March, 2009

Introduction

This report outlines the findings of the Election Reform Network with regard to the operation of the November 08 election in Montgomery County.  The vantage point of this report is that of the individual eligible voter or potential voter who interacts with the “apparatus” of election administration with the goal of casting a ballot that is recorded and counted accurately

This report is part of the continuing effort of the Election Reform Network to promote greater accountability between the operation and administration of elections and the citizens they serve by identifying problems and needed improvements in our system of elections.  The term “system” is emphasized since elections are a complex combination of interacting components ranging from voter registration and voter education to the employment of appropriate voting technology and the training of poll workers, with a wide range of additional functions and tasks in between.  Together, they comprise the overall system of elections and voting that is so crucial to governance in a democracy.  

We recognize that the administration of elections is an area of government that has gained significant prominence largely since the 2000 elections and that elections officials at the local, state and federal levels are striving to respond to the complex challenges and higher public expectations which they now face.  In Montgomery County, we believe that the situation parallels these broader trends – with dedicated officials and staff working with no significant increase in resources available to meet the stiff elections challenges of the 21st century.  

This report, then, does not set out to undermine the work of the Montgomery County Election Board or the Department of Voter Services.  Nor is it an attempt to evaluate in a comprehensive way the November election in the county from an operational standpoint.  It does not, for example, speak to the strengths of the process, including the hard work and dedication of many staff at the Department of Voter Services and the some 2000 citizen-poll workers across the county, without whom no election could occur.  Rather, the report identifies specific incidents and problem areas and speaks directly to some longstanding systemic issues related to the elections process, which are generally not unique to Montgomery County.  In fact, the problems we cite reflect those that are, for the most part, prevalent both across the Commonwealth of Pennsylvania and the nation.   

Nevertheless, because the operation and administration of elections is lodged principally with county government in Pennsylvania, this report is directed primarily to the Election Board.  It identifies specific problem areas based on our on-site documentation and reporting. We then hope to use this report to continue our dialogue with the County in the interest of improving our elections and strengthening the democratic process, which is, of course, the lifeblood of our system of government.
The body of the report is a listing of Election Day incidents documented by Election Reform Network poll watchers in more than 100 target precincts.  This is followed by a brief discussion of the significance of our findings along with recommendations and some outstanding questions.

Election Day Incidents

Election Reform Network (ERN) poll watchers monitored more than 100 target precincts and identified systemic problems in the following key areas: provisional ballots; emergency paper ballots; voter registration ; vote machines; poll worker issues; wait-times; intimidation of poll watchers; inadequate access to polling places for the disabled; and inadequate signage.  

It is important to note that due to our limited resources as a citizens organization, our coverage of slightly less than one-quarter of all precincts raises the question of the extent of these problems countywide – and one that can, of course, be debated.  We recognize that this report is more of a broad snapshot of issues and problems than a scientific survey yielding comprehensive results.  Two areas, for example, access for the disabled and inadequate signage, have only one documented incident each, yet we believe them to be areas of serious concern worthy of action.  In one case, we ask below (see Information Requests) whether the county has data on accessibility standards for individual precincts.  In the other, it is the sense of ERN poll watchers and monitors from their travels across the county during the last few elections that finding individual polling places is often a challenge for voters and more signs - strategically placed - would make a major difference.   

Provisional Ballots

Lansdale 1-2, Lansdale Middle School:  
Three prospective voters not found in registration data base during the morning were not offered provisional ballots.  The Judge of Elections (JoE) told an ERN poll watcher that he was told by the Board of Elections (BoE) that because the three names were not in the system, they should not be given provisional ballots.  In the afternoon, the JoE told the same ERN watcher that he had given out “about four” provisional ballots, explaining that the BoE now says to issue provisionals if voters can not be found in the system.

Two college students were not found in the system and not offered provisional ballots.

Lower Providence 2-4, Eagleville School:
A voter was rejected because the name on his drivers license differed from the one in the poll book.  When a poll watcher asked the JoE why the man was not offered a provisional ballot, the constable got involved, accusing him of being disruptive.  The watcher then called the BoE and was told that the JoE and constable were correct – that provisional ballots “wouldn’t be counted anyway” and that they “just made more work” for staff.

