** Attachments may require Adobe Acrobat Reader to view .pdf files. Click here to download Adobe Acrobat Reader for free.
|April 9, 2002
TO: EACH SUPERVISOR
FROM: Conny B. McCormack, Registrar-Recorder/County
Following my oral report to your Board on March 12, 2002, motions were introduced by Supervisors Zev Yaroslavsky and Michael D. Antonovich requesting a written report describing the difficulties encountered in administering the March 5, 2002 Primary Election, plans for preventing recurrence of the problems, and identification of additional resources needed to implement those plans .
This report responds to those
Board motions and provides a context for examination of the problems experienced
on election day. While the report principally focuses upon addressing
the dilemmas relating to changes in assigned voting locations and recruitment/retention
of precinct poll workers (and emergency replacement of an unprecedented
number who cancelled), the report also describes recent major changes
in election laws that have significantly and fundamentally altered the
environment for conducting elections now and into the future.
The problems experienced in the March 5, 2002 Primary Election cannot be analyzed in a vacuum. This election was conducted under new and unusual circumstances that greatly contributed to voter and poll worker confusion including: 1) a difficult redistricting process, significantly foreshortened due to the first Gubernatorial Primary held in March instead of June, 2) the confusing new "Modified Closed" Primary election rules that, for the first time, allowed voters registered as "decline to state" or non-partisan to cast ballots in some, but not all, political party primaries, and 3) the first statewide election held under the truncated 15-day voter registration deadline.Some of the resulting problems were predicted prior to the election in a news release  issued in conjunction with a press briefing the Registrar-Recorder/County Clerk (RR/CC) held in Los Angeles on February 20, 2002. The Registrars of Voters from Alameda, San Bernardino and Riverside Counties also participated in this press briefing in a show of solidarity highlighting the statewide nature of anticipated problems in administering this election.
The complexity of the "Modified Closed" Primary format, which replaced the popular Open or Blanket Primary rules that were in place for the 1998 and 2000 primary elections , confused voters and poll workers alike (colorful graphics are illustrative and found at Attachment 3. The new format, combined with the effects of political jurisdictions' redistricting following the 2000 U.S. Census, taxed our existing systems and staff to the limits of their ability to cope with polling location shortages and last minute poll worker cancellations of unprecedented proportions.
Additionally, the March 2002 Primary Election was the first statewide election since the law was changed shortening the voter registration deadline by two weeks from 29 days to 15 days prior to every election. This greatly impacted election preparations as lists of registered voters by precinct must be printed 19 days before the election in order to complete distribution of election supplies to the County's 4,845 voting locations .
As is fully described later in this report , the most significant problem that occurred in administering the March 5th election was the unprecedented number of Precinct Inspectors (supervising poll workers in charge at each voting location) who cancelled their service at the last minute. This resulted in late opening of 121 polling locations  because the RR/CC lacked sufficient emergency response mechanisms to accommodate the quadrupling of cancellations compared with past elections. Post-election research has involved telephoning cancelled inspectors to get their feedback on reasons for decisions to cancel their previously agreed upon service and that data is included herein.
The RR/CC department needs additional financial resources for poll worker pay incentives, replacement of 30-year old precinct election supply boxes and additional staff positions. As described on pages 19-21 of this report, augmented staffing and poll worker pay increase proposals were identified in a management audit of the Department by the outside consulting firm, Strategica, in a comprehensive report released in November 2000.
The Department subsequently submitted in February 2001, in conjunction with the FY 2001-2002 County budget process, a proposal requesting appropriate staff to adequately address the year-round requirements of coordinating and expanding newer areas of poll worker recruitment from the ranks of County employees, private businesses, high school and college students. This strategy is essential to both supplement and be prepared to replace the rapidly aging veteran force of poll workers, many of whom have served for decades but are now unavailable in record numbers due to retirement, physical infirmity or death.
After funding of new positions was denied in FY 2001-2002, pared down proposals were compiled and presented over the past year to the CAO and to Board offices as detailed in Attachment 5 (.pdf file). Additional resources are also needed to expand the Adopt-A-Poll and Poll Worker Academy  programs that the department initiated in 2001as pilot projects but which have been unsustainable due to lack of staff to assign to coordinate the development and maintenance of such new programs.
Background and Unusual Circumstances Surrounding March 5, 2002 Primary
Impact of Redistricting
Once per decade, following the U.S. Census, Federal and State laws require that all political jurisdictions redraw their boundary lines to ensure that the population of each district (congressional, state assembly, state senate, board of supervisors, city councils, schools, etc) is approximately equal.
The impact of redistricting is felt at the local level - the County. The redistricting process is not complete until County Registrars receive all new district lines and adjust the voting precinct boundaries to avoid bisecting any of the new political districts' lines (a massive task in Los Angeles County given the huge number of 5,000 voting precincts). Redistricting always has an impact on the elections immediately following implementation, but the affect on the March 2002 Election was more widespread than ever due primarily to the State Legislature's decision not to "nest"  two State Assembly Districts contiguously within the boundaries of each State Senate District.
A decade ago following the last redistricting, even though "nesting" of State Senate and Assembly districts was respected and when the 1992 Primary Election was held in June instead of March, our systems were barely able to process the data and re-draw precinct boundary lines in time to complete mandated election events. In 2001/2002, the process was cut short by a full three months due to the first ever Gubernatorial Primary Election held in March. The 2001 precinct redistricting included an unprecedented number of changes described in a memo sent by the RR/CC to your Board in November 2001 .
