1 of 20

TRUST AND ABSENTEE VOTING IN BLIND AND LOW –VISION VOTERS

University of Michigan, School of Information

Lauren Trimble, MA

Robin Brewer, Ph.D. (Advisor)

2 of 20

THE CONTEXT

Blind and low vision people (BLV) often engage in trade-offs to access essential services. Banking or buying groceries might mean disclosing personal information to sighted human intermediaries in scenarios where this information could be weaponized. Voting is a specific instance of this sort of high-risk privacy scenario.

With the advent of COVID-19 during the 2020 presidential election, states expanded access to mail-in ballots.

More than 46% of voters voted absentee.
Blind and low vision voters voted in substantially fewer numbers than voters without disabilities and had the lowest turnout of people with disabilities.
Although legislation has generally improved the accessibility of in-person voting, absentee voting is much less accessible.

2

In July 2017 Cynthia Morales, a blind person in Maryland used a self-checkout kiosk at a Walmart in Owings Mills. She asked for help using from an assistant who then used the cashback option to steal money from her. Blind and low vision people regularly exchange personal information with and without assistance from others in something I’ll be calling high-risk privacy contexts. These high-risk privacy contexts have placed BLV community members in the position of trusting an intermediary or being unable to complete an essential task. Voting is a specific instance of this sort of high-risk privacy scenario.

*touch on COVID 19 section

Why is absentee voting less accessible? Absentee ballots are generally paper, which is considered the “gold standard of election security,” meant to provide a trail that can be examined by hand if there is suspicion of election fraud. In 38% of states, voting absentee requires filling out a paper ballot and returning it by mail or drop box to the voters’ local election official. A voter must receive a ballot, fill it out by hand, sign it, put it in (generally) two envelopes, add postage and return it. This does not take into account any advance preparation done to receive absentee ballots. The Disability and Voting Accessibility report stated that 22.1% percent of voters with vision impairments reported difficulty in voting, as opposed to 2.1% of voters without a disability.

3 of 20

THE STUDY

3

We used the 2020 US presidential election as a case study to investigate the role privacy trade-offs play for blind and low vision (BLV) people, especially in terms of civic participation.

We conducted interviews with 12 BLV voters to understand how they approach the need for human assistance where options for independence are limited.

Using our findings, we proposed some design interventions to improve voting accessibility.

4 of 20

THE STRUCTURE

4

After receiving approval from our university’s Institutional Review Board, we recruited subjects for interview participation through the National Federation of the Blind. The recruitment announcement was sent through an NFB email listserv, and we received over 100 responses. We narrowed the pool to 19 and ultimately interviewed 12.

Interviews were recorded, transcribed and anonymized. Eventually similar comments and perspectives were voiced with increasing regularity, and we felt we had reached saturation.

We analyzed transcript to understand themes and highlight points of interest in a structured open-coding process. Based on the transcripts and mutual discussion, we built a codebook that summarized major observations and contextualized interviewees’ lives. We coded the transcripts.

5 of 20

OUR PARTICIPANTS

5

12 blind and/or low vision registered voters.

Recruited via the National Federation of the Blind.
From a variety of states (Alabama, California, Colorado, Hawaii, Kansas, Michigan, Minnesota, Montana, Oregon, New Jersey, Pennsylvania, Wisconsin).
Average age 53 (ranging from 25 to 70).
5 men, 7 women.
3 identified as low vision, 9 identified as blind.

6 of 20

RESEARCH QUESTIONS

What are perceptions of the absentee voting process by members of the BLV community? What affects these perceptions?

Is absentee voting specifically inaccessible to BLV voters? If so, what are the causes of this inaccessibility?

How can researchers and designers address gaps in voting inaccessibility?

6

7 of 20

KEY FINDING #1

Participants faced a trade-off as they tried to navigate absentee voting in the U.S. Either they forfeited a secret ballot (a right afforded sighted voters and part of the Help America Vote Act of 2002) or they faced disenfranchisement.

7

8 of 12 participants spoke positively about their in-person experiences and felt that voting absentee for the duration of the pandemic had harmed their ability to vote independently.

8 of 20

PRACTICAL ABSENTEE VOTING BARRIERS:

8

Reading the ballot.

Braille and/or large print ballots do not exist in the US.
Most states that offer accessible electronic documents still require the voter to return a paper copy.

Lack of clearly designated signature area.
Ensuring no mistakes were made.
Returning the ballot via mail (putting voter address on return side, affixing a stamp, etc.)
Recognizing the ballot in a stack of mail.
Lack of individual printer access (for ballot or ballot paperwork).

No one that we interviewed was able to absentee vote without sighted assistance of some kind. There were generally two pieces of paperwork to receive, fill out and return: applications for an absentee vote and the absentee ballot itself. Processing either or both of these documents required sighted assistance.

I have been unable to find a single state that offers Braille or large print ballots (in-person or absentee). This like likely for several reasons: braille is incredibly expensive to print and ballots can change before an election. BUT The key reason is that both Braille and large text ballots would also single out the voter in what is meant to be a completely anonymous process.

States like Michigan, Colorado and Hawaii offer an accessible electronic ballot that voters could print out and mail in. However, if a voter is able to use the electronic ballot, they still needed assistance verifying their choices on the physical document, verifying that the printer had worked correctly and signing the ballot. "I think I did have my wife validate that the printer had printed the sheet, but I was pretty sure it had (P3).” In one case the process of requesting an electronic ballot felt so slow that the voter had a family member fill out the paper document she had already received.

