Number of Registered Voters Per Metropolitan-Wichita Poll Site |
Number of Sites (N=43) |
Percent of Metropolitan Sites |
5,000+ |
7 |
16 |
4,500-4,999 |
10 |
23 |
4,000-4,499 |
11 |
26 |
3,500-3,999 |
9 |
21 |
3,000-3,499 |
4 |
9 |
2,500-2,999 |
2 |
5 |
In designating sites and times to observe, SCVC placed special emphasis on polling sites assigned 4,000 or more registered voters during traditionally peak voting times: early morning, lunch, after work.
SCVC members designed two survey instruments to guide poll observers. The first instrument was designed for SCVC/LWV members to complete while observing the polls. For each polling site, the survey asked the observer to document:
- ADA-related and other features of the facility
- direct or indirect electioneering at the site
- the number of machines and poll workers present
- how poll workers handled voters who were confused about their proper polling site
- the physical arrangement, privacy features, and operation of the new voting machines
- the length of time it took randomly selected voters to complete the entire voting process.
The second survey instrument was brief anonymous exit interview questionnaire designed to see if voters encountered any difficulties in knowing where to vote or how to vote using the new machines, etc.
2. OBSERVATIONS ABOUT THE POLLING SITES
2.1 SIGNAGE
Observers found that there was no site-identifying signage at the street at 11% of the polling sites observed. All sites had signage at the door of the polling sites. All sites used directional arrows and posted signage on the door (including a sign banning electioneering within 250 ft of the site).
In spite of SCVC's repeated recommendation after the August Primary Election that the Election Commissioner dramatically increase the size of site-identifying signage near street entrances, observers all reported that the street signage provided by the election office--red, white, and blue signs measuring 26" x 16" (416 sq in) that are no larger than a typical yard sign--was too small, too low to the ground, and not vibrant enough to command attention.
2.2 ADA-RELATED ISSUES
Overall, the polling sites technically complied with ADA requirements with each having at least 2 handicap-designated parking spaces and entry and egress either level with the parking area or accessible via ramps and curb cuts. That said, observers noted that technical ADA compliance did not universally translate into convenience or ease of use:
- At many of the sites we observed, peak voting times on Election Day saw long lines to check-in at the registration tables and, subsequently, to cast ballots on the touch screen machines. As a result, while waiting chairs were provided at many sites, infirm voters, nonetheless, had to endure long periods of time on their feet in order to vote.
- At Faith Christian Church (2110 W 45th St S), the two designated handicap parking spaces were located some 70 feet away from the only door to the polling site. What is more, the door to the site was a heavy fire door with no automated way for a person in wheel chair to enter unassisted.
- At Glenn Park Christian Church (2757 S Glenn), the main point of entry to the polling site was about 25 feet from the designated handicap parking spaces but the door was too narrow to permit wheelchair-bound voters to enter. A sign posted at the door directed voters with disabilities to enter through the main doors of the building and navigate to the entry of the gymnasium to enter the polling area. However, our observer noted that voters with disabilities had to travel over 90 feet from the main doors to enter the gym. What is more, there was no signage inside the main doors of the facility to guide voters with disabilities to the gym.
- At Central Christian Church (2900 N Rock Rd), handicap parking, a curb cutout, and a ramp were all available but not convenient or in close proximity to the polling area. At the South end of the facility, only a voter acquainted with the building would have known where to go without searching and would, even then, been required to walk some distance to enter the polling area. Observers noted that the parking area at the South end of the site was not handicap designated, there was no curb cutout, and voters had to climb a significant number of stairs to enter the site. What is more, the registration tables were located at the center of the huge church complex, meaning that anyone entering from the North or South had a long hike to get to the check-in area.
- At Evergreen Recreation Center (2700 Woodland), our observer documented that the voters with disabilities had to travel some 120 feet from handicap accessible parking to the handicap accessible entry point.
- At Olivet Baptist Church (3440 W 13th), our observer documented the following:
"This church has horrible accessibility issues....The preferred entrance on Nov. 7th was the entrance on the west side of the building. This entrance does have a ramp, which goes to the sidewalk. There is no curb cut to the street. The small parking lot across the street to the west does not have handica[p] parking. There is one handicap parking space to the north of the church off of High [St], but it is not easily identifiable as belonging to the church. It is 54 steps from this [handicap parking] space to the door of the church. I saw a man with a cane struggle up the curb from the street. There is a parking lot to the north of the church, which has 3 handicap parking spaces (one was blocked off). It is 20 steps from the handicap [parking] space [on the north] to the door, then another 60 steps from that door to the entrance to the [polling area in the] gym. I saw an elderly man using a walker walk from [the north] lot to the polling place--I didn't think to time him, but he clearly was struggling with the distance he was required to walk."
Another ADA-related issue that came up at one poll site also deserves mention. At the Sedgwick County Health Department Administrative Center poll site, our observer documented 3 instances where low vision voters were not offered the touch screen machine's audio-assist feature in order to vote.
2.3 PARKING
In addition to the handicap-accessible parking issues noted above, at 8 of the sites we observed--Olivet Baptist, Dawson United Methodist, Center for Health and Wellness, Grace Presbyterian, Gloria Dei Lutheran, Pleasantview Baptist, Faith Christian, and Metro East Baptist--there was insufficient on-site parking places at peak voting times on Election Day to accommodate voters. Consequently, voters had to park on the street and/or on the grass of the polling site.
2.4 CONVENIENCE OF SITES
Results from the exit interviews we conducted at 13 poll sites found that 71% of voters interviewed did not have to travel more than 2 miles to get to their polling site. Twenty-nine (29) percent of those interviewed had to travel between 2-5 miles. The largest percentage of those interviewed who traveled more than 2 miles in order to vote was assigned to the Park City Senior Center.
When asked to respond to the statement "compared to your previous poll place, your new site is more convenient," the 35 voters we interviewed replied as follows:
Completely Agree: 31%
Somewhat Agree: 9%
Agree: 17%
Disagree: 9%
Strongly Disagree: 23%
No response: 11%
The greatest number of voters expressing dissatisfaction with the convenience of their post-consolidation polling site came from voters at the Park City Senior Center. Voters interviewed at the Hughes Metroplex, St Andrew's Lutheran, Westlink Church of Christ, Bel Aire City Building, and Gloria Dei Lutheran Church also voiced strong dissatisfaction with the reduction in poll sites.
