarding Those Random Audits... More here and a broader overview here at How North Carolina's Elections Will Be Audited
NC's election audit - for the November 2006 General Election
Here is the Nov 8, 2006 memo from the NC SBOE to all county election Directors discussing how the audits will be done, explaining why some counties will have to audit more precincts (to be statistically significant).
Here is the spreadsheet from that November 2006 audit.
Here is Dr. Kalsbeek's report on the Nov 2006 election.
The Public Confidence in Elections Act, ratified August 26, 2005
Audit requirements sections:
SECTION 5.(a) G.S. 163‑182.1(b) reads as rewritten:
"(b) Procedures and Standards. – The State Board of Elections shall adopt uniform and nondiscriminatory procedures and standards for voting systems. The standards shall define what constitutes a vote and what will be counted as a vote for each category of voting system used in the State...
..For optical scan and direct record electronic voting systems, and for any other voting systems in which ballots are counted other than on paper by hand and eye, those procedures and standards shall do both of the following:
(1) Provide for a sample hand‑to‑eye count of the paper ballots or paper records of a statewide ballot item in every county. The presidential ballot item shall be the subject of the sampling in a presidential election. If there is no statewide ballot item, the State Board shall provide a process for selecting district or local ballot items to adequately sample the electorate. The sample chosen by the State Board shall be of full precincts, full counts of absentee ballots, and full counts of one‑stop early voting sites. The size of the sample of each category shall be chosen to produce a statistically significant result and shall be chosen after consultation with a statistician. The actual units shall be chosen at random. In the event of a material discrepancy between the electronic or mechanical count and a hand‑to‑eye count, the hand‑to‑eye count shall control, except where paper ballots or records have been lost or destroyed or where there is another reasonable basis to conclude that the hand‑to‑eye count is not the true count. If the discrepancy between the hand‑to‑eye count and the mechanical or electronic count is significant, a complete hand‑to‑eye count shall be conducted....
...
SECTION 5.(b) G.S. 163‑182.2 reads as rewritten:
"§ 163‑182.2. Initial counting of official ballots.
....
(b) The State Board of Elections shall promulgate rules for the initial counting of official ballots. All election officials shall be governed by those rules. In promulgating those rules, the State Board shall adhere to the following guidelines:
(1) For each voting system used, the rules shall specify the role of precinct officials and of the county board of elections in the initial counting of official ballots.
(1a) For optical scan and direct record electronic voting systems, and for any other voting systems in which ballots are counted other than on paper by hand and eye, those rules shall provide for a sample hand‑to‑eye count of the paper ballots or paper records of a sampling of a statewide ballot item in every county. The presidential ballot item shall be the subject of the sampling in a presidential election. If there is no statewide ballot item, the State Board shall provide a process for selecting district or local ballot items to adequately sample the electorate. The sample chosen by the State Board shall be of full precincts, full counts of absentee ballots, and full counts of one‑stop early voting sites. The size of the sample of each category shall be chosen to produce a statistically significant result and shall be chosen after consultation with a statistician. The actual units shall be chosen at random. In the event of a material discrepancy between the electronic or mechanical count and a hand‑to‑eye count, the hand‑to‑eye count shall control, except where paper ballots or records have been lost or destroyed or where there is another reasonable basis to conclude that the hand‑to‑eye count is not the true count. If the discrepancy between the hand‑to‑eye count and the mechanical or electronic count is significant, a complete hand‑to‑eye count shall be conducted....
SECTION 5.(c) G.S. 163‑182.5 reads as rewritten:
"§ 163‑182.5. Canvassing votes.
The county board shall examine the returns from precincts, from absentee official ballots, from the sample hand‑to‑eye paper ballot counts, and from provisional official ballots and shall conduct the canvass.
SESSION LAW 2006-192 HOUSE BILL 1024, ratified August 3, 2006
Revisions to have audit targets selected after the election and in public meeting
Audit requirements sections:
SECTION 7.(a) G.S. 163‑182.1(b) reads as rewritten:
"(b) Procedures and Standards. – The State Board of Elections shall adopt uniform and nondiscriminatory procedures and standards for voting systems.
(1) Provide for a sample hand‑to‑eye count of the paper ballots or paper records of a statewide ballot item in every county. The presidential ballot item shall be the subject of the sampling in a presidential election. If there is no statewide ballot item, the State Board shall provide a process for selecting district or local ballot items to adequately sample the electorate. The State Board shall approve in an open meeting the procedure for randomly selecting the sample precincts for each election. The random selection of precincts for any county shall be done publicly after the initial count of election returns for that county is publicly released or 24 hours after the polls close on election day, whichever is earlier. The sample chosen by the State Board shall be of full precincts, full counts of absentee ballots, and full counts of one‑stop early voting sites. The size of the sample of each category shall be chosen to produce a statistically significant result and shall be chosen after consultation with a statistician. The actual units shall be chosen at random. In the event of a material discrepancy between the electronic or mechanical count and a hand‑to‑eye count, the hand‑to‑eye count shall control, except where paper ballots or records have been lost or destroyed or where there is another reasonable basis to conclude that the hand‑to‑eye count is not the true count. If the discrepancy between the hand‑to‑eye count and the mechanical or electronic count is significant, a complete hand‑to‑eye count shall be conducted....
SECTION 7.(b) G.S. 163‑182.2(b) reads as rewritten:
"(b) The State Board of Elections shall promulgate rules for the initial counting of official ballots. All election officials shall be governed by those rules. In promulgating those rules, the State Board shall adhere to the following guidelines:
(1) For each voting system used, the rules shall specify the role of precinct officials and of the county board of elections in the initial counting of official ballots.
(1a) For optical scan and direct record electronic voting systems, and for any other voting systems in which ballots are counted other than on paper by hand and eye, those rules shall provide for a sample hand‑to‑eye count of the paper ballots or paper records of a sampling of a statewide ballot item in every county. The presidential ballot item shall be the subject of the sampling in a presidential election. If there is no statewide ballot item, the State Board shall provide a process for selecting district or local ballot items to adequately sample the electorate. The State Board shall approve in an open meeting the procedure for randomly selecting the sample precincts for each election. The random selection of precincts for any county shall be done publicly after the initial count of election returns for that county is publicly released or 24 hours after the polls close on election day, whichever is earlier. The sample chosen by the State Board shall be of full precincts, full counts of absentee ballots, and full counts of one‑stop early voting sites. The size of the sample of each category shall be chosen to produce a statistically significant result and shall be chosen after consultation with a statistician. The actual units shall be chosen at random. In the event of a material discrepancy between the electronic or mechanical count and a hand‑to‑eye count, the hand‑to‑eye count shall control, except where paper ballots or records have been lost or destroyed or where there is another reasonable basis to conclude that the hand‑to‑eye count is not the true count. If the discrepancy between the hand‑to‑eye count and the mechanical or electronic count is significant, a complete hand‑to‑eye count shall be conducted. |
