Back to Board of Elections and Ethics main page
Government and People
After-Action Report for the November 2008 General Election
Report prepared by the Council Board of Elections and Ethics Investigation Special Committee
April 21, 2009
Mary M. Cheh, Councilmember, Ward 3
Council of the District of Columbia
After-Action Report for the November 2008 General Election
The Special Committee and Scope of this Report
On September 16, 2008, the Council of the District of Columbia established the Council Board of Elections and Ethics Investigation Special Committee (“Special Committee”) to investigate irregularities in the District’s Primary Election and make recommendations for reform of the administration of elections in the District. Since its creation, the Special Committee has held hearings related to the District’s elections processes and procedures, issued a preliminary report and recommendations, and taken steps to conduct a forensic investigation of the irregularities that appeared during the District’s Primary Election.
On November 4, 2008, with the assistance of lawyers from the D.C. Trial Lawyers Association, the Special Committee canvassed precincts throughout the District of Columbia to identify any potential problem areas or opportunities for reform, and to experience, first-hand, the challenges involved in administering an election.
While the overall investigation is ongoing, the Special Committee prepared this “afteraction report” to provide its findings about the conduct of the November General Election. In this report, the Special Committee focuses solely on matters that relate to the administration of the General Election. Unfortunately, at the present time, the Board of Elections and Ethics does not generate a comprehensive after-action report. The Special Committee therefore created this document in an effort to catalog the major statistics and the issues observed by the Special Committee.
The Special Committee will continue to solicit input from the public, and is currently in the process of preparing a comprehensive report that will identify areas for legislative and institutional reform.
In November 2008, the District of Columbia held a General Election, in which approximately 266,871 voters, or 62.53% of total registered voters, cast their ballots. As most predicted, the District had very high turnout in the morning, with some lines forming hours before the polls opened. Nonetheless, due to low voter turnout during the evening, the overall turnout was somewhat lower than expected. Some initially estimated that turnout would approach almost 80%. But the actual turnout was still higher than the 230,105 votes cast in the 2004 General Election and 205,748 votes cast in the 2000 General Election, representing total turnouts of 59.94% and 58.1%, respectively.
In this election, 27,955 voters cast absentee ballots, representing 10.48% of the total number of votes. Of those absentee voters, 14,201 voted absentee by mail (representing 5.32% of the total number of votes), 12,364 voted in person (4.63%), and 1,390 voted for federal offices pursuant to the Uniformed and Overseas Citizens Absentee Voting Act (0.52%).
Finally, the Board reported that 14,713 voters cast “special” ballots (sometimes referred to as provisional ballots), and of those special ballots, 10,544 were counted, representing 3.95% of the total number of counted votes. Of the 4,169 special ballots that were rejected, 996 were rejected because the voters voted in the wrong precinct and 3,172 were rejected because the voters were not registered.
In anticipation of the November General Election, voters in the District were excited about the historic nature of the election, but were hesitant about the District’s readiness to administer the election. Although the election ran relatively smoothly by almost all accounts, several problems, discussed below, reflect the need for improvement.
If a voter is unable to vote at the polls on the day of the election,1 the District permits that voter to cast an absentee ballot either by mail or in-person at the Board’s office. According to the Board, 19,109 absentee ballots were mailed, of which 14,201 ballots were returned and counted.
According to the Acting Executive Director of the Board, Sylvia Goldsberry-Adams, the Board has traditionally handled the processing and mailing of absentee ballots internally. In other words, a voter would send an absentee-ballot request to the Board and the Board would mail the ballot directly to the voter.
In an apparent effort to reduce administrative costs, the Board in 2008 agreed that Sequoia Voting Systems would print and deliver absentee ballots to voters. Thus, once the Board received an absentee-ballot request, the Board would forward the request to Sequoia, and Sequoia would print a ballot and mail it to the voter. In the days before the election, however, several dozen residents contacted various D.C. offices, including the Council of the District of Columbia, indicating that they had not received their absentee ballots. By most accounts, the Board was ill-equipped to handle the volume of complaints, and many voters complained about receiving inaccurate or incomplete information from Board employees.
Though the Special Committee is not currently in a position to conduct a comprehensive cost-benefit analysis of the decision to contract out absentee ballot delivery services, should the practice continue in the future, substantial improvements must be made to ensure the reliable and timely delivery of absentee ballots.