2.5 OTHER COMMENTS ON SITES OBSERVED
As during the August primary, the polling areas at Maranatha Worship Center and Central Christian Church were set up in large, open spaces and, thus, were not secure--church staff and persons not voting were able to walk through the polling area in close proximity to voters and voting machines from all directions all throughout the day.
Voting at the remaining poll sites was conducted in secure rooms and/or gymnasiums. However, our observers found that the rooms utilized at some of the sites were either too small or not arranged optimally to ensure comfort and convenience for voters, especially for an election with such sizable voter turnout. For instance, at the Independent Living Resource Center site, the polling area was set up in room of sufficient physical size to conduct the election. However, the actual setup of the polling area was very cramped because nearly one-third of the room's area was occupied by furniture that had been pushed into the corner to make room for the voting area. If this furniture had either been removed from the room or stacked compactly, the space for the voting area could have been optimized to improve traffic flow and comfort for those waiting to vote. Instead, during peak times there were occasions when the center of the polling area was occupied simultaneously by voters either waiting to check-in or waiting for a touch screen machine or waiting to pickup or return a paper ballot.
Similar observations about cramped, confusing, and/or poorly prepared voting areas were noted by observers at Park City Senior Center, Faith Christian Church, Mount St Mary's Convent, Gloria Dei Lutheran, St Andrew's Lutheran, Dawson United Methodist, and the Center for Health & Wellness.
3. OBSERVATIONS ABOUT THE VOTING PROCESS
3.1 GENERAL OBSERVATIONS
Throughout the period of their observations, our poll watchers were asked to randomly select voters as they entered the polling area and measure the length of time it took the voters to complete the voting process. While we found few, if any, delays during the August primary election, our results for the high turnout general election were quite different.
At many of the sites we observed, our observers found that voters who arrived at their correct polling site experienced, in general, a wait of no more than 5 minutes to check-in. However, we did document instances over the course of the day when the wait to check-in was far more substantial--ranging from 10 minutes between 7-8 AM at St Elizabeth Ann Seton to 24 minutes at the Park City Senior Center between 10:30-11:30 AM.
The delays experienced by voters to check-in, however, were nothing compared to the delays to cast ballots on the touch screen voting machines. The most egregious delays documented on Election Day were at the Metro East Baptist Church. The site was assigned 5,351 registered voters and had only 10 touch screen machines to accommodate them. Upon arriving at the site at 6:30 PM, our observer estimated that there were over 100 people who had checked-in and were waiting to access a voting machine. Inquiring with and following the subsequent progress of three voters in line, our observer noted that it took 65 minutes, 60 minutes, and 45 minutes respectively for each of these voters to access a voting machine. From 6:30 PM to 7:00 PM--official close of polls--our poll watcher observed a continuous line of 100 voters waiting for a machine. At 7:00 PM, however, the vast majority of those still waiting simply gave up and left the poll site.
Moderate to lengthy wait times for voting machines after checking in were also documented at the following polling sites:
- Glenn Park Christian Church: At noon, 26 voters were waiting in line for a voting machine. The longest wait time measured for a voter to get to a voting machine was 13 minutes.
- Evergreen Recreation Center: At 5:45 PM, 40 voters were waiting in line for a voting machine. The longest wait time measured for a voter to get to a voting machine was 35 minutes.
- Park City Senior Center: Around 11:00 AM the longest wait time measured for a voter to get to a voting machine was 23 minutes.
- St Elizabeth Ann Seton: At 6:30 PM there was a 30 minute wait to get to the voting machines. At 7:00 PM, there were still 18 voters in line to for voting machines.
- Evergreen Recreation Center: At 5:45 PM there were approximately 40 voters in line waiting to cast their ballots. The longest wait time measured for a voter to get to a voting machine was 35 minutes.
- Pleasantview Baptist Church, Derby: The longest wait time measured for a voter to get to a voting machine was 20 minutes. A poll worker indicated that there had been long lines consistently since 6 AM.
- Gloria Dei Lutheran: At 5:45 PM there were 55 people in line. By 6:10 the number of voters waiting in line swelled to 87. At 6:25, 63 voters in line.
- Westlink Christian: At 9:30 AM there were 26 voters in line waiting for machines; 33 at 9:50 AM; 36 at 10:00 AM. In the afternoon, 31 voters in line waiting for machines at 4:00 PM; 36 at 4:15 PM; 50 at 4:30 PM; 50 at 6:00 PM.
- Kechi City Building: A voter emailed us reporting that when he went to vote around 4:00 PM he found that "there was an hour wait in line with 3 machines and no [paper] ballots. They said the [paper] ballots were on the way."
- Church of the Holy Spirit, Goddard: A voter emailed us reporting that when she went to vote at 10:30 AM, there "was a line with a wait of about 30 minutes or more."
Also as with the primary election, we observed that the length of time voters took to vote and the extent to which additional assistance was required directly correlated to voters' age and technical savvy--overall, younger, technically savvy voters spent the shortest amount time voting and required little, if any, additional assistance whereas visually-impaired and elderly voters and those not familiar with computers required additional assistance more frequently to complete their ballots.
3.2 TOUCH SCREEN VOTING MACHINES
Prior to the general election, the Election Commissioner's office declined to make public its intended ratio of assigned registered voters to new touch screen voting machines. Data collected from site observations indicate that the ratio of assigned registered voters to touch screen machines varied dramatically between the sites--from a low of 429:1 at Eastminister Presbyterian in Northeast Wichita to a high of 585:1 at Pleasantview Baptist in Derby. Overall, the county-wide ratio averaged 510 registered voters to each touch screen machine.