Lower Providence 1-1, IBEW Local 380:
A voter (name available) from Collegeville claimed to have registered at Montgomery County Community College, but was not in the poll book.  The JoE called the BoE and was told that he should not issue a provisional ballot because “it would just be thrown out anyway,” since the voter was not in the system.  After discussing the matter with an ERN watcher, the JoE decided to issue a provisional ballot.

Norristown 1-1, Warehouse:
A voter not found in the poll book was found in the data base as having attempted to register two weeks after the deadline; he was not offered a provisional ballot.

Lower Merion 12-3, St. Joseph’s University:
A number of students were not found in the poll book and not offered provisional ballots.

E. Norriton 2-1, Norriton Fire Engine Co.:
Two voters were denied provisional ballots after the JoE was instructed by the BoE not to issue provisional ballots when names were not found in the system.

Abington 13-1, St. Paul’s Lutheran Church:
Student who had been studying abroad claimed to be registered for the Primary, but was not found in the poll book and was not issued a provisional ballot.

W. Norriton 4-2, Marshall St. Elementary School:
Young voter not found in the poll book claimed to have registered through the Pennsylvania Department of Transportation (Penn DoT).  BoE told JoE not to issue a provisional ballot “because it would be refused.”

Voter not found in the system but claims to have voted in 2004.  She was not allowed a provisional ballot.  (Have name and address.)

Voter not in poll book but has voter card.  Moved in July, 08 and claims to have reregistered through Penn DoT.  Voter was told computer says “cancelled.”  Not offered provisional ballot.  

Emergency Paper Ballots (EPBs)

Upper Dublin 1-1, Township Building:
Two machines were not operating at 7 am.  The JoE was advised by party committee people not to issue paper ballots.  BoE approved issuing EPBs at about 10 am.  Repairman arrived at 11 am but was unable to get them working.  At one point, the JoE told an ERN poll watcher that the BoE told him that there was no such thing as emergency paper ballots.

Machine Breakdowns, Operating Problems and “Fleeing Voters”

Upper Dublin 1-1, Township Building:
Neither of the two machines were operating at 7 am.  Repairman arrived at 11 am but was unable to get them working.  EPBs were eventually issued.

Lower Merion 10-2, Bryn Mawr Presbyterian Church:
While at the poll, a legislative candidate found that his name did not light up when “Straight Party” for his party is activated.  JoE initially refused to take the machine out of service and have it replaced.  The machine was eventually removed after an intervention by the Board of Elections.

Abington 11-3, Abington High School:
860 voters signed in by about 7:40 pm, while total machine count had 855 votes at that time.  

Norristown 4-3, Hancock School:
Machine fails to operate.  Replacement brought in at 11 am.

Red Hill, Borough Hall:
There was a disparity of 18 votes between the machine count and the numbers in the signature book by about noon.  The JoE was concerned but had no explanation.

Cheltenham 3-4, La Mott Com. Ctr.:
“Do not wish to vote” button stayed lit when voter pressed “cast ballot” button.

Cheltenham 5-3, Myers School/Modular:
“Do not wish to vote” button stayed lit when voter pressed “cast ballot” button.

Whitemarsh East 2, St. Paul’s UCC:
Unconfirmed report that the first 15-25 votes on a machine went unrecorded.
Poll Worker Issues - Understaffing

Lansdale 2-3, Knapp Elementary:
Understaffed; only one person at the table with 2 poll books.

Norristown 1-2 and 1-3, Stewart Middle School:
Both precincts understaffed.  JoE from 1-2 reassigned machine operator to sign-in table, leaving an 81-year-old individual running three machines.  A potential poll worker arrived at 8:30 am and requested a position.  When the JoE refused her, she went across the room to 1-3, where the JoE accepted her immediately.

Upper Gwynedd 6:
Understaffed; one machine operator for 3 machines; JoE was operating machines when machine operator was on break.