On October 26, 2008, the Special Committee learned that a voter in Single-Member District 2F03 (“SMD 2F03”) received an absentee ballot that did not have the Ward 2 Council or the Ward 2 School Board races on the ballot. Although the Board initially claimed that the voter received the correct ballot because SMD 2F03 was partially located in Ward 6 (and thus would not have Ward 2 races on the ballot), that claim proved to be untrue; no part of SMD 2F03 is in Ward 6. The Special Committee learned that the Board’s database contained inaccurate data about SMD 2F03’s jurisdictional boundaries.
Having discovered that the ballot was invalid, the Special Committee dispatched staff to review every ballot style to see if any others were invalid. The Special Committee was surprised to learn that, until that point, nobody had attempted to assess the accuracy of the ballot styles for the General Election. After the review, it appears to the Special Committee that the SMD 2F03 ballot style in question was the only invalid style.
The Special Committee later learned, though, that Sequoia also erred. Specifically, the Board had submitted a ballot style request to Sequoia for the proper ballot (i.e., the correct SMD 2F03 ballot), but Sequoia sent the voter an erroneous ballot style. After further investigation, the Special Committee learned that Sequoia may have mailed as many as 126 incorrect ballots to voters.
Thus, two errors were discovered. First, the Board’s database contained erroneous information about SMD 2F03’s boundaries. Second, Sequoia erred in improperly mailing an invalid ballot style to voters.
Problems with In-Person Absentee Ballot Voting
Unlike some jurisdictions, the District does not permit “no-fault absentee ballot” voting or early voting. Current District law requires a voter to either vote in person on Election Day or have a specific reason to cast an absentee ballot.2 Nonetheless, if a voter is eligible to cast an absentee ballot, that voter is permitted to cast an “in-person absentee ballot” (IPAB) at the main office of the Board of Elections and Ethics. In the District, in- person absentee voting begins 15 days before the election, and continues until the day before the Election Day. According to the Board, over 1,000 voters cast IPABs in the first two days of voting, and ultimately 12,364 voters cast IPABs in the two weeks leading up to the election.
During the first few days of in-person voting, the Board faced long lines outside its office on the second floor of the One Judiciary Square Building. According to testimony provided to the Special Committee, the Office of the City Administrator provided unsolicited advice to the Board to move the voting down to a larger space on the first floor of the building. After changing the voting location, the lines were substantially more manageable.
The testimony also suggested that, despite the Board’s independence, the City Administrator’s Office took a substantial role in administering the in-person absentee ballot process—going so far as to actually hand out ballots. If true, this was, at a minimum, “irregular” and perhaps improper. In any case, it highlighted the Board’s inability to handle the situation on its own.
The Board had significant success in poll-worker recruiting. Though the Board initially estimated a need for 2,500 poll workers, the Board ultimately trained 3,388 poll workers for the General Election.
Nevertheless, the Special Committee heard testimony that training was often inconsistent. In addition, many poll workers indicated that training was not tailored to match their experience levels, so that veteran poll workers were provided the same training as new poll workers. According to a report from the Center from Democracy and Electoral Management, university student poll workers who worked as precinct technicians received “very uneven” formal training from the Board, including one instance in which the Board employee responsible for training the students arrived “two hours late and did not bring all of the proper equipment and materials.”3
Another concern arises from the understanding of the Special Committee that the Board does not evaluate poll-worker performance.
Despite the overall sense that Election Day was administered successfully, it was not entirely so. For example, some precinct captains complained that they had not been given sufficient notice of which poll workers were to work at the polls, in some cases until the day of the election. In at least one instance, a pollworker was made a precinct captain “at the last minute.”
Several pollworkers also cited problems with communication between the Board and pollworkers. For example, precinct captains complained that they received no emails from the Board and were unable to get in touch with Board employees. Similarly, the Board failed to “inform precinct captains about student poll workers. . . . [and s]taff and student workers themselves got directly in touch with precinct captains, who gladly accepted the additional workers.”4
Finally, there is reason to believe that Election Day complaints may have been underreported because the Board’s hotline was not sufficiently advertised. The Special Committee learned that many of the precinct technicians were “not aware of a dedicated line to call with polling place issues and were unable to get timely tech support from the [Board] when problems occurred with voting machines.” Further complicating matters, the Special Committee noted that the Board printed hundreds of signs posted at polling places with an incorrect phone number for that dedicated line.