Number of Registered Voters Assigned per Voting Machine at 36 Sites Observed
|
SITE |
Registered Voters Assigned to Site* |
Number of Touch Screen Machines |
Ratio of Assigned Voters to Machine |
Bel Aire City Building |
4,163 |
8 |
520:1 |
Calvary Baptist ( Derby ) |
4,174 |
8 |
522:1 |
Central Christian Church |
5,229 |
10 |
523:1 |
Center for Health & Wellness |
3,203 |
6 |
534:1 |
Dawson United Methodist |
3,353 |
6 |
559:1 |
Eastminster Presbyterian |
3,859 |
9 |
429:1 |
Edgemoor Recreation Center |
4,164 |
8 |
521:1 |
Evergreen Recreation Center |
2,956 |
6 |
493:1 |
Faith Christian Church |
4,501 |
9 |
500:1 |
First Baptist Church |
3,111 |
6 |
519:1 |
First Mennonite Brethren |
5,060 |
10 |
506:1 |
First Methodist, Valley Center |
4,040 |
8 |
505:1 |
Glenn Park Christian |
4,154 |
8 |
519:1 |
Gloria Dei Lutheran |
5,209 |
10 |
521:1 |
Grace Presbyterian |
4,663 |
9 |
518:1 |
Independent Living Resource Center |
4,111 |
8 |
514:1 |
Linwood Recreation Center |
4,899 |
10 |
490:1 |
Maranatha Worship Center |
4,904 |
10 |
490:1 |
Metro East Baptist |
5,351 |
10 |
535:1 |
Mount St Mary's Convent |
3,785 |
7 |
541:1 |
Northside Church of Christ |
4,397 |
9 |
489:1 |
Olivet Baptist |
4,618 |
9 |
513:1 |
Park City Senior Center |
3,917 |
8 |
490:1 |
Pleasantview Baptist |
4,683 |
8 |
585:1 |
SG Co. Health Dept Admin Center |
3,812 |
8 |
477:1 |
Sharon Baptist |
4,401 |
9 |
489:1 |
St Andrew's Lutheran |
4,624 |
10 |
462:1 |
St Elizabeth Ann Seton |
4,948 |
9 |
550:1 |
Temple Emanu-El |
2,516 |
5 |
503:1 |
West Evangelical Free |
4,525 |
10 |
453:1 |
West Side Church of God |
4,469 |
9 |
497:1 |
Westlink Christian Church |
5,025 |
9 |
558:1 |
Westlink Church of Christ |
4,682 |
9 |
520:1 |
Westwood Presbyterian |
4,083 |
8 |
510:1 |
Woodlawn United Methodist ( Derby ) |
4,577 |
9 |
509:1 |
WSU Hughes Metroplex |
3,916 |
8 |
490:1 |
|
|
AVERAGE |
510:1 |
|
- Number of assigned registered voters is from data obtained from Election Commissioner's office, August 2006.
In general, the touch screen machines functioned properly and voters reported few problems using them and appreciated the ability to review and correct their ballots before submitting their vote. That said, our observers found multiple instances of poll sites throughout the county with one or more touch screen machines that were completely out-of order. At Glenn Park Christian, St Andrews Lutheran, Edgemoor Recreation Center, Maranatha Worship Center, and WSU Hughes Metroplex, our observers found one or more touch screen machines out of order due to reasons other than needing a replacement roll of paper. In some instances, county workers were able to get the machines on-line; in other instances not.
In addition, at numerous sites our observers witnessed technical problems with one or more machines. Most often reported were paper jams or running out of paper throughout the day. Given the length of ballot for the general election--four pages--each roll of paper in any given machine could record only 100 votes before a new roll of paper was needed. Due to strong voter turnout, there was a constant need to replace rolls of paper, which only added to the delays voters experienced waiting for machines. Our observers noted some instances when poll staff and/or on-hand supplemental county staff were able to efficiently change out the paper in the machines. In other instances, however, observers found that poll staffs were unable to efficiently change the rolls of paper. At one West central site, an observer noted that two machines ran out of paper shortly after 10 AM. Poll workers tried to resolve the matter themselves, but after 40 minutes of trying without success, they finally requested that a trained county worker be dispatched to resolve the issue. In the end, it took over an hour to get the two machines on-line again causing a backlog of voters waiting to vote on the 6 remaining machines. At another West side site, nearly all the machines ran out of paper at the same time resulting in lines of 30-50 voters waiting for machines. At still another poll site on the South side, an observer found that only 5-6 of the 8 machines at the polling site were working regularly during the busy lunch hour due to paper outages on the remaining machines. (Poll workers were able to change out the paper in these machines in an average time of 15 minutes.)
In addition to delays caused by inconsistent technical training on changing paper, our observers also documented instances where poll sites were insufficiently stocked with extra rolls of paper for the machines. At Gloria Dei Lutheran, our observer found that the site had been supplied with only 3 replacement rolls of paper for 10 machines. As a result, at one point during Election Day, 6 of the 10 touch screen machines at the site were unavailable for use due to the lack of replacement paper. Poll workers at the site reported great difficulty obtaining replacement rolls from the Election Office throughout the day.
As during the primary election, our poll watchers did not observe voters verifying their on-screen ballot selections against the voter verifiable paper audit trail (VVPAT) printout. The lack of widespread use of VVPAT by voters is not surprising, however, since our observers reported that some (but not all) poll workers informed voters about this feature at only 11 (or 30%) of the poll sites.
Observers also noted instances of voters who were uncertain how to finish and/or leaving their machines when they were done assuming that they had cast their vote when, in fact, they had not completed the process. (In these instances, observers witnessed poll workers either assisting voters or completing the process for the voters who had left the site without altering their ballot selections.) There was also voter confusion as to the procedure for writing-in candidates.
3.3 OPTISCAN VOTING MACHINES & PAPER BALLOTS
While all sites had a section of the polling area designated for voters to complete paper ballots, voters were not routinely and universally offered a choice of voting method when they checked in. In most circumstances, unless voters explicitly requested a paper ballot, poll workers directed voters to the queue for the touch screen machines. And, indeed, at several sites we observed--like the Independent Living Resource Center--voters periodically requested to vote by paper ballot, often indicating that they did not trust or feel comfortable with the touch screen machines.
