To view graphic version of this page, refresh this page (F5)

Skip to page body

Department of Elections

Department of Elections Report of the 2000-2001 San Francisco Civil Grand Jury


SUMMARY

The Department of Elections (DOE) introduced new election technology in the November 2000 general election. The Civil Grand Jury (CGJ) found that this technology appeared to work adequately. We found DOE employees conscientious, knowledgeable and willing to help.

The CGJ conducted interviews, reviewed previous reports regarding DOE, and observed the November 2000 election, December 2000 runoff election, Supervisorial District 7 recount, several meetings of the Citizen's Advisory Committee on Elections, and a hearing at the Board of Supervisors. We also reviewed the DOE election problem database and tabulated these results.

The CGJ recommends that:

  • DOE prepare written guidelines regarding election observation and handling of electioneering.
  • DOE review provisions to ensure ballot privacy.
  • DOE update pollworker and staff training.
  • DOE update voter instructions.
  • DPH recommit to monthly transmittal to DOE of voter roll updates.
  • The Board of Supervisors review options to handle unscheduled and potentially costly runoff elections.
  • DOE update and use the election problem database to analyze statistics. A pollworker pre-election checklist could preclude many problems identified in the election problem database.
  • DOE printed materials have version dates.

BACKGROUND

The CGJ investigated the operations of the Department of Elections (DOE). This investigation was begun, in part, due to DOE intentions to use new technology (the EagleTM optical scanning system) for the November 2000 election instead of punch cards. Also, the district election system, enacted by San Francisco voters in November 1996 (to elect supervisors by geographical region rather than at large by all City voters), was implemented in the November 2000 election.

Our review encompassed the November 2000 election, the December 2000 runoff election, and the Supervisorial District 7 recount. In addition, we attended several meetings of the Citizens Advisory Committee on Elections (CACE) and met several times with DOE management. We also attended several Board of Supervisors hearings regarding the proposed DOE reorganization and allegations regarding the November 2000 election.

We reviewed several previous investigations of DOE: the 1997-1998 Civil Grand Jury report; the CACE Special Report on DOE (sent by DOE to Supervisor Teng June 16, 2000); and the California Secretary of State Report (released January 26, 1998) regarding allegations of fraud in the San Francisco election of June 3, 1997. Additionally, the CGJ received input from citizens in the form of one complaint to the CGJ regarding the election of November 7, 2000, and several reports from the San Francisco Fair Vote Committee.

INVESTIGATION

This section is presented chronologically, by CGJ activity.

Presentation to CGJ prior to November 2000 Election

DOE management and staff gave a presentation to the CGJ prior to the November 2000 election. DOE appeared to have a good plan in place to educate the public regarding the new Eagle optical scanning equipment. DOE brought along an Eagle unit to the courthouse for demonstration. The Eagle system appeared to be very easy to use and have satisfactory backup systems to ensure adequate voting protocols, if used properly. The CGJ made suggestions to the proposed ballot template regarding highlighting of admonitions to vote for only one candidate, which DOE management stated they would take under advisement (and which DOE subsequently implemented for the November 2000 election).

November 2000 Election

We contacted DOE several weeks prior to the November 2000 election to notify DOE that the CGJ would like to observe vote counting activities at City Hall. DOE stated that accommodations would be made to enable CGJ review.

However, the evening of November 7, it was not clear what planning, if any, had been made to accommodate CGJ review. The CGJ members were issued badges, which clearly identified them as observers from the CGJ, then were asked to wait in the North Light Court where someone from DOE staff would meet them. The CGJ waited approximately one hour, during which we asked DOE staff several times when we would be able to proceed to the vote counting areas. After waiting one hour without accommodation, we proceeded to DOE offices in the City Hall basement. DOE staff were kind enough to put out chairs behind a rope, from which we were informed we could sit and observe vote counting activities. The CGJ observers noted that, from the specified location, very little of actual DOE activities could be observed and monitored. DOE staff informed the CGJ that this was the type of observation opportunity given to members of the public and was essentially the only option available.

The CGJ asked to talk with DOE management. Instead, a member of Administrative Services management came to talk with the CGJ. After some discussion, several CGJ members were given walk-through tours of vote count areas; however, the CGJ was not permitted to stay in any of the areas for long-term observation. When the CGJ attempted to view Eagle memory pack upload activities, we were asked to remain behind a rope at least 25 feet from the upload station, and were escorted two at a time over to the station for a short look at these activities. Essentially, the CGJ was given little opportunity to observe canvassing activities over any significant amount of time.