Voters by and large are oblivious to the redistricting process until many confront the fact that their polling place has changed (precinct boundary changes as a result of redistricting caused approximately 20% of L.A. County voters to be assigned to an unfamiliar polling place). Our post-election research has now revealed that most voters are unaware that when a new State Assembly, State Senate or Congressional boundary line, for example, has divided their neighborhood, requiring adjustment of the precinct boundaries accordingly as precinct lines cannot by law bisect any political jurisdictions' boundaries. The consequence for a number of voters was assignment to a new voting location different from their immediate neighbors. Several examples of complaints from voters expressing anger about having their voting location changed are illustrative of this problem and are included as attachments to this report .
While some problems resulting from statewide redistricting were predicted, the magnitude was not foreseen including: 1) the re-precincting process fell behind schedule, delaying the required relocation of hundreds of new voting sites, and 2) the number of distinct ballot combinations (unique set of candidates and propositions dependent upon geographic area and political district boundaries) dramatically increased. This explosion of ballot combinations in all Counties ultimately resulted in sample ballot booklet printing and delivery delays statewide and soaring sample ballot printing costs. In Los Angeles County these costs more than doubled from any previous countywide election (the former highest sample ballot printing bill was $1.4 million; the estimated cost for the March 5, 2002 election sample ballot booklets is $3 million) .
Changes in precinct boundaries have a "ripple" effect throughout the entire election preparation process. As the number of precinct boundary changes increase, so do the number of polling places that must be relocated and the number of poll workers who are reassigned to abide by election law that mandates poll workers' assignment to their "home" precinct whenever possible.
Confusing "Modified Closed Primary" Election Format
Much of our pre-election planning involved designing and disseminating outreach materials to educate voters and poll workers about the radically-changed voting rules for the Primary Election:
In spite of these efforts, post-election telephone surveys to precinct Inspectors and analysis of poll worker training class attendance statistics has now revealed that a significant number of poll worker cancellations was due to poll worker concerns that they would be unable to answer voters' questions about the unfamiliar new election rules (compared with the Open/Blanket Primary format used for the past two primary elections that was very popular as it allowed all voters a complete range of choice among candidates for all political parties). Evidence of poll workers harboring such concerns is apparent when examining the number of poll workers who attended training classes compared with the number who actually served on election day. Traditionally, almost 100% of poll workers who attend a pre-election training class show up to serve on election day. However, for this election, of the 15,788 poll workers who attended training, only 13,863 actually served.
Post-election research has also revealed that Los Angeles County was not alone among California counties experiencing unexpected and surprisingly high rates of poll worker cancellations and election day no-show rates. For example:
Additionally, in a conversation this month with the State of Washington's Election Director, the RR/CC discovered that the confusing format of that state's last presidential primary election resulted in poll worker mutiny and defections of unprecedented proportion. Indeed, the Washington Election Director stated that a large number of their experienced poll workers are now on record as stating that they are willing to continue as poll workers in every type of election except future presidential primaries conducted under their complex rules.
The pervasive nature of these similar experiences forebodes a repeat of such problems in future primary elections. This concern is heightened by the fact that California's next scheduled statewide "Modified Closed" Primary is the March 2004 Presidential Primary Election. Assuming the Democratic and Republican parties continue to allow decline-to-state/non-partisan voters to participate in their primary elections, non-partisans will be allowed to vote on those parties' candidates for congress, state assembly, and other partisan offices with the exception of the presidential contest. This prohibition from voting on the top of the ticket inevitably will lead to a level of non-partisan voter confusion and anger that can be expected to be directed at poll workers on election day - a situation that is bound to cause a crisis in poll worker recruitment and defections in high numbers of those who agreed to serve prior to learning about this onerous restriction.
Problem: Polling Place
Shortages, Cancellations and Location Changes
Polling locations must meet a set of criteria before being accepted. Criteria include adequate size, availability/permission to use, wheelchair access, adequate parking, appropriate lighting, available restroom facilities, etc. Inevitably, some polling places change with each election. Private businesses change ownership, buildings are torn down or undergo major reconstruction, and schools schedule functions that make facilities unavailable. Additionally, locations that are not wheelchair accessible are being phased out as accessible facilities become available.
However, as explained above, the number of new polling places that needed to be recruited for the March 5, 2002 Primary Election was higher than ever before. Additionally, many previously available polling locations refused permission for use, an unprecedented number canceled at the last minute after permission had been previously secured, and polling location recruitment got off to a late start awaiting the release of new legislative boundary lines following redistricting.
For decades, one of the most reliable sources for polling places has been the schools. We have been particularly hard-hit by changes in school availability (37% of the polls that cancelled were schools). Loss of schools are due to:
Poor Management Decision-Making
Post-election analysis recently revealed that Departmental mid-level management decisions, unconveyed to the Department Director or Chief Deputy, were poorly conceived including waiting until precinct boundary changes were completed in early January 2002 following the redistricting process to make polling place selections. It is now clear that this was far too late in the election preparation process to finalize polling place selection, especially when more than 20% of County precinct boundaries were so significantly altered that different polling places were required.
The delay in the poll site selection process adversely affected the most crucial factor in poll site selection - availability. The result was many "undetermined" polling places at 46 days prior to the election when software to match voters' names with the appropriate sample ballot booklets must be sent to the printer to begin the addressing of 4.1 million booklets. This resulted in subsequent mailing of sample ballot booklets  to approximately 10-15% of voters containing the statement "no polling place has yet been established - watch for a postcard containing voting location". Because many of these postcards were unable to be mailed until 10-14 days prior to the election, they were co-mingled with the avalanche of last minute candidate mailers. We now believe this resulted in many voters overlooking the polls address postcards even though these postcards contain an eye-catching patriotic design. 
As the deadline for finalizing sample ballot printing neared and a significant number of polling sites were still undetermined, virtually the entire staff of the RR/CC Polls Section was deployed to locate polling sites prior to the sample ballot mailing. This is the same staff that would normally have been telephoning and handling correspondence with poll workers during this time period. Ultimately, the overlap of tasks during the weeks prior to the election overtaxed staff to the point that key functions were competing for identical resources and the work processes fell behind. We now realize that an infusion of additional temporary staff should have been hired to accommodate these processes simultaneously due to the unusual circumstances of the once a decade redistricting process colliding with the Gubernatorial Primary election date three months earlier than ever before.