While the paper ballot option provided a useful safety valve to address the high volume of voters, our observers documented numerous instances when the Optiscan M100 machines either were not working at all or consistently failed to scan/read the paper ballots that voters tried to insert after they had voted. Regarding the former, our observer at Glenn Park Christian learned that the Optiscan delivered to the site had not worked since the polls opened. As for the latter, our observers found that numerous voter-completed paper ballots failed to be scanned by the Optiscan machines at Linwood Recreation Center, Evergreen Recreation Center, West Free Evangelical Church, Gloria Dei Lutheran Church, St Andrews Lutheran Church, Mount St Mary's Convent, and Olivet Baptist Church. The failure of these machines to scan/accept the ballots may have been due to ballot printing errors, voters using ink pens that bled, or the lack of clear instructions by poll staff telling voters how they should complete the ballot and/or voter error completing the ballot (e.g., placing a check mark or "X" in the oval for the candidate they chose rather than filling in the oval as required). Whatever the source of the error, in some instances (e.g., Olivet Baptist, Independent Living Resource Center, and West Free Evangelical) poll staff voided the errant ballot and let voters complete another (in several instances, multiple) paper ballot(s) until the machine accepted one. In most instances, poll staff directed the voter to insert the ballot into a secure slot on the front of the Optiscan machine to be scanned downtown after the polls closed. However, in one troubling instance (at Gloria Dei Lutheran) poll staffers left a pile of over 75 voter-completed, but unsuccessfully scanned, paper ballots unsecured in a pile by the provisional ballot table very near an outside door. Our observer noted that "with no one watching, I could have walked off with the entire pile of 75+ paper ballots." At still another site (Linwood Recreation Center), poll staff was unclear on the procedure for handling paper ballots and, initially, treated them as provisional ballots and sealed a number of them in envelopes.
In spite of these Optiscan machine problems, at some sites--like Glenn Park Christian--the high volume of voters led poll staff to approach and inform voters waiting for touch screen machines that they could fill out a paper ballot instead--an option a fair number of voters took. However, at other sites (such as Gloria Dei Lutheran) our observers found that due to the Optiscan machine spitting back paper ballots, poll staff were directed not to offer voters the option of paper ballots when they checked in. This decision was made even though it was 6:25 PM with 63 voters in line waiting for voting machines.
Another problem created by high voter turnout and the recurrent errors with paper balloting was a shortage of paper ballots that many sites experienced throughout and, most especially, at the end of Election Day. Our observer at Gloria Dei Lutheran reported that the site had run out of paper ballots for some precincts by 6:50 PM. At Metro East Baptist--where 100+ voters were waiting in line for a voting machine up until the polls closed at 7:00 PM--the site had completely run out of paper ballots by 6:30 PM. Our site observer reported that the supervisory judge "discovered" 8 paper ballots around that time and a mad scramble ensued as voters in line tried to be one of the lucky eight to be able to vote via paper ballot. Some voters even asked if they could cast their vote on the sample ballots, but they could not.
A final note about the handling of non-provisional paper ballots during the 2006 general election. Observers who stayed after the close of the polls to watch the initial poll site canvasses (as permitted under Kansas election law(1)) noted that the Election Commissioner's office directed all supervisory judges to manually input the votes cast on non-provisional paper ballots that were not scanned by the Optiscan machines into the touch screen machines. In ordering this, the Commissioner's objective was to have (provisional ballots aside) near complete original canvass results delivered by each site to the central Election Office in the memory cards and printouts from the voting machines in order to more quickly provide election results to the public.
While perhaps well intentioned, the coalition is concerned with the directive the Election Office issued. Ordinarily, the accumulated non-provisional paper ballots that could not be successfully scanned would be manually tabulated at the original poll site and then transported to the Election Office where the manual totals would be added to the county-wide totals accordingly. However, the Election Office's directive required already fatigued poll staffs to suspend closing and totaling out the voting machines and, instead, strive to accurately program the touch screen machine for the proper precinct and then enter the choices voters' made on paper ballots for each contest into the four page touch screen ballot. (Our observers noted that this was made trickier owing to the fact that the order of the races on the paper ballots did not always correspond to the order of the races on the touch screen ballot.)
4. OBSERVATIONS ABOUT VOTER CONFUSION
At 27 (or 75%) of the sites observed, we witnessed voters over the course of Election Day who were told by poll workers that they had come to the wrong polling location. During 33 periods of observation at the 36 polls we watched, our observers found 24 instances when between 1-5 voters per hour arrived at the wrong poll site; 7 instances when 6-12 voters per hour arrived at the wrong poll site; 1 instance when more than 12 voters per hour were told that they were at the wrong polling site.
As we found during the primary election, the site with the largest occurrence of confused or "lost" voters trying to determine where to vote was at Central Christian Church in northeast Wichita. A poll worker our observer spoke with estimated that by 8:30 AM, 60 people had attempted to check-in at Central Christian, assuming it was their designated poll site, only to be told by the poll workers that they had come to the wrong site. Other sites with high incidences of "lost" voters per hour include Edgemoor Recreation Center, First Baptist (Wichita), Glenn Park Christian, Gloria Dei Lutheran, Grace Presbyterian, Olivet Baptist, and West Evangelical Free Church.
The majority of these voters did not voice any frustration, but observers did note several instances of irate voters--one voter at Faith Christian Church, overheard by an observer, was very upset, having already been to two sites already trying to find his official site; he demanded the name of someone to whom he could issue a formal complaint. In several cases (at Central Christian Church, Pleasantview Baptist (Derby), Woodlawn United Methodist (Derby), and the Sedgwick County Health Department Administrative Center), voters reported having been sent to multiple polling sites--as many as three others--in an effort to find their correct site.
Voters most often stated that they had either voted at the site in past elections--"This is where I've always voted"--or assumed that the site was correct because it was closest to their home. As during the primary election, voters arriving at incorrect sites also expressed frustration with the precinct map printed in Wichita Eagle on Election Day stating that they came to site after analyzing the map only to find out that they were at the wrong location when they arrived.
Another source of voter confusion on Election Day came as result of voters who had received mail-in ballots that they insist they had not requested. At Evergreen Recreation Center, poll workers reported to our observers that over 30 voters came to vote claiming that they received a mail-in ballot by mail that they had not requested. Similar occurrences were reported by our observers at Gloria Dei Lutheran and Maranatha Worship Center. (In each instance, our observers noted that poll staff let these voters cast provisional ballots.)
4.1 POLL WORKER RESPONSES TO CONFUSED OR "LOST" VOTERS
Observers found that poll worker responses to confused voters were inconsistent across the polling sites observed and demonstrate a clear need for further improvements in training, tools, and procedures in this area. Voters whose name did not appear in the registration book were most often directed to a poll worker working the area of the site for casting paper and provisional ballots. This worker then tried to identify the voter's correct polling site by consulting a very large, detailed precinct map--there was no uniform procedure established to call the election office or consult a computerized database. It must be noted that the GIS maps prepared for the general election provided much more detail for the area of the county surrounding the poll sites than the maps used during the August primary election. (At one polling site in Derby, our observer noted that, as during the August primary election, poll workers set up an ad hoc system of calling between the three Derby sites in order to precisely direct voters to the site that had their name on the rolls.)