It is understandable that DOE management and staff's first priority is to make sure that an election complicated by district elections, important initiatives, and new voting technology, goes well. However, it is equally necessary that the public have confidence in the integrity of the vote counting process, especially with elections involving these same factors. Accommodation of qualified observers is fundamental to ensuring transparency of the vote counting process.

The CACE Special Report on DOE discusses the issue of DOE policy changes that impeded the efforts of several observers to monitor vote counts. The CACE Report section "Transparency of Canvassing" observes that DOE "...has not developed written standard operating procedures...to provide adequate monitoring of the vote counting process while preserving the integrity of the process and minimizing interference with the canvassing duties of the Department. In recent elections, since the return of the Department to City Hall, observers' movements have been restricted (due, in part, to the layout of the basement of City Hall)."

The California Elections Code, Sections 335.5, 353.3, 15104, 15204, and 15360 state that the canvass is to be a public process (texts provided in Attachment 1). Section 15104 particularly notes that:

"...any member of the county grand jury...shall be permitted to observe and challenge the manner in which absentee ballots are handled, from the processing of absentee ballot return envelopes through the counting and disposition of the ballots."

Section 15104 further notes that:

"...observers shall be allowed sufficiently close access to enable them to observe and challenge whether those individuals handling absentee ballots are following established procedures, including all of the following: (1) Verifying signatures and addresses by comparing them to voter registration information. (2) Duplicating accurately any damaged or defective ballots. (3) Securing absentee ballots to prevent any tampering with them before they are counted on election day....".

While it is not clear if absentee ballots were being counted as part of the canvass process the evening of November 7 (it is possible they were counted in the following several days), it remains clear that there is a need for written policies regarding vote count monitoring. Our investigation has identified several other California counties that do have such written policies in place.

Subsequent to November 7, the CGJ discussed with DOE management the lack of written policies regarding accommodation of groups wishing to observe and monitor vote counting processes such as campaign representatives, media, official government observers (Secretary of State representatives or the CGJ), representatives of public interest groups, and the general public. DOE management agreed that written guidelines for both DOE staff and for such observers was needed and should be prepared. DOE management stated that it appeared to make sense to take a several tier approach to level of observation, with official groups being given more latitude in observation activities.

Election Statistics

DOE maintains a database that documents election process problems and their resolutions. The problems are also documented on hard copy by way of a Problem and Incident Reporting (PAIR) form (see Attachment 2). The PAIR form specifies code numbers to be used for specific problem types (e.g., Code 21 is used when someone calls to note a problem with a locked polling place), and these code numbers are used in the database. The CGJ asked if any group had ever analyzed these data to see what trends, if any, could be determined. DOE did not know of anyone who had ever done this. The CGJ determined that analysis of the data could possible identify areas where attention could profitably be directed for improvement.

The CGJ requested a copy of the election problem database for the November election (and the December runoff election), then tallied the problems by type. The data are not perfect. The database contained many duplicate entries and some problems had been mislabeled with one code type when another code clearly was more proper. For example, for the November election, Code 27 (which is not defined on the PAIR) was used as a "grab bag" for a variety of problems, all of which could be placed into already defined categories. In order to prepare and review statistics from the printouts, duplicate entries were deleted and recategorization was performed as necessary.

The results are presented in Attachment 3, Table 1.

Some observations from the data include:

  • The "Pollworker Manual - Election 2000," pp. v-viii, has a detailed checklist for closing the polls. However, there does not appear to be a pre-election checklist for use by precinct inspectors. If a detailed pre-election checklist were used by the inspector to verify materials, it might prevent many of the problems identified on the database.
  • The December election was presumably easier than the November election due to fewer items on the ballot and fewer overall voters going to the ballot box, yet the number of pollworker problems was significantly higher.
  • The number of polling places that were locked was consistent across the two elections (21 in November, 20 in December). San Francisco has 647 voting precincts; therefore, the data suggest that there may be a consistent 3% of polling places that are locked on Election Day. This issue should be strongly considered for review and resolution by DOE.
  • Overall, there appears to be a definite learning curve regarding the Eagle optical scanning system. The December election had slightly fewer than one-third of the problems experienced during the November election. (Note: November election personnel specified individual Eagle problems with sub-codes in some cases, but for the December election all Eagle problems were lumped into one category.)
  • The CGJ was interested to note that there were 64 calls regarding electioneering problems in the November election, and 31 calls for the December election. It is worth noting that the 1998 Secretary of State Report also notes electioneering to be a problem, and suggests that DOE obtain the name and description of potential electioneering violators when visiting precincts in order to assist law enforcement in any follow-up investigations. As noted by the Secretary of State, no acts of electioneering are acceptable. However, the Secretary needs definitive evidence identifying violators before he can pursue charges. Currently, DOE only records the name of the campaign associated with electioneering.