Changes in management reporting are now in place to ensure that key resources and timelines are monitored at a higher level. Benchmarks are being developed that will trigger appropriate warnings to upper management so that additional staff and other resources can be requested and deployed when needed earlier in the process.
Voter Confusion Due to Voting Location Changes
Although all 4.1 million voters received information on their polling place address either on the back cover of their sample ballot booklet or on a subsequent polls address postcard, (and, if those materials are misplaced, can find it interactively by entering their address on the RR/CC website or similarly via the Department's 200 interactive phone lines) voters are creatures of habit and often assume that their voting locations are unchanged. Consequently thousands appeared at poll sites that were no longer where they had been in past elections.
Voters are informed of last minute polling place changes that occurred after mailings of sample ballot booklets via signs posted at the previously selected polling place in order to direct wayward voters to the correct location. However, it is not possible to post signs at every location that has ever been previously used as a polling place for a County or City election.
Election day confusion was compounded by a large number of voters who called our office, your offices and City Clerks to complain that their polling place was "not open" when, in fact, the voters went to a location where they had previously voted but was not assigned as a polling place for the March 5th election. Post-election research has revealed that almost 50% of the "polls not open" reports received by our office were from such confused voters (this number was reported to your Board on March 12th as 236 polls opening late when in fact the actual number was 121 as shown on Attachment 4).
We were also dismayed to learn that even some of the poll workers assigned to serve in this election neglected to read their appointment notices  advising them of the polling place to which they were assigned. Consequently some poll workers went to the wrong location on election day further compounding the problems.
We will investigate the feasibility and costs of expanding election day phone banks, including the possible use of the County's Emergency Operations Center or contracting with the City of Los Angeles Elections Department. The goal will be enhancement of our capability to handle and prioritize the ever increasing volume of calls and to better distinguish between calls from voters who may be confused and those placed by poll workers.
Unfortunately, it is now clear that the impact of redistricting on polling place changes will continue through the November 2002 Election. This is due to several factors including an expected voter turnout in the County's November 2002 Election of 1 million additional voters who did not participate in the March Primary election . Additionally, the redistricting process has not yet been completed as cities and school districts will not finalize their new boundary lines until July 2002 or later . This will result in the necessity to adjust numerous precinct boundaries again to coincide with these districts' new lines. Therefore, many of November's voters will be confronting polling place changes as a result of redistricting for the first time when they go to vote on November 5, 2002. To prepare for the expected onslaught of inquiries from confused voters our mechanism for determining the validity of voters' calls to report their "polling place not open" is being re-tooled for the November 5, 2002 General Election.
Solutions to Polling Place Problems
Earlier Recruitment of Polling Locations
The RR/CC is committed to earlier recruitment of polling places. We have already begun contacting facilities to secure commitments for the November 2002 General Election. We will obtain overlapping commitments for use of facilities for voting and, where necessary, will cancel locations if final redistricting of cities and school districts' boundaries shows them to be redundant.
More Prominent "Voter Alert" of Possible Polling Place Changes
It has consistently been our policy to display on the sample ballot booklet front cover the message: "Voter Alert! Your polling place may have changed. Poll location is shown on back cover."  For the March 2002 Election, our concentration on disseminating information on the first ever "Modified Closed" Primary rules prompted placement of a "New Election Law" alert near the top of the sample ballot cover. This relegated the size and location of the standard polling place alert to a less prominent location on the front cover of the sample ballot booklets for the March 5, 2002 Election. For the November 2002 and subsequent elections we will again prominently display the polling place "Voter Alert" message in large, contrasting typeface on the cover of all sample ballots.
In our continual partnership with City Clerks, we have recently advised them of plans to obtain their early review, prior to each election, of our tentatively selected polling places to get their input regarding suitability of the sites in their Cities. We also will seek to expand City Clerk assistance in locating and obtaining permission for use of the best facilities in their Cities as voting locations.
Polling Place Rental Increase
For many years the County has paid $25 compensation for rental of polling places. Cities throughout the County increased this fee to $50 per site several years ago. If approved, such an increase would cost an additional $125,000 per countywide election for the 5,000 precincts.
Problem: Poll Workers Cancellations, No-Shows and Late Opening Polls
No other issue has dominated the long-term planning of election administrators nationwide over the past few years than the question, "Who is going to staff the polls?" The pool of poll workers was once fairly stable, consisting primarily of retired persons and housewives. However, the number of veteran, trained volunteers has diminished alarmingly as the traditional group ages (many are in their 80s). Statistics nationwide show volunteerism has dropped off in the past decade. Additionally, many women, once the backbone of the poll worker corps, have joined the workforce in record numbers. Long-time poll workers complain that working a 14+ hour day  for a very small stipend, and the increasing complexity of election laws and procedures, reflected in the weight of the precinct election supplies, has led to their decision to forgo service in future elections. Indeed, sustaining a viable poll worker force was a key concern identified in the management audit of the Department conducted by Strategica in 2000. The audit stated, "Over the next few years, the RR/CC is likely to reach a point where the margin for error in having enough poll workers to conduct a major election disappears." 
In a County with 4,845 polling locations that must each be staffed with a supervising Precinct Inspectors and 3 or 4 poll worker Clerks  (total of approximately 22,000 poll workers), a few late-opening polls inevitably occur in every election. We know that poll workers sometimes have personal emergencies at the last minute that prevent them from serving or arriving exactly on time. For every election the RR/CC routinely anticipates a given percentage of poll worker cancellations and election day no-shows and plans accordingly.