Poll Worker Issues – Lack of Training or Competence Concerns

Lansdale 3-1, Sons of Italy:
JoE required identification from every voter, an overtly illegal practice.  She defended the practice to the ERN watcher, saying: “If there are 8-9 Mohammads, I have to know which one it is.”  A Lansdale Democratic Committee representative had already notified the BoE, but the practice continued.  When the Council President of Lansdale spoke to the JoE, she threatened to eject him and called the BoE to complain.  The JoE said: “If you’re an American, you should have an ID!”  Observers reported to the ERN watcher that a number of people were sent home for ID; it was not known whether they returned.

Limerick 2, Township Building:
JoE required identification from every voter and told the ERN watcher s/he did it in order to speed the lines.  The JoE eventually agreed to stop the practice.

Abington 11-3, Abington High School:
An elderly poll worker with a severe hearing problem was the cause of much confusion and long lines.  She made many mistakes, including having people sign in on the wrong line and forgetting to have them sign in at all.  This caused a major slow-down in the morning.  In addition, to compensate for the poll worker’s hearing problem, voters often volunteered their ID for her to read.  This appeared to some observers that IDs were being required, which caused considerable controversy.

Abington 5-2 and 8-1, Willow Hill School:
Elderly poll workers were very slow, contributing to a 1 ½ hour wait in the morning.  The BoE was notified and another more efficient poll worker was sent in.  Comr. Hoeffel appeared and granted the poll worker(s) who were replaced a full day’s pay.  Insufficient organization and staffing also caused complaints from both JoEs about the division of the poll books.  The JoE for 8-1 complained about having only two poll books, while the JoE for 5-2 complained about having three books, rather than two. When 5-2 integrated the three books into two, wait-times seemed to be reduced.

Long Wait-times

Abington 5-2 and 8-1, Willow Hill School:
One and one-half hour wait in the morning, gradually dwindling to 10 minutes by noon.

Norristown 3-1, Eisenhower Middle School:
One hour wait in the morning.

Norristown 4-1, Laborers Local:
One hour and forty minute wait in the early morning.

Norristown 1-1, Warehouse:
Up to forty minute wait in the morning.

Norristown 3-2, Public Library:
One hour wait in the morning.

Inadequate Access for the Disabled

Lansdale 3-1, Sons of Liberty:
An ERN poll watcher witnessed a number of voters who turned around and went home without voting because of their inability to navigate the stairs.

Intimidation of Poll Watchers

Red Hill, Borough Hall:
An attorney-poll watcher was threatened and forced to leave by the constable after the JoE had enlisted her help regarding the problem of a disparity of 18 votes between the machine count and the numbers in the signature book.

Lower Providence 2-4, Eagleville School:
A voter was rejected because the name on his drivers license differed from the one in the poll book.  When a poll watcher asked the JoE why the man was not offered a provisional ballot, the constable got involved, accusing him of being disruptive.  The constable got increasingly agitated, shouted and, although unprovoked, attempted to use physical force to remove the watcher.   

Inadequate Signage

Collegeville, St. Eleanor’s Catholic School:
“Vote Here” signs placed outside of the church, rather than the school, which was the voting site.

Voter Registration Problems

Horsham 4-4, Landis Park Field House
Two prospective voters claimed to have moved and filed for re-registration at the same time, but only one was registered.

Abington 13-1, St. Paul’s Lutheran Church
Student claimed to be registered for Primary but was not found in the system.

W. Norriton 4-2, Marshall St. School
Young voter not found in system.  He believed that he had registered through Penn DoT.

W. Norriton 4-2, Marshall St. School
Voter not found in poll book.  BoE was called and said that individual last voted in 1992 but is          not registered now.  Voter claims his name was not in the book in 2004 general election but was allowed to vote at that time.  (name and address available)

W. Norriton 4-2, Marshall St. School
Voter not in book but has voter card.  Moved in July and claims to have voted in Phil. for Primary.  Computer says “cancelled.”  (name and address available)

W. Norriton 4-2, Marshall St. School
Individual thought that he had registered on-line.  He was not aware that he had to print out the application and mail it in.  (name and address available)