Citizen Complaint

Following the November 2000 election, the CGJ received a citizen complaint which voiced concern that voting booths had not yet been set up at their polling station when they came to vote at approximately 8 a.m. As a consequence, the complainant and others at the polling station were required to vote out in the open while standing or while sitting on couches or chairs, without privacy. The complainant also notes that their marked ballots were visible to pollworkers taking the ballots.

As an additional comment regarding the secrecy of voting, one member of this CGJ had the experience of a pollworker at their polling station ask for their ballots as they came to the Eagle unit. The pollworker then tore off the voting stubs and gave one ballot at a time (the November election had a total of three ballot sheets) to the CGJ member to feed into the Eagle unit. We note that a February 2001 Report and Recommendations from the Maryland Secretary of State Special Committee on Voting Systems and Election Procedures states that voters in jurisdictions using optical scan voting systems often complain that pollworkers can view their ballot as it is inserted into the unit, and that this voter concern should be emphasized during pollworker training.

Ballots used in the optical scanning process are much larger than the cards previously used in the punch system. Unlike the punch system, in which it was difficult to ascertain what had been voted, the optical scanner ballots are quite easily readable if privacy of the voting booth and use of secrecy envelopes are not maintained. Further, The California State Constitution, Article 2, Section 7, states, "Voting shall be secret." Maintenance of ballot secrecy is a fundamental cornerstone of voting rights.

The issues of polling place equipment and pollworker training are addressed further in the "Findings and Recommendations" section below.

December 2000 Runoff Election

Several members of the CGJ observed the December 2000 supervisorial runoff election activities. Several CGJ members observed activities at City Hall, and several members observed activities at Taraval Station. Taraval Station was a collection point for ballots and memory packs, as well as an uplink site for reading of memory packs.

During the counting process for the runoff, members of the CGJ were allowed much greater latitude in observing election and vote counting processes at City Hall. This may be due to a combination of our discussions with DOE and Administrative Services management in November, and DOE staff not having as many demands on their time due to better familiarity with the Eagle system processes and the lesser volume of ballots voted. No significant findings were observed at City Hall. As noted in "Election Statistics" above, data from this election are presented in Attachment 3, Table 1.

Several observations were made at Taraval Station. As a follow-up to ballot secrecy issues from the November 2000 election, the CGJ members asked several pollworkers if they had observed any ballot secrecy issues with the new ballots. The pollworkers agreed that ballot secrecy was a problem, both with respect to the large ballots that had the potential to be easily read and the secrecy envelopes, which did not always ensure privacy.

The CGJ also stayed to observe closing of the polling station, receipt of ballots and memory packs from other polling stations, and uplink processes from the Taraval station to City Hall. The CGJ was impressed by how quickly the ballots and memory packs were delivered to Taraval Station. We observed that personnel performing the uplink process initially had problems fitting the memory packs into the slot for reading. After declaring several packs to be "fried", they discovered that the tolerances for fitting the packs into the reading slot were fine, and that they just had to take a little more time to put the packs more deeply into the slot and more firmly on the pin connections.

Both of these issues will be discussed below in the "Findings and Recommendations" section.

Lastly, it appears that San Francisco will face the certainty of runoff elections in future supervisorial elections. Such runoff elections can be expensive. San Francisco should give consideration to methods to reduce such costs, such as the "instant runoff" method used recently in elections in England.

Supervisorial District 7 Recount

In the certified runoff vote count from December 2000, incumbent Mabel Teng lost to challenger Tony Hall by 39 votes out of 18,627 votes cast. By letter dated January 4, 2001, the San Francisco Neighbors Association requested that DOE perform a recount. The recount officially began Friday, January 12, 2001. DOE released the final recount tally on January 31, 2001, showing a net gain of one vote for Teng.

The CGJ observed some portion of almost every day of the recount process. We were accorded the same observer status given to observers from the California Secretary of State and CACE. The representative from the Secretary of State's office attended much of the first several days of the recount process. There appeared to be at least one person present from the A. Phillip Randolph Institute (a public group) for virtually all of the recount process. We note that there appeared to be much less monitoring on the part of CACE in comparison to these other groups and the CGJ.