To prepare for the anticipated number of last minute emergencies we always hire approximately 200 "precinct coordinators" who are specifically trained to be available to work as Inspectors or election day roving troubleshooters. These temporary workers (coordinators) are paid $350 to be available and on-call in the days preceding the election for deployment at the 11th hour anywhere in the County on election day. An additional pool of up to 100 temporary workers are hired and paid $200 as "precinct reservist clerks" to appear at the RR/CC office early on election morning to serve as stand-by poll worker clerks to augment locations where poll worker no-show rates are high.
Surprising Number of Inspector Cancellations & No-Shows/Actions Taken
The number of last minute Precinct Inspector cancellations is typically no more than 50-70 in the five days prior to a statewide election. However, as is shown on the pie chart found at Attachment 4 the number of Inspectors who quit on election day or in the five days prior to the March 5 Primary Election totaled an unfathomable 231. Indeed, as the chart shows, an unprecedented number of Inspectors - 797 - cancelled from the time of initial agreement to serve through election day. The chart lists the reasons the Inspectors gave for canceling when post-election.
Consequently, we were forced to assign our entire pool of 200 precinct coordinators by the weekend before election day, a situation never before faced. We had limited success on Monday, March 4th tracking down the 100 temporary staff hired as election day reservist clerks and pleading with them to take on the rank of Precinct Inspector for emergency precincts. Aware that this left us vulnerable to further election eve cancellations and election day "no-shows," early on March 4th we hastily trained approximately 90 of our RR/CC staff and pressed them into service as emergency Inspectors at locations far and wide throughout the County.
Attempts to elevate previously recruited precinct poll worker Clerks to the rank of Inspector in the crisis precincts were largely unsuccessful. We were repeatedly told by poll worker Clerks, who are paid a stipend of $55 compared with $75 for Inspectors, that they are unwilling to take on the added responsibilities of being in charge on election day and also entails the difficulty of hauling heavy election supply boxes  to the polls and later transporting the same supplies to the collection centers after voting is completed.
Through a combination of all of these emergency last-minute efforts, by shortly after midnight on election eve all 4,845 precinct election supply boxes had been placed.
Nevertheless, the number of election day Inspector problems reached an unprecedented level resulting in late openings of 121 polling locations. Several charts at Attachment 4 depict the reasons these Inspectors gave for being late or not arriving at all. The profile that has emerged reveals that almost half of these Inspectors (52 of 121) were first-time, new Inspectors and 75% were in their 20s, 30s and 40s and 50s. This is worrisome for the future because long-time, primarily elderly Inspectors will inevitably cease to be available forcing greater reliance on younger less experienced poll workers.
Due to the size of L.A. County elections and inevitable emergencies with some poll workers, a small number of late opening polls occur in every election. Reviewing historical records, the RR/CC achieved the best on-time polling place opening statistics during the Presidential General Election in November 2000 when 27 polls opened late, less than one-half of one percent of the total 4,986 voting precincts. Most of these 27 were opened by 8 a.m. and the last poll opened by 10 a.m. In order to meet or hopefully exceed this standard for future elections, we have analyzed the specific factors that were present for the November 2000 election that were not there for the March 2002 Primary Election.
The factors present in November 2000 that were absent in March 2002 included:
High Voter and Poll Worker Interest
Presidential General Elections always result in the highest voter turnout of any type of election. High voter interest translates to high poll worker interest. Attendance at poll worker pre-election training has always been higher for general elections in presidential years; poll workers are more willing to serve when they feel they are a part of an "important" election.
No Extraordinary Time Pressures
The November 2000 Presidential Election did not follow a redistricting period, our usual polling places did not have to be disrupted to re-draw precinct boundary lines, and the election was at a "normal" time on the calendar, not months earlier, as was the March Primary Election.
Extra Contact With Poll Workers
The first use of touch screen voting equipment for "early voting" in conjunction with the November 2000 Presidential General Election required telephoning every Precinct Inspector over the three days preceding the election to inform them who in their precinct had voted early on a touch screen device (as it was too late in the process to mail out a list of these voters). Similar to the lists of absentee/mail voters, which are mailed to Inspectors four days prior to every election, Inspectors also needed to mark the names of touch screen voters in their precincts' rosters to prevent the possibility of those voters casting ballots again on election day).
This extra pre-election contact provided an opportunity for Inspectors to alert us if they wanted/needed to cancel prior to election day, allowing time to replace them with a precinct coordinator. As plans call for touch screen voting to again be offered in conjunction with the early voting period prior to the November 2002 General Election , this extra contact with Precinct Inspectors will again be done for that election.
Solutions to Poll Worker Recruitment Problems
Poll Worker Team Re-Building
The major changes required in precinct boundary reconfiguration as a result of legislative redistricting prompted RR/CC staff to use this opportunity to comply more diligently with State law that directs assignment of poll workers to their "home" precinct whenever possible. Unfortunately, the unintended consequence of this effort was a significant group of poll workers who, upset at the break-up of traditional poll worker teams, became disgruntled, and some canceled their service as a result. We plan to call each Precinct Inspector this Spring to get their input on re-forming poll worker teams in preparation for the November General Election.
Precinct Supply Box Replacement
As demonstrated for your Board during my oral presentation on March 12, the existing precinct election supply boxes date from the early 1970s and both the boxes, and the poll workers, are collapsing from the weight. When loaded with required materials, these supply boxes weigh a burdensome 54 pounds. This does not include the 20-pound weight of the separate suitcase that houses Votomatic devices/machines. The large supply boxes hold official ballots, sample ballot booklets and state pamphlets (translated into seven foreign languages in compliance with Federal and State laws), the roster of voters for the precinct, instructional materials and signs. The weight of these boxes must be borne by each Inspector to his or her polling place on election day.