When compared to the February and September elections, the District’s experience with the voting machines was substantially improved in the November election. The Special Committee is of the opinion that having an additional precinct-count optical scan voting machine at each precinct greatly contributed to the overall success of the election. Many precinct captains expressed gratitude that, when one machine failed, voters could use the other machine without significant delay. Thus, technicians were able to respond to voting machine issues without substantially affecting the election.
The only issue with voting machines that arose during the day pertained to confusion with the touch-screen voting machines and the At-Large Council race. Specifically, because voters could select up to two candidates, voters expressed concerns that if they selected only one candidate, the machine warned them that their vote was “incomplete.” Although this was not actually incorrect, it caused some voters to be confused about their options.
One of the major continuing challenges facing the District over the past several election cycles has been that a number of precinct locations are structurally and operationally inaccessible to persons with disabilities. The Special Committee observed that the precincts varied greatly in terms of available space and accessibility. In some cases, the space was so limited that voters could hardly expect to maintain total privacy with respect to their votes.
The Special Committee’s observations were consistent with other accounts as well. Indeed, according to a recent report of the November election, 4.8% of the precincts were structurally inaccessible and 8.7% of the precincts were operationally inaccessible.5 In addition, the report indicated that several other precincts suffered from problems with accessibility, including problems with doorbells and with insufficient turning radii for voting machines (e.g., wheelchair accessibility).
Finally, though the District has continued to expand “curbside voting,” which has increased the accessibility of the District’s elections, it is not an entirely satisfactory alternative because voter privacy cannot be assured.
Before election night, a number of people requested and received permission to observe the election. On election night, though, the Board, citing space restrictions, initially refused to allow all of the observers to view the tabulation process from the office. And, when the Board ultimately did permit observers to view the tabulation process, the observers were restricted to two viewing areas. The Board did not allow observers to view the Board members or senior staff’s review of the unofficial election results, and the Board did not make any attempt to explain the tabulation process. Furthermore, the Board’s failure to adopt consistent policies about observer conduct led to disputes on election night, including one instance where an observer was prevented from taking photographs of the vote-tabulation process.
In the Preliminary Report released before the General Election, the Special Committee recommended that the Board “design and implement a protocol to ensure timely releases of unofficial election results without sacrificing accuracy.”
The Special Committee suggested that “the Board would do well to release the results in smaller batches.” The report also warned the Board that:
In the aftermath of the primary election, there may be a tendency to proceed overly cautiously in the release of results. It would be a mistake, however, to unnecessarily slow the release of results. Though accuracy must be ensured, by unnecessarily or unreasonably delaying the release of results, the public’s confidence in the Board’s ability to efficiently process vote tabulation could be further eroded.
Unfortunately, only half of this advice appears to have been taken by the Board. Although the Board released the results in smaller increments, it did not communicate the process that it would use to review the election results. Further, the Board took a very long time to release the first batch of results. The first results, with 20 precincts of 143 reporting, came almost 2 ½ hours after the polls closed, and more than an hour after the voting cartridges had already been processed.
More troublingly, though, was the lack of a clear, publicly available checklist for evaluating the accuracy of election night review of the unofficial election results. According to a letter submitted to the Special Committee on the day of the election, the Board “developed a checklist of checks and balances for review of the unofficial election night results [to be] completed and attached to each report prior to release of any information to the public.” However, that checklist was never provided, as an attachment to each report or otherwise, either to the Special Committee or to the public.
Though further investigation is necessary, the Special Committee has significant concerns about the roughly 4,169 special ballots that were rejected. According to anecdotal evidence heard by the Special Committee, there is reason to believe that the voter registration roll in the District of Columbia is both over-inclusive and under- inclusive. In particular, the Special Committee understands that the voter rolls have a large number of voters who have moved or otherwise no longer are eligible to vote, but remain on the voter list. Furthermore, the Special Committee received a very large number of complaints from voters who claimed to have timely registered but learned that their registrations had not been processed.