The recount process was well organized and, for the most part, proceeded smoothly. DOE staff were quick to answer questions and resolve issues as necessary for both campaigns as well as for observers. For example, changes to the process were made as necessary during the recount. At the request of the campaign representatives, changes were made to the way in which DOE staff held up each ballot for examination. Also, originally only one representative for each campaign had been issued a "challenge" card to hold up when challenging a ballot; at the request of the campaigns, "challenge" signs were also given to the second official observer for each campaign.

The recount process showed that the Eagle optical scanning system is very accurate. The manufacturer states a 99% accuracy, but the recount results indicate much better accuracy than that.

The recount process was also beneficial in that it also brought forth several other issues that require resolution by DOE. We have summarized the issues in Attachment 3, Table 2.

Board of Supervisors Hearing, May 31, 2001

Close to the end of this Grand Jury's term of service, DOE Acting Director Dr. Phillip Paris sent a letter dated May 17, 2001, to City Attorney Louise Renne in which several allegations were made regarding DOE.

The allegations were:

  • Signing of fraudulent time sheets.
  • Unauthorized commitment of fund expenditures in excess of $1.6 million.
  • Improper certification of the November 2000 election results.
  • Inability to account for approximately 3,600 ballots from the November 2000 election.

The CGJ attended a Board of Supervisors committee hearing on May 31, 2001, at which the issue of the vote discrepancy was discussed. The Supervisors noted that further hearings would be held to obtain information as necessary about the votes and the other allegations. The following discussion summarizes testimony presented at the hearing.

The vote discrepancy arises from a mismatch between what is termed the "voter/voted" number and total number of ballots cast. The voter/voted number is developed by reviewing precinct rosters, documenting the number of persons signed in to vote at each precinct, then adding up the various precinct data. The final voter/voted tally is apparently not used in any official manner; however, the roster data are used to update DOE files that document individual voting histories. The total number of ballots cast is acquired from the Eagle data and is published in the final DOE Statement of Vote.

In January 2001, a citizen requested the voter/voted and ballots cast numbers from DOE, compared them, and identified a mismatch of 9,007 votes (324,031 ballots cast, and 315,024 voters as derived from roster data). This discrepancy was brought to the attention of DOE management, who asked DOE staff to investigate. In simple terms, it appeared that there were 9,007 more ballots cast than persons who voted.

The number of ballots cast, as determined by the Eagle system, has a high degree of reliability, as demonstrated during the recount process. In contrast, DOE staff investigation determined that there were several problems related to the voter/voted data obtained from the rosters. The method by which roster data are obtained is as follows. After the election, the rosters are brought back to DOE offices. There is a bar code next to each name on the roster, and for each roster signature the appropriate bar code is scanned by a hand-held device ("wanded"), which is then downloaded to a larger machine. Temporary workers are used to perform this scanning process. There are apparently a number of things that can possibly go wrong during this process, including:

  • Failure of voter to sign roster, or failure of pollworker to have voter sign roster.
  • Failure of voter to sign in manner that clearly documents intent to vote (e.g., some voters apparently use an "x" or just a simple line), thus making it difficult for DOE staff to later determine if the mark is random or if the voter actually voted.
  • Loss of a roster (during the hearing, DOE staff noted that two rosters from the November election may have been lost and, therefore, not scanned).
  • A voter signs the roster, but leaves before actually voting.
  • Rosters placed in the wrong bin at DOE offices (and thus assumed to have already been scanned).
  • Failure of temporary workers to appropriately scan rosters, perhaps due to inattention or fatigue.
  • Failure to appropriately download roster from hand-held device.
  • Data for provisional voters, who by definition cannot sign the precinct roster, must be input by hand (with the attendant possibility of human error).

A DOE staffperson investigated and resolved several of these issues and reduced the voter/voted tally and Statement of Vote difference to 3,496 ballots. (This apparently is the source of the number used in Dr. Paris' letter, though mis-stated there as "approximately 3,600 ballots"). The staffperson also noted that some precincts showed more voters than ballots, and some less voters than ballots. If one takes the absolute value of these differences, rather than sums the positive and negative differences, the overall discrepancy is on the order of 8,700 ballots. These results were discussed with DOE management in early February. Apparently, DOE management did not request this staffperson to perform any further analysis for this issue following the early February discussion.