W. Norriton 4-2, Marshall St. School
Voter not in the poll book. Was found registered at an old Phil. address; had expected Penn DoT to change registration when he moved.  (name and address available)

W. Norriton 4-2, Marshall St. School
Voter not in poll book, had apparently moved.  BoE said he was registered at the warehouse. (name and address available)

W. Norriton 4-2, Marshall St. School
Voter not found in book but has registration card.  She claimed to have moved in January and re-registered in July.  Eventually found in the system with address in Whitpain 11, Stony Creek Elementary.  Allowed to vote, name added in back of book. (name and address available)

Write-in Votes

The ERN has learned of reports that all write-in votes in Montgomery County, at minimum 718 for three candidates for president, remain uncertified.  We request an explanation of the certification process of write-in votes, the total number of write-ins recorded in the 08 general election and their certification status.

Significance of Our Findings

Provisional Ballots:

The administration of provisional ballots remains a glaring weakness in election operations in Montgomery County.  Federal and state law require that they be offered as, in effect, a “last resort” to voters who believe that they are registered, even if their name is not found anywhere in the poll book (including supplemental lists) or registration data base on election day.  It should be clear that the likelihood of a registration decision favorable to the voter has no bearing on the obligation to provide provisional ballots.  Yet election judges in a significant number of precincts (see above: Election Day Incidents, Provisional Ballots) refused to provide provisional ballots based on the rationale that they “wouldn’t be counted anyway.”  In some cases, however, JoEs eventually reversed themselves and offered a provisional ballot in a given instance after consulting with an ERN poll watcher.

Perhaps more disturbing, for the second election in a row (including the 08 Primary), ERN poll watchers found, through conversations with JoEs, that BoE staff were instructing JoEs not to provide PBs in instances where the voter was not found in the registration data base system or was found at an address outside of the county.  This practice is in clear contravention of federal and state law (25 Pa. C.S. 3050 (a.4) (1)).

At this point, after some five years since the provisional ballot law took effect, there is no justification for this lingering problem.  It is beyond the time for the county to take remedial action to ensure the proper administration of provisional ballots, including, if necessary, termination of employment for any employees found to have knowingly failed to comply with federal and state requirements.

Another concern related to provisional ballots is Montgomery County’s process for determining whether and to what extent provisional ballots are valid and therefore counted.  State law establishes a means of public observation of decisions on provisional ballots as well as an opportunity to challenge any decision – requiring a hearing within seven days - by allowing party representatives to witness the process (Pa. Election Code, 25 P.S. 3050).  

Unfortunately, Montgomery County routinely provides no advance notice of the date, time and location of final determinations on provisional ballots and consequently, no public observers attend.  (This practice, which, in effect, evades the observation requirement, is, unfortunately, probably not uncommon in other counties across Pennsylvania.)  During the process of compiling the 08 general election official count, an observer from the Election Reform Network - with the required credentials - had explicitly requested attendance at the process determining PB eligibility, but was not advised of the opportunity to attend.  The Department of Voter Services later indicated that this had been an oversight.  

To ensure that this practice ends, we suggest that for future elections, the Montgomery County Election Board /Department of Voter Services institutes a regular practice of announcing the date, time and location of the process prior to election day.

Emergency Paper Ballots:

There is no question that the issue of EPBs was a difficult and controversial issue in the run-up to the 08 general election.  The BoE’s decision to  produce EPBs (in blue) distinct from provisional ballots (in green) was a significant contribution, and we hope one that will be adopted as part of normal procedures.  Of even greater significance was the federal court order handed down on October 29, which finally clarified that EPBs “shall be distributed immediately” when 50% or more of electronic voting machines in a precinct are inoperable.  

EPBs were issued at some Montgomery County precincts on election day.  In Upper Dublin 1-1, an ERN observer was told by poll workers on the scene that two machines failed to start up in the morning, yet EPBs were withheld until 10 am, a full three hours after the problem arose. (Our information indicates that they were the only machines deployed at the precinct, although that information has not been confirmed.)  At one point, the JoE told an ERN poll watcher that the BoE told him that there was no such thing as emergency paper ballots, yet another indication that a new approach to poll worker training is needed.  