4.2 INCONSISTENCIES WITH PROVISIONAL BALLOTS
As during the August primary election, the options poll workers presented "lost" voters diverged dramatically:
- At some sites poll workers followed proper procedure by offering voters the option of either casting a provisional ballot or traveling to their assigned site to vote(2). These workers further advised voters that casting a provisional ballot at a site other than their assigned one would likely deny them the ability to vote on races specific to their home address. Observers noted that among voters so informed some opted to vote provisionally, others chose to go to their proper site to ensure that they could vote on races specific to their address.
- At other sites--chiefly a polling site in a large facility in northeast Wichita--poll workers informed voters of what appeared to be their "correct" polling place after reviewing the precinct map and then sent them off without offering them the choice of casting a provisional ballot. One poll watcher at this site observed numerous voters arriving who did not appear on the rolls. Most of the time, the poll staff simply told the voters that they were at the wrong site. Sometimes the poll staff directed the "lost" voter to a precinct map in an attempt to help the voter figure out where they needed to go. Poll staff did not routinely offer or explain provisional ballots. And, in one instance when a provisional ballot was offered, our observer found that it did not occur without a struggle: "A young woman [whose name was not found in the registration books] insisted that she was already late for work and could not go to [another poll site]. A check-in worker took her to the paper ballot area and told the worker to give [the woman] a provisional ballot. Much arguing among the two workers ensued before one produced a [provisional] ballot. [Neither worker] explained to the voter what a provisional ballot was. When she was finished, she gave the ballot to the worker who scanned it [into the Optiscan machine] and did not put it in a provisional envelope."
- At still other sites voters were not given full information to make an informed choice: Voters who were at the wrong polling site were offered a provisional ballot, but were not told that the ballot would likely not be correct for their address.
The SCVC is extremely concerned with the emerging pattern of inconsistent responses by poll workers to "lost" voters, especially at the northeast Wichita polling sites we observed. The failure to provide "lost" voters with complete information to permit them to make an informed choice risks disenfranchising them, either because they are frustrated at being told to go elsewhere to vote or because they are not informed of the differences in the provisional ballot from their official precinct ballot and are assured that their ballot will count.
For most other occurrences--e.g., change of address, name change, unused mail-in ballot--observers noted that provisional ballot procedures were followed correctly.
5. OBSERVATIONS ABOUT ELECTIONEERING
Our poll watchers found few examples of overt electioneering by candidates at the poll sites we observed.
However, at St Andrews Lutheran, Westwood Presbyterian, and Eastminister Presbyterian campaign signs were found near the entrances to the polling sites early in the morning. In each instance the signs were promptly removed by poll staff. However, in the case of Eastminister Presbyterian, our observers were informed by the supervisory judge that after she had initially removed the campaign signs, they were put back up later that morning. The supervisory judge took them down again and there was no further incident.
While some voters were concerned, and even upset, by the presence of poll watchers from political parties behind poll workers at the check-in tables, their presence was completely permitted under Kansas election law. And at no time did we witness any of these observers saying or doing anything advocating for a specific candidate over another.
At Central Christian Church, our observer noted the presence of large banner behind that check-in table that read "Reaching for the World of Jesus." Poll staff made no effort to remove or cover the banner. At Westlink Church of Christ, religious materials were on display on a table within 20 feet of the polls. Again, no attempt was made to cover or remove the materials.
6. OBSERVATIONS ABOUT POLL WORKERS
In general, observers were impressed by the hard work and dedicated efforts of poll workers to assist voters with the dramatic changes in polling sites and voting equipment, especially in the context of such high voter turnout. Data from the 36 sites observed indicates that the Election Commissioner's office achieved a 1 to 1 ratio (often better) of poll workers to machines and an average ratio of 1 poll worker for every 424 registered voters.
While helpful, efficient, and competent, SCVC observers noted areas where poll worker training needs to improve. In addition to persistent problems administering provisional ballots (see section 4.2), poll workers were not fully versed on the new equipment. As noted above (section 3.2), the chief problem encountered was the need to change rolls of paper for the touch screen machines. Due to strong voter turnout, there was a constant need to change out rolls of paper, which only added to the delays voters experienced waiting for machines. Our observers noted some instances when poll staff and/or supplemental county staff on-hand were able to efficiently change out the paper in the machines. In other instances, however, observers found that poll staffs were unable to efficiently change the rolls of paper, resulting in lengthy delays.
Poll workers we spoke with wanted more touch screen machines as well as staff so their sites could better process the volume of voters that turned out for the general election.
Also, poll workers observers spoke with universally complained about the 15-hour day and poor pay. Many suggested having two shifts and increasing the pay. Taking steps like these could increase the available pool of workers and ensure less fatigued and error-prone workers.
7. CONCERNS ABOUT VOTER PRIVACY
The coalition commends the efforts taken by Election Commissioner's office between the August primary election and the general election to ensure that county poll sites were arranged and that voting machines were augmented to better comply with the voter privacy provisions of Kansas election law(3).
At nearly all the sites we observed, our poll watches reported that the side doors (or wings) of the touch screen machines were oriented perpendicular to the touch screen to better guard against voters on either side of the machine seeing what the voter was doing. In addition, poll staffers were provided with hoods--made of single piece of black plastic or paper material--to place on top of the machine and fasten to the side doors by means of two metal binder clips. Overall, this inexpensive fix achieved technical compliance with Kansas election laws, though our observers found several problems with the solution. First, the touch screens of wheelchair bound voters were still quite visible to any voter or poll worker passing behind the voter's machine. Second, the height of the hoods attached to the wings of the machine made it difficult for tall voters to cast their ballots without having to kneel down or bend over to see the screen. Third, the opaque hoods and side doors effectively screened out or blocked overhead light, which made it very difficult for voters to be able to see the voter verifiable paper audit trail printout and, thereby, ensure the accuracy of their vote. Finally, the hoods were not used at every poll site for the entire day. Our observer present at the start of the day at Linwood Recreation noted that the privacy hoods were not affixed to the machines when the polls opened. (A poll watcher from one of the major political parties pointed this out to the presiding judge and the hoods were quickly attached to the machines.) Another observer, present mid-afternoon at Westlink Church of Christ, also found that none of the voting machines at her site had the hoods affixed to the machines. Similar findings were noted for Calvary Baptist (Derby), Eastminister Presbyterian, Pleasantview Baptist (Derby), West Evangelical Free Church, and Westwood Presbyterian Church.