Testimony was presented at the hearing that other California counties apparently have the same general problem with the voter/voted numbers not matching the Statement of Vote. However, even though there are many reasons to cause the voter/voted tally to be inaccurate, it should also be noted that the 3496 vote discrepancy is greater than 1% of the total number of votes identified in the Statement of Vote (3494 / 324,031 = approximately 1.1%). Though the voter/voted totals are not used in an official capacity, given the past allegations of vote problems for DOE and given the several close candidate and initiative races in the November 2000 election, the somewhat large discrepancy could give rise to suspicions that indeed more votes were cast than there were people voting.

Following the issuance of Dr. Paris' letter, the Secretary of State's Office has initiated an investigation into the voter/voted and ballots counted number mismatch, and the California Attorney General's Office has initiated an investigation into the alleged financial malfeasance issue. The Chief Administrator has also contracted with a retired judge to perform an independent evaluation. These investigations are expected to conclude following the end of this CGJ's term of office.

Follow-up to Issues from Previous Audits

The 1997-1998 CGJ Report, the 1998 Secretary of State, CACE, and Fair Vote Reports all mention failure to update voter rolls as a potential issue. In response to the 1997-1998 CGJ Report, the Department of Public Health (DPH), Bureau of Records and Statistics, had committed to send DOE a diskette on a monthly basis with which to update the voter roll database.

Discussion with DPH staff the first week of June 2001 indicated that, until December 2000, information had been sent monthly to DOE. However, personnel changes at DPH evidently prevented any transmission of data after that time. Discussion with DOE staff in the same time frame confirm that finding. A follow-up discussion several days later with DOE staff indicate that a quite large transmittal of information had been received from DPH. We contacted DPH, who confirmed the transmittal of information; however, the transmittal contained only all deaths in 2000, which was sent to ensure that DOE had all the pertinent 2000 data. Due to staffing constraints, DPH staff have not yet had time to input any 2001 death certificate information into the database; therefore, no data regarding any 2001 deaths have yet been sent to DOE. DPH states that they expect to be able to begin inputting the 2001 data into the database sometime in July 2001.

DOE staff also receive database update information from the California Secretary of State, the California Department of Motor Vehicles, the United States District Court (felony listings), and other Registrars. DOE staff state that, since January 2000, they have updated the DOE database six times to incorporate the information received from these sources.

FINDINGS

Our findings are listed below, with notations where a similar concern has been identified elsewhere. Recommendations follow in the next section.

  1. DOE does not have adequate guidelines in place regarding how groups are permitted to observe vote counting processes, nor which types of groups have what levels of access.
    • Identified by CGJ during November 2000 election
    • Also a concern noted in the June 2000 CACE Report
  2. DOE needs to review secrecy envelope provisions to ensure ballot privacy for the larger Eagle system ballots.
    • Identified by CGJ during November 2000 election
    • Citizen complaint, November 2000 election
    • Identified by CGJ during December 2000 runoff
    • Also a concern noted in the 1998 Secretary of State Report
  3. DPH is not meeting its commitment to send monthly voter roll updates to DOE.
    • Identified by this CGJ (as follow-up item)
    • Also a concern noted in the 1997-1998 CGJ Report
    • Also a concern noted in the June 2000 CACE Report
    • Also a concern noted in the 1998 Secretary of State Report
    • Also a concern noted in the Fair Vote Report
  4. Currently, DOE only records the name of the campaign associated with electioneering. However, in order to pursue any cases of electioneering, the Secretary of State must have definitive evidence identifying violators.
    • Identified by this CGJ in analysis of election database statistics section
      Also a concern noted in the 1998 Secretary of State Report
    • 5. Pollworker training needs to be updated
  5. Pollworker training did not prevent potential ballot secrecy issues
    • Supervisorial District 7 Recount indicated problem with incorrectly processed provisional ballots by pollworkers
    • Possible problem with voters signing rosters noted in Board of Supervisors hearing
    • Potential loss of rosters noted in Board of Supervisors hearing
    • Pre-election checklist could avoid many problems in the election database (pollworker manual has detailed post-election checklist, but pre-election material is spread across several paragraphs).
  6. The Board of Supervisors should move forward with legislation to remedy potentially costly runoff elections.
    • Observation from December 2000 runoff election
  7. DOE staff training needs to be updated
    • Identified by CGJ during December 2000 runoff
    • Supervisorial District 7 Recount, see Attachment 3, Table 2
    • Board of Supervisors Hearing, wanding and roster handling
    • Also a concern noted in the June 2000 CACE Report
    • Potential loss of rosters noted in Board of Supervisors hearing
  8. DOE election database statistics appear to be potentially useful data by which to identify further areas for review for possible improvement
    • Database apparently does not allow for automatic summation of data
    • DOE staff sometimes make duplicate or inaccurate data entries
  9. Voter instructions appear to be inadequate
    • Supervisorial District 7 Recount, see Attachment 3, Table 2
    • Board of Supervisors Hearing, roster signing
  10. Some DOE printed materials do not have revision dates on them (for instance, the PAIR form).
    • CGJ finding
    • Also a concern noted in the Fair Vote Report