The ERN has received reports that EPBs were issued at additional precincts.  We think it vital that a full report comprising a description and explanation of all situations related to EPBs in the 08 general election be made public by the Board of Elections.  This should include, at minimum: circumstances and timeframes related to the failure of the machines, including an estimated number of voters who were unable to vote as a result; when EPBs were issued; the number of EPBs issued and the number available at the time; and a full explanation of any and all situations involving practices which were out of compliance with the October 29 court order of the United States District Court for the Eastern District of Pennsylvania.

Poll Worker Recruitment, Appointment and Training:

The general election 08 again found significant understaffing at the polls, although anecdotal evidence indicates that the situation was improved over recent elections.   The Election Reform Network, in conjunction with Poll Workers for Democracy, recruited some 80 people to fill poll worker positions, yet barely a few were eventually appointed.  Even on election day, we had reports of people who had been trained but unable to gain appointments, even after witnessing vacancies at the polls and requesting appointments on the spot, only to be turned down again.  

We ask that the Board of Elections/Department of Voter Services provide the ERN a full roster of the poll worker assignments for the 08 general election and identify any and all vacancies at the polls on November 4.

The recruitment and appointment system for poll workers needs to be revamped with control centralized.  Despite a significant level of vacancies, eligible citizens attempting to apply for poll worker positions for the November 08 election experienced what amounts to a “run-around,” as they were shifted from the Election Board to various levels of their party organization and still were unable to secure a position or even gain the courtesy of a rejection.  Clearly, a new approach is needed.  We urge the Board of Elections to sit down with the political parties and non-partisan voters’ organizations to work out a new arrangement that streamlines the poll worker recruitment and appointment process.   This will enable some welcome “new blood” to flow into poll worker positions.  There is no reason why capable, responsible citizens with the motivation to serve should continue to be shut out of a system in desperate need of exactly what they offer.  

It is also clear that poll worker training needs a new approach, a point made in past years by the ERN.  Some poll workers were, for example, completely unfamiliar with EPBs, continued to require IDs for all voters and failed to issue provisional ballots, in violation of federal and state election law.  A modest amount of resources needs to be put in place to design a new training structure that supplants the current lecture format (sometimes supplemented with groups converging around machines for demonstrations).  

The curriculum should be systematized and expanded so that, for example, emergency paper ballots are included in all training sessions, rather than at some and not others.  Training methods need to be professionalized, grounded in adult learning concepts and tools so that poll workers have an opportunity to engage with the material rather than just be talked at, and more time needs to be set aside for questions and clarifying discussion.  Paying for attendance at training was an important innovation and should be continued, but more is required.  Expanding and improving recruitment would also allow training to become genuinely mandatory, while minimizing any threat of a drop-off in staffing, if people do not show up.  An increase in stipends should also be considered.  

While this is an awkward time for a budget increase, there is strong reason to believe that the Board of Elections/Department of Voter Services in Montgomery County has been substantially underfunded – other than for vote machine technology - for a long time.  The increase in voter trust and efficiency should generate an attractive return on investment.   

We ask that the Board of Elections/Department of Voter Services provide the ERN with the number and proportion of poll workers serving on Nov. 4 who attended one or more trainings just prior to the election.

Machine Problems and Voter Confusion

Twelve years after the introduction of the Sequoia direct recording electronic vote machines to Montgomery County, many voters continue to be confused about using them, often for good reason.  Prior to the general election and on election day 08, there was an outpouring of concern about whether the use of the “Straight Party Vote” touch-screen would work correctly - recording and counting votes accurately.  Voter concern was substantial and real, regardless of whatever the motivations may have been of those responsible for raising the issue in the first place.  

We believe that Boards of Election in conjunction with the Department of State should have responded with a substantive public information campaign to allay fears by explaining the workings of straight party voting.  Public demonstrations and explanations of the process could have been undertaken, the press could have been mobilized and much of the concern may have been alleviated.  Instead, concern festered because a clarifying response from elections officials never materialized.