An apparent, though unconfirmed, additional measure taken by the Election Commissioner's office to provide greater voter privacy was giving administrative guidance to supervisory judges to physically arrange the touch screen machines at their sites so that the voter's back was to a wall while they voted. At sites like Central Christian Church, this machine arrangement (along with spacing the machines well apart from each other) was quite effective. However, at too many other sites, we found that while the machines were arranged to face the wall, voters and poll workers still had to walk behind those in the process of casting their vote to get to open machines, which essentially nullified the goal of voter privacy.
There is also some suggestion that the Election Commissioner's office issued guidance for poll workers to space the touch screen machines as far apart from each other as practicable at each site. In practice, however, our observers found that the efforts of poll site staffers to comply with the Commissioner's guidance were spotty. At too many sites, our observers found that the touch screen machines were within 3 feet of each other in violation of KSA 25-1332(4). In some cases the close proximity of the machines to one another was due to either the small physical size of the designated polling area (e.g., Mount St Mary's Convent or Faith Christian Church) or the suboptimal use of the available space in the polling area (e.g., Independent Living Resource Center) (see section 2.5 above). In many cases, though, the close proximity of the machines was due to the daisy-chain arrangement of the machines, which was done to permit all the machines to share power from a single wall socket. This machine configuration occurred in spite of the fact that each machine had multiple feet of electrical cord (wound up on the back of the each machine) available to space the machines further from each other.
In addition to the machines being too close together, at sites like Grace Presbyterian, Sedgwick County Health Department Administrative Center, and Linwood Recreation Center, the machines were arranged in two rows facing each other. As a result, one row faced the wall while the other faced the center of the polling area creating wide open space behind the voter, thereby allowing anyone to walk up from behind to see what the voter was doing. The absence of privacy due to this orientation was especially evident when observing wheelchair-bound voters using the machines. The screens on their machines were clearly visible to all--you did not even have to be close up to see how these voters were voting. Our observers documented at least two voters who complained to the presiding judge at Grace Presbyterian about the lack of privacy over the course of Election Day.
As during the August primary election, the coalition commends poll staffers at Woodlawn United Methodist Church in Derby, which was the only site we found that most fully addressed the privacy issue. Poll workers at this site spread the voting machines apart from each other and arranged them in a semi-circle to prevent voters from being able to see each other's screens. In addition, opaque vertical panels were placed close behind each voter to prevent their backs from being exposed while they voted.
8. COMMENTS ON THE PRESENCE AND ASSISTANCE OF COUNTY STAFF
As with the August primary, the coalition commends County Manager Bill Buchanan and Public Safety Director Bob Lamkey and their respective staffs for the assistance they provided the Election Commissioner's office throughout the election process. Our observers noted that county workers were very helpful. At all the sites observed, county workers were either present or had been there more than once during the day to assist voters and poll workers and troubleshoot technical problems. Poll workers were also able to call a central number to request supplies or information they were lacking: e.g., extra paper ballots, additional signage, instructions for the machines, etc.
Based on our observations, the coalition would like to request--in cases of general elections involving Federal, statewide, and local contests--that workers trained on the technical aspects of the voting equipment be present at each poll site for the entire duration of Election Day to aid poll workers with troubleshooting machine problems. This measure will help ensure the efficient operation of voting machines to minimize delays experienced by voters.
9. CONCLUDING COMMENTS
The November election was the first general election involving Federal, state, and local contests since the Election Commissioner radically consolidated county poll sites and procured new voting equipment. The higher than anticipated voter turnout throughout the county on Election Day along with the commitment of voters to vote in spite of delays and lines at many poll sites was encouraging. However, the lines and delays we documented above cannot be viewed as acceptable, especially looking ahead toward the 2008 Presidential election. One central lesson of the November election is a need to reverse the radical consolidation of poll sites undertaken by the Election Commissioner in order to minimize voter delays and instances of overwhelmed poll workers. Another key lesson of this election seems to be that the existing stocks of new voting equipment are not sufficient to meet the demand voters place on polling sites in a well-participated election. While voters got through registration relatively quickly, they had to endure long lines and wait times for machines to become available. These lines were aggravated by frequent interruptions as machines required paper rolls to be replaced--a process that most poll staffs had not mastered.
The coalition commends the overall efforts of dedicated poll workers and supplemental county staff in their efforts to efficiently and fairly administer the first general election since the radical consolidation of county poll sites and procurement of new voting equipment. That said, as noted throughout the report above, there is a fundamental need for improved poll worker training and oversight, especially with regard to administration of provisional ballots, technical trouble-shooting of machine problems, and physical arrangement of poll sites and voting equipment to ensure compliance with state voter privacy election laws. In addition, there is a real need for the Election Commissioner's office to reconsider the suitability of certain poll sites based on practical accessibility for people with disabilities (see section 2.2 above) as well as their physical size (see sections 2.5 and 7 above).
The much heralded Optiscan safety-valve proved only as good as the available supply of paper ballots and pens and the ability of the (usually) one poll staffer assigned to provisional/paper ballots to process and accommodate those desiring paper ballots. Even then, our observers found that the reliability and functionality of the Optiscan hardware was suspect, calling into question the accuracy of vote totals as well as furthering delays both for voters during Election Day and for poll workers after the polls closed.
Finally, high voter turnout on Election Day 2006 at county polling sites provided probably the best evidence that the large majority of voters in Sedgwick County still associate the first Tuesday in November with the civic act of voting and a preference to cast their ballots at that time. The very limited efforts by the Election Commissioner's office to educate voters about and promote satellite advance voting as an alternative to Election Day precinct-based voting show little promise in altering voter behavior. In addition to the societal convention of precinct-based voting on Election Day, many voters may not wish to vote in advance on the off-chance that developments in the waning days of the campaign season may alter the way they vote. Whatever the basis, given the enduring and deeply imbued culture of voting on Election Day, rectifying the problems voters experienced on November 7 at a great many metropolitan poll sites will require the Election Commissioner to revisit his poll consolidation plan and take significant measures to spread out and more capably handle the large volume of voters during the upcoming 2008 Presidential election season.