Many persons who are otherwise well-qualified to perform the duties of precinct Inspector are physically unable to lift these heavy boxes without injuring themselves. This is particularly true when poll workers are assigned to a location that does not have immediately adjacent parking. To assist in recruitment and retention of Inspectors, we request funds to purchase 5,000 new precinct supply boxes designed with built-in wheels and a handle similar to a large molded plastic suitcase.
Our neighboring counties of Riverside and San Bernardino have purchased such rolling supply boxes and report a correlation between recruitment and retention of Inspectors as they are pleased with these more maneuverable boxes. Although we do not yet have firm estimates from manufacturers, based on the price of similar equipment in other counties, we believe the one-time cost for new supply boxes will be approximately $600,000. We need to order these new boxes this Summer, in time for use in the November 2002 General Election.
Expansion of County Pollworker Program
The Voluntary County Poll worker Program began in November 1998 and has become so successful that it has since been emulated by other election offices in California and nationwide as a model program. As shown on the chart at Attachment 15, the RR/CC is making progress toward achieving the goal of eventually placing a County poll worker in each of the 5,000 polling sites for every countywide election. The number of County employees participating grew from 360 in November 1998 to 1,160 in November 2000 to 2,120 in March 2002 elections.
The County Poll Worker Program is vital to the immediate as well as long-term success of polling place management. County poll workers have proven to be an exceptionally reliable source of polling place staffing, with virtually no cancellations, an excellent record of attending poll training classes and appearing on time on election day. An added bonus is many are bilingual, and are able to offer translation assistance to English as a Second Language (ESL) voters.
We are encouraged by the number of County employee poll workers who, in post-election telephone surveys conducted over the past month, have expressed a willingness to serve in future elections. Such continuity will be essential to the sustaining and expanding the program. County poll workers are the backbone of our ongoing preparation for the time when Los Angeles County completes the process of phasing-in modern electronic voting equipment. County employee poll workers are computer literate, a factor in that will prove invaluable when a new, sophisticated, computerized voting system is introduced at the precinct level.
In order to maintain the momentum of this excellent program, we need additional staffing to coordinate County poll worker activities on a year-round basis. Current staffing for the County poll worker program consists of two clerks in the RR/CC Polls Section (and consequently required significant diversion of Polls Section management staff at the expense of their responsibilities). The two clerks were overwhelmed in the March 2002 Primary Election due to the dramatic increase in the number of County employees volunteering and the paperwork required to secure permission of their supervisors, track their applications, report participation rates back to each County department, maintain a County Poll Worker website, and finally place County employees in poll locations as convenient to their homes as possible. Additional staff is needed to support a year-round recruitment program .
Expansion of Student Poll Worker Program
California legislation enacted in 1997 allows high school seniors who are U.S. citizens with at least a 2.5 grade point average to work at the polls. By participating as poll workers, students gain hands-on experience in the election process while also serving their community. Many of these students are bilingual, providing further assistance to voters in need of translation services.
The growth of this program has been stymied due to lack of appropriate year-round staff to dedicate to sustain and develop it. Beginning with the 1998 Primary Election, 255 high school students were recruited. This number grew to a high of 969 students in the 2000 Primary Election but diminished to 596 in the March 2002 election as is shown at Attachment 16. Currently one temporary clerk and three temporary college interns struggle to recruit and place high school students during major election periods.
The Student Poll Worker Program will not be sustainable unless permanent higher level staff are dedicated to maintaining and developing it year-round in order to be available to meet with teachers to market the program at hundreds of high schools and devise appropriate support and appreciation events for the program. Teachers who participated in the program in one election expect to be contacted for the next election. RR/CC current staffing with temporary help only at election periods has not been able to provide the necessary continuity to administer the program appropriately and reach full potential for expansion of this program.
The Strategica Management Audit of the Department released in 2000 specifically cited the recruitment of student poll workers as a significant achievement, yet criticized the Department for not running a larger program with administrative employees and more clerical support staff to provide the outreach needed, necessary analytical skills, and tracking and record-keeping.
It has become increasingly challenging for even the most dedicated and intelligent of poll workers to perform well in the current environment of constantly changing, complicated election laws and procedures. Over the past few years, new laws such as the Open Primary, the Modified Closed Primary, Failsafe Voting, Provisional Voting, and the implications of a 15-day voter registration deadline are but a few of the examples of the exponential increase in complexity and expectation for poll workers to comprehend.
Given the extremely difficult job of finding 22,000+ poll worker volunteers willing and able to take on these challenges, the Department established a pilot Poll Worker Academy Program in August 2000 to begin the process of enhancing the skills and knowledge of poll workers. Approximately 200 veteran poll workers participated in the maiden program (representing less than 1% of all poll workers). The objectives of the program include: 1) upgrading the skills and knowledge of poll workers through intensive, hands-on training over the course of three days, and 2) weeding out poor performers and to minimize election day mistakes. Our ambitious goal is 2,000 Poll Worker Academy graduates per year.
It is important to note the distinction between the Poll Worker Academy concept and the routine 300+ pre-election training classes that are always offered by the Department throughout the County prior to every election. The curriculum of the Academy was designed to provide in-depth, hands-on understanding of the reasons for each election procedure, in the course of an intensive three-day program rather than the two-hour pre-election classes that focus only on the highlights of how to deal with key situations.
The Poll Worker Academy pilot program was organized using five temporary clerks borrowed from the RR/CC Polls Section. Again, the key resource needed to continue this program is several dedicated full-time staff at appropriate levels to manage and expand this program similar to the needs for expanding the County employee poll worker, student poll worker, and Adopt-A-Poll private business poll worker programs.