The “Do Not Wish to Vote” button is another source of confusion, although not with the level of importance of straight party voting.  First, some voters noticed that the light for the “Do Not Wish to Vote” button stayed on, even as the “cast vote” button was pushed.  Not surprisingly, they questioned if their votes were going to be recorded and counted when the machine seemed to be registering a non-vote.  Moreover, the real function of the “Do Not Wish to Vote” button remains something of a mystery, considering that, based on a voting official’s explanation, activating it actually does not nullify any selections that have been made when the “cast vote” button is pushed.  The obvious question, then, is why have the option at all?

The problem of “fleeing voters,” a catch-phrase for voters who make selections but do not have their votes counted because they neglect to press the “cast vote” button prior to leaving the booth, lingers on.  A preliminary study done by the Election Reform Network in 2007  based on the 2006 election indicated a disparity of about three votes per precinct between the total vote count on the machines and the number of voter sign-ins.  Reports of a disparity of as much as 18 votes between the machine count and the signature count for a single precinct (Red Hill) in the last election alone argues for a comprehensive comparison between machine vote counts and total signatures for every precinct.  After twelve years with the Sequoia AVC Advantage machines , it is time for a definitive assessment of the magnitude of the problem of “fleeing voters.”  More than 1000 voters across the county may be disenfranchised every election because of the inadequate interface between voters and the machine voting process.  

Summary Recommendations and Outstanding Questions

The Board of Elections should undertake a full investigation of the administration of provisional ballots and take comprehensive remedial action to address wrongful practices through training and other means, including, if necessary, termination of employment for any employees knowingly failing to comply with state and federal requirements.  

The Board of Elections should institute a regular practice of announcing the date, time and location of the process of determining the validity of provisional ballots prior to election day.

The Board of Elections should fully disclose all situations involving emergency paper ballots and the procedures employed for their use in the 08 general election. This should include, at minimum: circumstances and timeframes related to the failure of the machines, including an estimated number of voters who were unable to vote as a result; when and where EPBs were issued; the number of EPBs issued and the number available per precinct.  The report should also include an explanation of any and all situations involving practices which were out of compliance with the October 29 court order of the United States District Court for the Eastern District of Pennsylvania.

The recruitment and appointment system for poll workers needs to be revamped – within parameters of current law - with control centralized.  We urge the Board of Elections to sit down with the political parties and non-partisan voters’ organizations to work out a new arrangement that streamlines the poll worker recruitment and appointment process.  Expanding and improving recruitment would also allow training (see below) to become genuinely mandatory, minimizing any threat of a drop-off in staffing, if people do not show up.  An increase in stipends should also be considered.  

Poll worker training should be redesigned.  Training methods need to be professionalized, grounded in adult learning concepts and tools so that poll workers have an opportunity to engage with the material rather than just be talked at, and more time needs to be set aside for questions and clarifying discussion.

A definitive assessment of the magnitude of the problem of “fleeing voters” should be undertaken.  Reports of a disparity of as much as 18 votes between the machine count and the signature count for a single precinct (Red Hill) in the last election alone argues for a rigorous comparison between machine vote counts and total signatures for every precinct.  Thousands of ballots may be at stake.

Information Requests:

We ask that the Board of Elections/Department of Voter Services provide the ERN a full roster of poll worker assignments for the 08 general election and the number and proportion of vacancies at the polls on November 4.
We ask that the Board of Elections/Department of Voter Services provide the ERN with the number and proportion of poll workers serving on Nov. 4 who attended one or more trainings just prior to the election.

Please provide the ERN with a list of polling places which do not comply with federal and state disability standards.

Please explain the certification process of write-in votes and provide the total number of write-ins recorded in the 08 general election and their certification status.

Please explain the definition of “Under Votes,” as posted on the County website as part of the unofficial and official election results and indicate why the presidential undervote increased by 15 from the unofficial to the official/certified results.



We look forward to receipt of the County’s own report on the 08 general election and to a frank and constructive discussion of the data and future opportunities for innovation in election practices.  The Election Reform Network stands ready to work with the Board of Elections and Department of Voter Services on the ways that citizens and the county can collaborate in the interest of the continuous improvement of the system of elections in Montgomery County.