10. RECOMMENDATIONS
Based on direct observation of the November general election and comments made by numerous concerned voters, the Sedgwick County Voter Coalition makes the following formal recommendations for county-wide changes and improvements in the administration of elections to be made prior to the 2008 primary election:
1. The Election Commissioner, in conjunction with the Board of Sedgwick County Commissioners, should form a bi-partisan citizen Advisory Task Force, as previous Election Commissioners have done, that works with the Commissioner in decision-making and advises for or against changes.
2. Restore polling sites throughout the county based on a uniform standard of no more than 3,000 registered voters per poll site AND each polling location within 2 miles of any given voter's home. Publicly-owned facilities should be given preference when designating (or re-designating) polling locations(5).
3. Guarantee voters' right to vote in private in accordance with KSA 25-1332(6). While cosmetic improvements were made to achieve general compliance with KSA 25-2703, KSA 25-1310, KSA 25-1327, the Election Commissioner's office must now study the physical dimensions and features of the polling area at each poll site and prepare mandatory floor plans for the arrangement of each polling site that both eliminates unscreened paths behind machines and ensures sufficient spacing between machines to maximize voter privacy. Based on the results of its site studies, the Election Commissioner's office should work with County elected and appointed officials to obtain whatever additional materials and/or labor is necessary to ensure that the poll areas in each poll site are configured in the manner mandated by the Election Office at the start of Election Day.
4. Ensure that polling areas, voting equipment, and ballots are secure from unauthorized access at all times prior to and during Election Day. Polling areas should be dedicated exclusively to voting. Finally, where polling sites are located in places of worship or religious-affiliated facilities, the Election Commissioner's office should direct poll staff to remove or cover all religious and political literature in the area designated for voting.
5. Procure additional voting machines to achieve a ratio of 300 registered voters to each touch screen machine at all county polling sites(7).
6. Assure that all existing and/or potential polling sites are both formally and practically accessible by all voters. Voters with disabilities must be assured convenient and practical access to the polling area at all polling sites. Voters with disabilities, especially those with visually impairments, must also be universally offered the audio ballot feature on the touch screen machines.
7. Ensure that polling areas within facilities are of sufficient physical size and climate controlled for the comfort of poll workers and voters.
8. Devise and implement an ongoing voter education and marketing strategy of polling sites and the new machines. Working with elected and appointed County officials, the active marketing strategy devised by the Election Commissioner's office should include, but not be limited to, the following:
- Mail large, brightly colored cards to all voters identifying their designated Election Day polling site as well as information on advance voting options--mail-in ballot and satellite voting--prior to start of advance voting
- Produce PSAs as well as paid advertisements
- Conduct demonstrations of new machines throughout the county with a special emphasis on educating individuals who have limited technological savvy
- Produce a virtual interactive or streaming video voting machine tutorial on the county website
- Produce a 5-10 minute video explaining how to use the new machines, as well as providing other relevant voting information, and place it in heavy rotation on Cox Channel 7
- Contact and coordinate with school districts throughout the county to educate high school students of, or near, voting age
- Aid the print media by providing detailed, accurate county precinct maps for publication on or about Election Day
9. Expand satellite advance voting sites in areas of high population density and include high foot-traffic locations such as large commercial facilities. Satellite advance voting should be available, at a minimum, from 8 AM to 8 PM on the Thursday, Friday, and Saturday prior to the election. Advance voting should also be available at all satellite locations on the Monday prior to Election Day as other large Kansas counties have done successfully.
10. Poll worker training should emphasize and ensure the proper, consistent, and uniform administration of provisional ballots by poll workers at all polling sites. In addition to circumstances when a provisional ballot is mandatory, poll worker training should require that the option of completing a provisional ballot be extended to all voters whose names do not appear in the registration book at any given site(8). Additionally, poll workers must be required to inform voters that not all provisional ballots cast will count and may not include contests specific to their home address.
11. In cases of general elections involving Federal, statewide, and local contests, workers trained on the technical aspects of the voting equipment should be present at each poll site for the entire duration of Election Day to aid poll workers with troubleshooting machine problems in order to ensure the efficient operation of voting machines and minimize delays experienced by voters
12. The physical size of poll site-identifying signage at street entrances should be increased from the current size of 26" x 16" (416 sq in) to 48" x 30" (1,440 sq. in.) or 60" x 37" (2,220 sq. in.).
13. The Election Office should increase the percentage of paper ballots available at each polling site at the start of Election Day for each precinct. Further, the Election Office should order the printing of a certain percentage of extra or reserve ballots for each precinct in case poll sites run out of allotted ballots due to provisional balloting as well as voter, poll worker, and/or Optiscan machine error.
The coalition further calls on the Kansas Legislature to:
14. Change Kansas state election law to reduce the length of the work day for poll workers from 15 to 8 hours. Require supervisory judges to work the full Election Day, with the option of two shifts for other poll workers(9). Creating a shorter work day and offering schedule flexibility would likely increase the available pool of poll workers and, more importantly, ensure less fatigued and error-prone workers.
15. Change and/or amend Kansas state election law to require a paper record audit trail for all electronic direct recording devices (including touch screen machines) authorized for use in elections by the Kansas Secretary of State. The law should further codify that paper record audit trails are the true and correct record of votes cast in any given election and should be used as the official legal records for any audit and/or recount(10). In cases of discrepancies, the vote(s) recorded on the voter verifiable paper audit trail are to be seen as the true and correct vote(s) cast by individual voters.
16. To ensure the accuracy and security of direct record electronic voting equipment (including touch screen machines), Kansas election law should be changed to require supervisory judges at all polling sites to perform random periodic audits of direct record electronic voting equipment/touch screen machine tallies against the paper trails of the machines being audited throughout Election Day. Further, all poll workers should be required to make each voter aware of and how to use the VVPAT feature on the touch screen machines.