Expansion of Pilot Program - Adopt-A-Poll
At the March 12th meeting of your Board, a suggestion was made to form an Adopt-a-Poll Program. In June 2001, our Department developed a pilot Adopt-a-Poll Program targeting chronic hard to recruit areas of the County. Following identification of possible partner organizations, during the summer months of 2001 staff made 33 presentations to a wide variety of community groups throughout the County (brochure with program highlights is found at Attachment 17. The program's goal is to obtain both a set of poll workers from a church, club or other community organization and a polling location (where needed) to take on the responsibility as a fundraising opportunity for their organization.
Unfortunately, none of the initially contacted organizations followed-through and the pilot program was not successful. It has been unsustainable without full time staff to manage and promote the program. After several years of working on a similar program in Ventura County, the Registrar achieved moderate success this year but significant staff resources have been dedicated to that County's program.
Other Avenues of Poll Worker Recruitment
Your Board also suggested at the March 12th meeting that the American Association of Retired Persons (AARP) be contacted along with other senior citizen organizations. In the past we have contacted AARP as well as Rtired and Senior Volunteer Program (RSVP) and the County Employees Retirement Association. We have placed articles seeking poll workers in these organizations newsletters without success to date.
We have forged strong partnerships with community organizations through our Department's Community Voter Outreach Committee (CVOC). CVOC meetings, held quarterly, are well attended by representatives of the League of Women Voters, the National Association of Latino Elected Officials, Chinese Americans United for Self Empowerment, the Asian Pacific American Legal Center and several other Asian organizations, the Western Law Center for Disability Rights and representatives of the political parties.
Members of CVOC and participants of these organizations have responded enthusiastically to the hundreds of requests our Department receives from a variety of schools and community events to provide voter registration, voter education, and hand-on practice with voting devices. We provide each of these organizations with materials to conduct voter outreach and also to recruit poll workers. With the exception of the League of Women Voters, we have had minimal success with persuading these organizations' representatives to assist with poll worker recruitment as they agree such recruitment of volunteers is very difficult.
Expansion of Pilot Program of Inspector Supply Pick-up
For the November 2001 Uniform District Elections (UDEL) the RR/CC established a pilot program for Precinct Inspectors to pick-up their election supplies on specific dates at selected locations. UDEL was the perfect election to begin such a new program as the number of voting precincts in off-year elections is approximately 1,000 compared with 5,000 for countywide elections. Inspectors responded well to the program - fully 93% picked up their supplies leaving only 7% to be delivered. Inspectors told us they enjoyed controlling the time of day when they picked up their election supplies compared with waiting for hours on supply delivery day.
The benefit to the RR/CC was realizing that Inspectors who showed-up to collect their supplies were demonstrating a level of responsibility that provided us assurance that these individuals would reliably show up on election day. The 7% of Inspectors who did not pick up supplies provided a "warning signal" that additional contact was needed. We were able in many cases to avoid a "no-show" situation on election day in November 2001 due to the warning of an Inspector who did not pick up supplies.
There was insufficient time to expand this pilot program countywide between the November 2001 and March 2002 elections. We now plan to expand the supply pick-up program to the extent possible for the November 2002, perhaps concentrating the program in those Cities where poll workers are familiar with this approach as it is the norm in several smaller Cities when they conduct elections.
Increasing Pay for Poll Workers
Poll workers have traditionally been considered community volunteers. Although they receive a small stipend for long hours of service, the rate of poll worker pay is significantly less than minimum wage . The responsibilities of the Inspector in charge at each voting precinct are significant and include:
The Strategica Management Audit stated, "In the future, securing enough poll workers may have to rely less on the spirit of civic duty and more on economic incentives. One strategy for ensuring adequate staffing would be to increase stipends to $200 to $300 a day, or whatever it takes to attract enough people. Including training and recruitment costs, the overall poll worker payroll budget for a major election (now at approximately $1.5 million) may increase to $7 million per election."
Given the County's budget situation, the RR/CC recognizes that such a dramatic increase is not feasible. Additionally, an across-the-board increase of the stipend for all poll workers, or even for Inspectors, may not be the best long-term solution to the poll worker recruitment crisis. Some targeted methods of compensation may be more effective.
For example, in September 1999 your Board approved the RR/CC's request to increase the stipend for attending training class for Inspectors from $5 to $25 and also extended this $25 training "bonus" to all poll worker clerks. Clerks had previously received no compensation for attendance at a poll worker training class. Since implementation this new policy has increased costs approximately $200,000 for each countywide election. However, the goal of achieving a more highly trained poll worker force, especially in the environment of so many new election laws, has been met. Statistics reveal that the added financial incentive for clerks to attend training classes has resulted in triple the number of trained clerks while the number of Inspectors, many of whom attended in the past, has also increased.
Continuing to offer Inspectors the base stipend of $75 but adding a recruitment bonus of $25 for each clerk they are able to recruit in geographic areas of chronic recruitment difficulty and to supplement where poll worker clerk cancellations occur may provide appropriate incentive for Inspectors to assist in recruitment of clerks. We estimate that such a recruitment bonus would be paid to approximately half of the Inspectors. Assuming approximately half of our precinct Inspectors (2,500) assisted in recruiting two clerks each, the added cost of this proposed program would be $125,000 per election.
Similarly, we believe emulating San Diego County's model  of creating a position of Assistant Inspector, responsible for stepping up to take charge of the precinct in the event of the Inspector's cancellation, would be a valuable program to prevent late opening polls. San Diego County pays each Asst. Inspector $20. The cost to implement such a program in our County would be $100,000 per election (5,000 Asst. Inspectors at $20 each).
We believe that the costs of offering Inspectors a recruitment bonus and creating the position of Asst. Inspector would be partially offset by a reduction in RR/CC temporary Polls Section staffing needs. If financial resources are provided, the RR/CC would like to pilot these bonus programs in conjunction with the November 2002 Election.