RECOMMENDATIONS

  1. The CGJ recommends that DOE prepare written departmental guidelines regarding how groups are permitted to observe vote counting processes, and which types of groups have what levels of access.

    Required Response
    Department of Elections - 60 Days

  2. The CGJ recommends that DOE review secrecy envelope provisions to ensure ballot privacy for the Eagle system ballots.

    Required Response
    Department of Elections - 60 Days

  3. The CGJ recommends that DPH recommit to monthly transmittal to DOE of voter roll updates. The CGJ acknowledges that budget restrictions may impact the ability to meet this previous commitment. However, maintenance of the voter rolls is an important function. Therefore, we encourage DPH and DOE to work together towards a solution. Possible methods for resolution of this recommendation might include:

    • DPH and DOE jointly hiring a temporary or full-time information specialist, who would be responsible for database update and maintenance in both departments:
    • DPH hiring temporary workers to handle data update
    • DOE staff review daily obituary notices to flag names in the voter rolls, until such time as the next data transmittal from DPH is received.

    Required Response
    Department of Public Health - 60 Days
    Department of Elections - 60 Days

  4. The CGJ recommends that DOE initiate an electioneering policy or procedure by which both the names of persons and the campaigns are recorded such that there is definitive evidence identifying violators.

    Required Response
    Department of Elections - 60 Days

  5. The CGJ recommends that pollworker training needs to be updated regarding the following issues identified in this report:

    • potential ballot secrecy issues
    • incorrectly processed provisional ballots by pollworkers
    • proper methods for voters to sign rosters
    • traceable control of election materials, including rosters
    • pre-election checklist


    Required Response
    Department of Elections - 60 Days

  6. The CGJ recommends that the Board of Supervisors move forward with legislation to remedy potentially costly runoff elections.

    Required Response
    Board of Supervisors - 90 Days
    Department of Elections - 60 Days

  7. The CGJ recommends that DOE staff training be updated regarding the following issues identified in this report:

    • Guidelines for preparation of duplicate ballots
    • Handling of ballots
    • Wanding of rosters
    • Positive traceability of election materials, including rosters
    • Handling of Eagle memory packs during upload process


    Required Response
    Department of Elections - 60 Days

  8. The CGJ recommends that DOE continue to use the election problem database, and analyze statistics. Personnel updating the database should be given guidance as to avoiding duplicate or inaccurate information. The PAIR codes should be reviewed and updated (e.g., Eagle codes used during elections do not appear on current PAIR form). Items currently logged as "other" under each major subcategory should be reviewed to see if new categories are advisable.

    Required Response
    Department of Elections - 60 Days

  9. The CGJ recommends that voter instructions be reviewed for possible update regarding the following issues identified in this report:

    • More warnings regarding the consequences of overvoting.
    • Actions to avoid when filling out absentee ballots (do not sign, cut in half, etc.).
    • Reminder to have current signature on file at DOE, or absentee ballot could possibly be rejected for signature mismatch.
    • How to properly sign precinct roster.


    Required Response
    Department of Elections - 60 Days

  10. The CGJ recommends that DOE printed materials have revision dates on them (for instance, the PAIR form).

Required Response
Department of Elections - 60 Days

ATTACHMENT 1
APPLICABLE SECTIONS OF THE CALIFORNIA ELECTIONS CODE

335.5 The "official canvass" is the public process of processing and tallying all ballots received in an election, including, but not limited to, provisional ballots and absentee ballots not included in the semifinal official canvass. The official canvass also includes the process of reconciling ballots, attempting to prohibit duplicate voting by absentee and provisional voters, and performance of the manual tally of 1 percent of all precincts.

353.5 The "semifinal official canvass" is the public process of collecting, processing, and tallying ballots and, for state or statewide elections, reporting results to the Secretary of State on election night. The semifinal official canvass may include some or all of the absentee and provisional vote totals.