Endnotes
(1) KSA 25-3005 states that "[a]t all elections authorized poll agents shall be allowed to be present and observe the proceedings at all original, intermediate and final canvasses of elections and at the time and place of casting ballots...The supervising judge of each voting place shall be in charge thereof and may direct authorized poll agents as to their conduct within the voting place, but such directions shall not favor agents of one kind or party over agents of another kind or party, and such directions shall not be contrary to law..."
(2) "When there is any doubt about a voter's eligibility to vote at the precinct where the voter attempts to vote, a provisional ballot should be provided, even if the voter is apparently unqualified. The facts of the matter can be investigated after election day before the county canvass." [KSA 25-409] Source: http://www.kssos.org/forms/elections/election_standards/ChapII-ElecAdmin.pdf at p. 49.
(3) KSA 25-2703 requires that "County election officers shall arrange for voting places...to enable the voters to prepare their ballots, screened from observation....Each booth shall be designed so as to protect the privacy of the voter." KSA 25-1310 states that "[a] kind or make of voting machine approved by the secretary of state:... must provide for voting in absolute secrecy in voting, except as to persons entitled to assistance." KSA 25-1327 requires that "[a]ll voting machines used in any election shall be provided with a screen, hood or booth which shall conceal the voter and his action while voting."
(4) KSA 25-1332 requires that "[e]ach voting machine shall be placed at least three (3) feet from every wall and partition of the voting place and from any other voting machine, and at least four (4) feet from the table of the election board and other furniture." (Emphasis added.)
(5) The coalition feels in light of the 2006 General Election experience that this recommendation represents a necessary step toward a rational and equitable county-wide voting system; especially in light of Federal legislation under consideration that would require that the county have no more than 1,000 registered voters per polling site. Section 303A of HR 4989 requires that "[e]ach State shall establish the number and location of polling places used for each election for Federal office held in the State on the date of the election, and the number of check-in stations, poll workers, ballot marking stations, and voting systems assigned to each such polling place on the date of the election, so that to the greatest extent practicable--(A) the polling places established within a jurisdiction are equally convenient for and proximate to the registered voters in the jurisdiction and provide poll workers, voters, and individuals waiting to vote with shelter from the elements and protection from excessive heat and cold; (B) the maximum waiting time at the polling place is equal for all voters in the State; (C) the time spent by any voter at the polling place from the time of arrival at the check-in station until the time of departure after voting does not exceed 1 hour; and (D) the maximum number of registered voters assigned to any single polling place on the date of the election (as determined one week prior to the date of the election) does not exceed--(i) in the case of a jurisdiction using as its primary voting system a system under which ballots are marked by the voter and are read by a voting device (such as an optical scan or other ballot reading system), 3,000 (except that a greater number shall be permitted at any polling place at which no voter in fact spent more than 1 hour from the time of arrival at the check-in station until the time of departure after voting); or (ii) in the case of a jurisdiction using as its primary voting system a system under which voters use a voting device which directly records the votes (such as a direct recording electronic or touch screen system), 1,000." (Emphasis added.) For full text of bill, go to http://holt.house.gov/pdf/HR_4989.pdf
(6) KSA 25-1332 requires that "[e]ach voting machine shall be placed at least three (3) feet from every wall and partition of the voting place and from any other voting machine, and at least four (4) feet from the table of the election board and other furniture." KSA 25-2703 requires that "County election officers shall arrange for voting places...to enable the voters to prepare their ballots, screened from observation....Each booth shall be designed so as to protect the privacy of the voter." KSA 25-1310 states that "[a] kind or make of voting machine approved by the secretary of state:... must provide for voting in absolute secrecy in voting, except as to persons entitled to assistance." KSA 25-1327 requires that "[a]ll voting machines used in any election shall be provided with a screen, hood or booth which shall conceal the voter and his action while voting."
(7) The coalition feels in light of the 2006 General Election experience that this recommendation represents a necessary step toward a rational and equitable county-wide voting system; especially in light of Federal legislation under consideration that would require that the county have no more than 200 registered voters per touch screen machine per polling site. Section 303A of HR 4989 requires that "the designated minimum number of voting systems used at a polling place within a jurisdiction is--(A) in the case of a jurisdiction using as its primary voting system a system under which ballots are marked by the voter and are read by a voting device (such as an optical scan or other ballot reading system)--(i) one ballot reading system (together with an ample number of ballots and ballot marking supplies and a sufficient number of secure ballot boxes), with at least one readily available spare ballot reading system for the county in the jurisdiction which uses such a system, and (ii) one ballot marking station for every 200 registered voters assigned to the polling place (as determined one week prior to the date of the election); and (B) in the case of a jurisdiction using as its primary voting system a system under which voters use a voting device which directly records the votes (such as a direct recording electronic or touch screen system)--(i) one system for every 200 registered voters assigned to the polling place (as determined one week prior to the date of the election), with at least one readily available spare of such direct recording electronic or touch screen system available for every 5 polling places within the jurisdiction." (Emphasis added.) For full text of bill, go to http://holt.house.gov/pdf/HR_4989.pdf
(8) "When there is any doubt about a voter's eligibility to vote at the precinct where the voter attempts to vote, a provisional ballot should be provided, even if the voter is apparently unqualified. The facts of the matter can be investigated after election day before the county canvass." [KSA 25-409] Source: http://www.kssos.org/forms/elections/election_standards/ChapII-ElecAdmin.pdf at p. 49.
(9) SCVC's recommendation is supported by the findings and recommendations of a comprehensive empirical study of election administration processes completed by the League of Women Voters Kansas in 2003. Analysis of survey responses from election administrators in 101 or 105 Kansas counties led LWV-Kansas to adopt a consensus position that the option of split shifts for poll workers (excluding supervisory judges) on Election Day should be required. See http://www.lwvk.org/files/ea.pdf
(10) This recommendation follows closely on the Federal legislation (HR 550) proposed by Rep. Rush Holt to amend the 2002 Help America Vote Act to require that direct record electronic voting equipment (including touch screen machines) used for Federal elections be equipped with paper record audit trails. The Bill also mandates the legal primacy of paper record audit trails for audits and/or recounts. For the text of the bill, go to http://holt.house.gov/pdf/HR_550.pdf
Comments, suggestions, questions? Contact our
webmaster.
Last revised: July 8, 2008 22:47 PDT.
© Copyright
League of Women Voters Wichita - Metro, Kansas. All rights reserved.
|
![[LWV]](images/lwv76.gif)