Financial Resources Needed for Improvement
Board motions passed at your March 12th meeting requested identification of the additional resources needed to make improvements and to enhance the quality and quantity of poll workers. Throughout this report a range of financial resources have been described and requested including augmented permanent staffing to sustain and expand the wide range of poll worker recruitment programs identified herein, to replace 30-year old election supply boxes for 5,000 voting precincts and to offer a small increase in precinct Inspector pay via a bonus program for those Inspectors who recruit poll worker clerks where vacancies exist in chronically hard to recruit areas and to create and fund the position of Assistant Inspector.
However, augmenting RR/CC financial resources as described above can only be effective in conjunction with retaining the Department's current net County cost (NCC). The CAO's current recommendation for the RR/CC's FY 2002-2003 budget reflects a reduction of $2.5 million in NCC that includes $2.2 million in election services funding reduction and a third of a million dollars in other miscellaneous reductions. Additionally, we presented information to the CAO at the Department's March 1, 2002 budget hearing that clearly showed the financial impact of redistricting, combined with the Modified Closed Primary format, caused sample ballot printing costs for the March 2002 Primary Election to double from a budgeted $1.8 million to approximately $3.7 million.
The current CAO recommended budget for the RR/CC does not contain any augmentation for sample ballot printing costs which are also anticipated to increase by at least 40% for the November 2002 General Election.  Consequently, our Department's budget is slated to begin FY 2002-2003 approximately $4.4 million in the hole . It will do little good to augment the RR/CC budget by the $1,712,000 identified on the chart above for needed improvements if the proposed FY 2002-2003 budget cuts are approved by your Board. For perspective, the additional $1,712,000 (of which $1,112,000 would be on-going NCC) represents an increase of only 5% of the average $20 million cost of election supplies and services per each countywide election.
The Registrar-Recorder/County Clerk Department experienced a number of problems conducting the March 5, 2002 Primary Election. The most significant involved recruitment and retention of polling locations and poll workers principally due to an unprecedented number of last-minute poll worker cancellations that resulted in 121 of the County's 4,845 poll sites (2.5%) opening after 7 a.m. The problems experienced in administering this election were compounded by the unusual circumstances surrounding the collision of the first March Gubernatorial Primary Election, held three months earlier than ever before, with the effects of major political jurisdictional boundary lines changes from the recently completed decennial redistricting process. Because legislative redistricting created more changes than ever before, there was a much greater impact on reconfiguration of voting precinct boundaries which, by law, cannot bisect any other political jurisdictional lines. Consequently, approximately 20% of the County's voters were affected by re-drawn precinct boundaries and were assigned to unfamiliar polls. The redistricting process is still on-going for cities and school districts and will impact the conduct of the November 2002 election. 
The volume of the problems
encountered on March 5 came uncomfortably close to fulfilling the alarm
sounded in the comprehensive management audit of the RR/CC
This audit recommended the infusion of additional financial resources, including augmenting staff with administrative and mid-level management positions, to prevent a crisis from occurring. However, as was described more fully earlier in this report (Attachment 5), audit-driven personnel positions requested in the Department's budget submission for FY 2001/2002 were denied and the CAO's current recommendation for FY 2002-2003 includes a budget reduction that accumulates to $2.5 million . As requested, this report identifies (on page 21) an additional request totaling $1,712,000 for resources needed to improve the conduct of elections and avoid recurrence of problems experienced in the March 5, 2002 Primary Election.
The problematic conditions were compounded by introduction of the most confounding primary election rules ever devised - the "Modified Closed" Primary. This new format is sometimes referred to as the "Slightly Ajar" Primary because it is a confusing hybrid of the former Closed and recently "Open/Blanket" primary rules. 
Post-election research has revealed that registrars of voters statewide struggled with similar problems including an unprecedented number of poll worker cancellations for the March 2002 Primary Election . The pervasive nature of these experiences forebodes a repeat in the next primary election, the March 2004 Presidential Primary. Assuming retention of the "Modified Closed" Primary format and the expected continuation of the Democratic and Republican parties decision to allow non-partisan voters to cross-over to their elections, the problems with this format will be much more severe as non-partisan voters will be prohibited from voting on the presidential contest. That key restriction will undoubtedly confuse and possibly enrage non-partisan voters who can be anticipated to direct their anger at poll workers. Such a situation is bound to cause poll worker recruitment problems as well as defections on election day.
Additionally, the acceleration of major structural changes to electoral laws is continuing at a dizzying pace. A well-funded initiative is currently circulating in California to place "same day" voter registration on the November 2002 ballot. The recent change from 29 days to the 15-day voter registration deadline caused problems in electoral administration  and permitting voter registration on election day would undoubtedly complicate matters. Also, the U.S. Senate is poised to pass legislation  that would create a two-tiered system of voter identification with first time voters in Federal elections falling under a requirement to show identification at the polls (or when submitting an absentee ballot) while previously-registered voters would not incur an identification requirement.
At a two-day hearing of the State Assembly's Elections and Reapportionment Committee held in January 2001 (immediately following the November 2000 Presidential Election with the stated goal of taking testimony from experts on how to avoid Florida-like electoral administration problems), several registrars of voters testified that electoral processes must be simplified to prevent an increase in the number of errors on the part of voters and election administrators alike. Unfortunately, recently enacted laws such as the 15-day voter registration deadline, advancing the Gubernatorial primary election to March instead of June combined with the format change to "Modified Closed", and extending permanent absentee voter status to all voters, to name a few, have vastly complicated the process of administering elections.
Into this environment of unrelenting change looms the requirement for California counties that currently use punch card voting systems (including Los Angeles County) to convert to a different voting system no later than the March 2004 Presidential Primary election .
The election process is fragile. With the uncontrollable external forces mentioned above causing continual changes to the election environment, the RR/CC must have sufficient staff and other financial resources identified herein to create the necessary internal controls and structure to conduct elections appropriately in order to maintain public confidence in the democratic process.