15104. (a) The processing of absentee ballot return envelopes, and the processing and counting of absentee ballots shall be open to the public, both prior to and after the election. (b) Any member of the county grand jury, and at least one member each of the Republican county central committee, the Democratic county central committee, and of any other party with a candidate on the ballot, and any other interested organization, shall be permitted to observe and challenge the manner in which the absentee ballots are handled, from the processing of absentee ballot return envelopes through the counting and disposition of the ballots. (c) The elections official shall notify absentee voter observers and the public at least 48 hours in advance of the dates, times, and places where absentee ballots will be processed and counted. (d) Absentee voter observers shall be allowed sufficiently close access to enable them to observe and challenge whether those individuals handling absentee ballots are following established procedures, including all of the following: (1) Verifying signatures and addresses by comparing them to voter registration information. (2) Duplicating accurately any damaged or defective ballots. (3) Securing absentee ballots to prevent any tampering with them before they are counted on election day. (e) No absentee voter observer shall interfere with the orderly processing of absentee ballot return envelopes or processing and counting of absentee ballots, including touching or handling of the ballots.

15204. All proceedings at the central counting place, or counting places, if applicable, shall be open to the view of the public but no person, except one employed and designated for the purpose by the elections official or his or her authorized deputy, shall touch any ballot container. Access to the area where electronic data processing equipment is being operated may be restricted to those persons authorized by the elections official.

15302. The official canvass shall include, but not be limited to, the following tasks: (a) An inspection of all materials and supplies returned by poll workers. (b) A reconciliation of the number of signatures on the roster with the number of ballots recorded on the ballot statement. (c) In the event of a discrepancy in the reconciliation required by subdivision (b), the number of ballots received from each polling place shall be reconciled with the number of ballots cast, as indicated on the ballot statement. (d) A reconciliation of the number of ballots counted, spoiled, canceled, or invalidated due to identifying marks, overvotes, or as otherwise provided by statute, with the number of votes recorded, including absentee and provisional ballots, by the vote counting system. (e) Processing and counting any valid absentee and provisional ballots not included in the semifinal official canvass. (f) Counting any valid write-in votes. (g) Reproducing any damaged ballots, if necessary. (h) Reporting final results to the governing board and the Secretary of State, as required.

15360. During the official canvass of every election in which a voting system is used, the official conducting the election shall conduct a public manual tally of the ballots tabulated by those devices cast in 1 percent of the precincts chosen at random by the elections official. If 1 percent of the precincts should be less than one whole precinct, the tally shall be conducted in one precinct chosen at random by the elections official. In addition to the 1 percent count, the elections official shall, for each race not included in the initial group of precincts, count one additional precinct. The manual tally shall apply only to the race not previously counted. Additional precincts for the manual tally may be selected at the discretion of the elections official.

ATTACHMENT 2
DEPARTMENT OF ELECTIONS
PROBLEM AND INCIDENT REPORTING (PAIR) FORM

ATTACHMENT 3, TABLE 1
STATISTICS DERIVED FROM ELECTION PROBLEM REPORTS

Problem Type

Problem and Incident Reporting Form Code

November 2000

December 2000

POLLWORKER

Need Inspector

11

30

33

Need Clerks

12

68

94

Need Bi-lingual Pollworker

13

9

8

Other

14

46

61

Subtotal

 

153

196

       

POLLING PLACE/EQUIPMENT

Polling Place Locked

21

21

20

Can't Locate Booths/Equipment

22

28

7

Port-a-Potty Issue

23

3

2

Electricity Off

24

2

3

Other

25

37

18

Subtotal

 

91

50

       

EAGLE

Eagle Trouble

26

165

70

Ballot Jam

26.1

57

- na -

Tape Out/Tape Won't Print

26.2

4

- na -

Memory Pack

26.3

5

- na -

Memory Pack

26.4

1

- na -

Subtotal

 

232

70

       

SUPPLIES

No Ballots

31

22

9

Wrong Ballot-type

32

4

1

Running Out of Ballots

33

29

1

No Roster

34

11

3

No Provisional Envelopes

35

10

1

No VIP (voter information pamphlet)

36

1

1

Stationary Supplies

37

14

2

Need PM Pick-up

38

77

35

Other

39

3

42

Subtotal

 

171

95

       

ELECTIONEERING/OTHER

Electioneering

41

3

9

Media/Pollster Interference

42

0

1

Parked Car

43

1

0

Campaign Signs

44

24

15

Other

45

36

6

Subtotal

 