As described above, the Board successfully recruited a large number of poll workers for the November election. Nonetheless, the Committee received complaints from poll workers who did not receive compensation in a timely manner. Though the Committee has not precisely determined the cause of the untimely payments, the Committee is of the opinion that one of the causes may be the lack of a comprehensive poll-worker database.
In many respects, the November General Election was a success. The Board’s use of additional voting machines at the polling place and the Board’s success in recruiting and training poll workers contributed to the overall efficiency of the voting process. Moreover, the implementation of redundancies in the voting tabulation process and adoption of ballot auditing rules increased the public’s confidence in the integrity of the voting results. Despite these gains, there is still substantial room for improvement. The lack of clear guidelines and process workflows led to missteps ranging from the Board’s inability to respond to misdelivered absentee ballots to the late payment of poll workers. In order to improve the management of the next election, the Board should critically evaluate what elections processes demand clearly defined workflows, establish those workflows, drill its employees on the workflows, and make every effort to ensure that the Board and the public understand them.
The Special Committee is continuing to investigate potential legal reforms based on its observations during the election. Major recommendations and legislative proposals will be part of the Committee’s final report, which is scheduled to be issued in May. However, in the interim, the Special Committee proposes the following programmatic improvements to the District’s administration of elections.
1. ESTABLISH INTERNAL OFFICE WORKFLOWS FOR ROUTINE TASKS, SUCH AS PURCHASING RELEVANT SUPPLIES AND PROCESSING POLL-WORKER PAYMENTS, AS WELL AS CHECKING THE DESIGN AND ACCURACY OF BALLOT CONFIGURATIONS
2. PUBLISH ANY WORKFLOWS, CHECKLISTS, PROCEDURES, OR PROTOCOLS USED BY THE BOARD FOR ADMINISTRATION OF ELECTIONS ONLINE IN ADVANCE OF THE ELECTION, AND SOLICIT FEEDBACK FROM THE PUBLIC
3. WORK WITH THE OFFICE OF THE CHIEF TECHNOLOGY OFFICER (OCTO) TO DEVELOP A POLL-WORKER DATABASE
4. CREATE CLEAR GUIDELINES REGARDING ACCESS FOR ELECTION OBSERVERS
5. CONTINUE TO DEVELOP STRONGER RELATIONSHIPS WITH GROUPS SUCH AS THE LEAGUE OF WOMEN VOTERS, CENTER FOR DEMOCRACY AND ELECTION MANAGEMENT, THE POLLWORKER INSTITUTE, AND OTHERS TO CONTINUE TO IMPROVE POLL-WORKER RECRUITMENT AND TRAINING
6. IMPROVE THE SUMMARY-REPORT-SIGNOFF CHECKLIST, TO INCLUDE COMPARISONS WITH PRECINCT TAPES AND BALLOT ACCOUNTING SHEETS, AND PUBLISH THE SIGNED CHECKLIST WITH THE SUMMARY REPORTS
1 D.C. law permits a registered voter to cast an absentee ballot only if the voter: “is temporarily outside the territorial limits of the District”; “is unable to vote at the regular polling place due to a permanent or temporary physical handicap, disability, or illness”; “is confined in an institution for mental treatment within or outside the District, but not judicially declared incompetent”; “is incarcerated in an institution, but not for conviction of a felony”; “is a Board employee who is necessarily absent from the precinct of registration on election day”; “resides outside the United States, and has registered to vote pursuant to the Uniformed and Overseas Citizens Absentee Voting Act, but only for candidates for federal offices”; “expects to be absent from the District of Columbia on election day”; “expects to be hospitalized on election day”; or “declares personal religious tenets prevent the voter from voting on election day.” 3 D.C.M.R. § 715.2.
2 D.C. Official Code § 1-1001.09(b).
3 American University's Center for Democracy and Election Management, Final Report of the Poll Worker Recruitment and Training Project 2008, at 3 (January 2009).
5 University Legal Services, Inc., Accessible voting in DC—Still beyond reach: Individual precinct results of accessibility survey for the November 4, 2008 General Presidential Election (Jan. 16, 2009).
Back to top of page
Send mail with questions or comments to firstname.lastname@example.org
Web site copyright ©DCWatch (ISSN 1546-4296)