** Attachments may require Adobe Acrobat Reader to view .pdf files. Click here to download Adobe Acrobat Reader for free.
1. Board motions are found at Attachment 1 (.pdf file).
2. See Attachment 2 for February 7th press release entitled New Laws Will Significantly Impact March 5 Primary Election (Press Briefing Scheduled) in which the RR/CC was quoted as saying "I am very concerned that confusing new state election laws will result in unintended consequences, including much greater opportunity for misunderstandings and errors on the part of voters and precinct poll workers alike."
3. Under the previous Open format, any voter, regardless of party affiliation, could vote for candidates of any political party. When the U.S. Supreme Court declared California's Open primary unconstitutional in July 2000, the state legislature wanted to preserve the opportunity for nonpartisan registered voters to participate in primary elections if the political parties chose to allow it. While this demonstrated legislative good intentions, the new "Modified Closed" Primary format resulted in confusing inconsistency as four of the seven political parties (Democratic, Republican, American Independent and Natural Law parties) chose to allow non-affiliated voters to vote for their partisan candidates while the other three (Libertarian, Green and Reform) did not. Furthermore the Democratic and Republican parties decided to preclude nonpartisan voters from voting on party Central Committee contests, while the American Independent and Natural Law parties permitted this. Because this format is neither Open nor completely Closed, Registrars dubbed it the "Slightly Ajar" Primary.
4. The registration deadline for the March 5th election was actually 14 days prior to election day due to the President's Day holiday. Additionally, voter registration forms mailed by voters on the 14th day often do not arrive at the RR/CC office until the 12th or 11th day preceding the election but are still legally valid and must be entered immediately into the voter registration database. Consequently, many voters have to be added to supplemental voter lists that are mailed to poll workers a few days before the election. For the March 5th Primary Election this process was even more complex because voters who executed registration forms between 19 and 14 days prior to the election in order to change their party affiliation nevertheless appeared on precinct voter rosters as being registered with their former political party because their changes in party affiliation appeared only on the mailed supplemental lists that poll workers often forget to reference. The listed party affiliation on the precinct rosters is very important in primary elections because different types of ballots must be issued for each political party whereas in November general elections all voters in a given precinct receive the same ballot.5. See pages 11-13.
6. As shown in Attachment 4, a record-high 797 Inspectors canceled service including 231 who canceled within five days of the election. Prior countywide elections have experienced an average of 50-75 Precinct Inspector cancellations during the 5 days before the election. The 121 precincts that opened after 7 a.m., constitute 2.5% of the County's 4,845 voting precincts. Of the 121 late opening polls, 51 opened by 8 a.m., 52 between 8 a.m. and noon, 17 opened after noon and one did not open but voters were directed to vote at another polling location down the street.
8. Following the 1990 census, the State Senate and Assembly districts were configured using a "nesting" system in which two Assembly districts were contained completely within each State Senate district. This configuration greatly minimized the need to alter precinct boundary lines.
See Attachment 6 (.pdf file),
11/26/01 memo "Re-Districting Voting Precincts - Process and Timelines"
10. See Attachment 7 (.pdf file).
11. See Attachment 8 for visual depiction of the soaring number of different ballot combinations and sample ballot booklet printing cost history.
See Attachment 9.
13. Attachment 8 visually depicts the unprecedented explosion of 3,184 different types of ballot combinations required for March 5, 2002 Election (this number had been 198 for the 2000 Open Primary Election).
Shown at Attachment 11.
15. See Attachment 12 (.pdf file) for a Sample of a Poll Worker Appointment Letter.
16. Voter turnout in the County for the March 5, 2002 was 26%. Past trends indicate that turnout for November Gubernatorial General Elections is much greater than for primary elections. Voter turnout is anticipated to double to approximately 50% for the November 2002 election.
17. See Attachment 13 (.pdf file) for a chart on local districts re-districting schedules and also news articles relating to City of Los Angeles and LAUSD redistricting plans.
19. A split shift option has been available for poll worker clerks for several years (not for the Inspectors in charge). This option is infrequently used because it requires the two poll workers sharing the shift to split the stipend among themselves (based on the number of hours each worked) and also both must sign an agreement acknowledging that if the worker assigned to the afternoon shift does not show up, the lead worker who arrived in the morning must remain throughout the day.
The audit suggested a number of dramatic potential solutions including:
1) increasing the poll worker stipend from the current $55-75 to between
$200-300, 2) implementing regional "voting center" (not currently
permitted under State law that limits maximum precinct size to 1,250 registered
voters, and 3)
"drafting County employees to reduce poll site
21. State law requires a minimum of three poll workers at precincts with under 1,000 registered voters and four workers in precincts between 1,000 and 1,250 registered voters (1,250 is the legal maximum in Counties with a population of over 1 million).
22. Supply boxes containing all election supplies (ballots, roster of voters, sample ballot booklets and state pamphlets in 7 required languages, etc.) weigh 54 pounds and must be taken, along with the 20-pound case containing Votomatic devices/machines to and from the polling location.
24. The financial resources needed for this program are identified on page 21.
27. Sample ballot printing costs are directly tied to the number of unique ballot combinations (candidates and issues) appearing on the ballot. For November General Elections, the number of ballot combinations has never exceeded 300 in the past (the numbers are higher for primary elections due to the separate combinations for each political party). For the November 2002 General Election the number of unique ballot combinations has been determined to be 503.
29. Fully explained at the top of page 10.
31. Fully explained on pages 5-6.
32. See Attachment 4.
33. As described on page 2.
34. A vote is expected this month of S565. The U.S. House of Representatives passed their version of election administration reform in December 2001. A conference committee would be convened to work out differences should S565 pass, as expected.
** Attachments may require Adobe Acrobat Reader to view .pdf files. Click here to download Adobe Acrobat Reader for free.
| Back to Top |