64

31

       

MEMORY PACK

Pack Not Read

53

- na -

1

Pack at Wrong Uplink Sit

54

- na -

1

"Fried" Pac

62

- na -

7

Late Pac

63

- na -

1

Subtotal

 

10

ATTACHMENT 3, TABLE 2
ISSUES FROM THE SUPERVISORIAL DISTRICT 7 RECOUNT

ISSUE

FINDING AND RECOMMENDATION

DOE STAFF TRAINING ISSUES

A duplicate ballot is made whenever the Eagle unit cannot read a ballot. DOE does this for absentee ballots; the voter can rectify ballots voted on Election Day. Absentee ballot problems caused by voter include:

  • Ripped/torn ballots
  • Stains (coffee, other)
  • Ballots cut in half, then either left that way or stapled back together
  • Use of red/blue ink, or pencil
  • Creases/folds
  • Ballot signed by voter
  • Candidate name circled rather than arrow filled in, or hole punched in ballot to indicate preference

FINDING: It is possible that better or clearer instructions to people voting by absentee ballot could reduce the types of problems seen.

RECOMMENDATION: Instructions to voter need to be reviewed for possible improvement.

One remade duplicate by DOE staff clearly should not have been made. The original ballot was an undervote (no vote for either candidate), and furthermore had been signed by the voter. DOE staff remade a duplicate and assigned the vote to Teng. A second duplicate had not been stamped "DUPLICATE" as required. On a third, the original ballot looked OK and it was not obvious why the duplicate had been made.

FINDING: There were several DOE staff errors regarding duplicate ballots.

RECOMMENDATION: Instructions to DOE staff need to be reviewed for possible improvement, or prepared where none are written.

Evidently, one precinct listed one more vote for Teng in the original certification than had been found during the recount. DOE stated that they were sure the vote was good since they had documentation that DOE had received the ballot; the ballot probably was misfiled. DOE would continue to look for the misfiled ballot, and in any case would credit it toward the Teng vote total.

FINDING: A ballot was misfiled.

RECOMMENDATION: Instructions to DOE staff need to be reviewed for possible improvement, or prepared if none are written.

There was substantial organization of ballots by DOE staff. During the recount, for several precincts, there were long runs of ballots for each candidate, much more than would be accounted for by statistical variation.

FINDING: There was substantial organization of ballots by DOE staff.

RECOMMENDATION: Instructions to DOE staff need to be reviewed for possible improvement, or prepared if none are written.

VOTER GUIDANCE ISSUES

There were a fair number of absentee ballot overvotes (an overvote is when a voter votes for two or more candidates in the same race).

FINDING: There are a fair number of absentee ballot overvotes.

RECOMMENDATION: Instructions to voter need to be reviewed for possible improvement.

There was evidence that several people voted twice; DOE had both a provisional and absentee ballot from these people. There were several of these two-ballot situations for each campaign. Counsel for Teng cited specific sections of law indicating that the law requires both votes to be eliminated from the count. DOE stated they currently only cancel the second vote and let the first stand.

DOE management states that they have turned over 10 names to the San Francisco District Attorney for further investigation and possible prosecution. Note that double voting is a repeat finding from the 1998 Secretary of State Report.

FINDING: DOE policy may not be consistent with state law.

RECOMMENDATION: DOE should review state law applicability to policy regarding canceling only one vote instead of both.

DOE does not notify a voter if that voter's absentee ballot has been rejected. Rejection may be due to voter's signature not matching the signature on file. Some people do not know that they may need to update their signature at the DOE office if they change signature style.

FINDING: Absentee voters are not informed if their ballot is rejected.

RECOMMENDATION: DOE should review policy regarding possible notification of voter following rejection of absentee vote.

POLLWORKER GUIDANCE ISSUES

Some provisional ballots from Laguna Honda Hospital (LHH) clearly used LHH identification cards for the voters, but no addresses were put on the provisional envelopes. DOE guidelines require an address for provisional ballot; a ballot is rejected without one. DOE states that they need such guidelines so that staff do not get bogged down in tracing down information for each ballot cast.

FINDING: Though some LHH residents voted provisionally, some ballots were rejected because the address information box was not filled out.

RECOMMENDATION: DOE should review instructions to pollworkers filling out provisional ballots so they know that the ballot will be rejected if the boxes are not filled out.


Last updated: 4/3/2012 11:22:19 AM