Another issue was the lack of a clear space for signatures. Most participants were unable to find the signature spot on their own. Sighted assistance in this capacity was described as frustrating by several participants. “...the most frustration was signing it because I couldn’t see it...and believe it or not, I know that sounds ridiculous. I had to sign that thing five times, he had to print it out five times for me to sign it (P9).”

Mailing difficulties posed further barriers. Addressing the absentee ballot envelope was often a challenge. One participant preferred to request a regular absentee ballot in lieu of printing out an accessible one because the requested ballot came pre-addressed. Recognizing the ballot in a stack of print mail was also a difficulty. “So, remember filling everything out all out online, and then getting materials and print not knowing what they are (P8).”

The lack of individual printer access is also key. Fewer and fewer people have printers and the cost of maintaining a printer imposes a barrier.

9 of 20

BUREAUCRATIC ABSENTEE VOTING BARRIERS:

9

State ID requirements.
Additional paperwork to receive ballot:

Absentee ballot application.
Agent delivery designation form (PA).

“The process to get an online ballot to fill out is not accessible. I didn’t try it, a friend of mine did, and he couldn’t make it work. I didn’t bother.” (P3)

Many participants experienced additional steps to getting an absentee ballot, which also required assistance. These barriers included stringent voter ID requirements and any required additional paperwork to receive an absentee ballot.

Voter ID requirements varied from state to state and often required additional steps and/or inaccessible paperwork. One participant hoped to use an accessible online form but didn’t have the correct form of ID to do this. “I was hoping that I would be able to register online to receive the absentee ballot, unfortunately, because I did not have a Pennsylvania driving non driver’s license, I had a passport card. I had to fill out the form by hand on paper, and I had to make sure that it was sent, and I had to put postage on it and everything (P7).” In Alabama, a voter has to send in a photocopy of a valid ID and have a witness sign their Absentee ballot application. A participant from Alabama found this so onerous that they opted to apply for a ballot in person.

States with the additional step of filling out an absentee ballot application, placed another (often inaccessible) form as a barrier. “The process to get an online ballot to fill out is not accessible. I didn’t try it, a friend of mine did, and he couldn’t make it work. I didn’t bother (P3).” Available PDFs may not be properly tagged.

Some states require an application for each election while other states allow voters to be placed on a permanent list.

Just to give you a specific example I’m going to pick on Pennsylvania, not that they’re alone in creating barriers. Pennsylvania has the option of being added to permanent absentee ballot list. This requires a doctor’s certification that the voter is “permanently disabled, and either physically unable to attend the polls or physically unable both to operate a voting machine (or mark the ballot) and to orally request assistance to do so).” Pennsylvania will also allow someone to deliver your ballot (in lieu of getting it through the mail) but this requires filling out another form and signing it. I did a quick spot check in Acrobat on the absentee voter form and the Agent Delivery form: the former is incorrectly tagged and the later isn’t tagged at all.

10 of 20

IN-PERSON VOTING BARRIERS:

10

Travel logistics.
Lack of staff knowledgeable about accessibility features of voting equipment. The consequences of this could be:

Long wait times for BLV voters.
BLV voters needing sighted assistance.�

“But when it comes to using the machine and setting it up...there seems always to be confusion and how to set it up. And then something happens in the end where the ballot doesn’t print out properly or print out and slightly tears. And then I still need help.” (P1)

Most of our participants preferred in-person but there were distinct caveats. Even those with very positive in-person voting experiences noted that there were often practical barriers to voting.

Travel logistics played a role (either you need a driver or a robust public transportation system).

Untrained poll workers were a commonly cited barrier. Several participants described this barrier as time consuming and inconvenient. At worst, a lack of experienced poll workers could lead to needing direct assistance filling out the ballot anyway.

“The biggest challenge was having the poll workers knowledgeable enough to set up the machine in audio mode, so I could utilize the feature...But when they first started, for the 2004 presidential election, they had some problems. And as a result, the relative who was with me had to assist me in completing a ballot (P4)”

A side effect of untrained poll workers were long voting times for BLV voters. Several voters recalled being instrumental in guiding poll workers through the machine protocol and/or helping them troubleshoot issues, which took time.

Some participants felt frustration at occasionally needing assistance to vote despite the availability of a machine. “But when it comes to using the machine and setting it up...there seems always to be confusion and how to set it up. And then something happens in the end where the ballot doesn’t print out properly or print out and slightly tears. And then I still need help (P1).” However, the ability to vote independently often outweighed possible barriers. Interviewees who preferred in-person voting continued to express this preference after bringing barriers to our attention. Many also voiced the fear that other blind and low vision community members may not vote because of an inaccessible process.

11 of 20

KEY FINDING #2

Finding accessible election information was difficult for participants. This included information on local races and information on the process of voting.

11

Or you can maybe get online somewhere to get all the information, just all the sample ballot and all that kind of thing. That isn’t very accessible.” (P8)

“Because here’s the thing when you do a check voter registration at checkyourvoterregistration.com. The result is that a combo box and or edit Feld will not align with the cursors, causing a screen reader to… Can you do this with me? Causing the screen. to CRASH.” (P10)

12 of 20

KEY FINDING #3

Participants had trusted proxies (family members, friends, etc.) fill out or examine printed ballots. Despite these being trusted people, most voters expressed a preference for voting privately.

12

“And I usually don’t like to share who I am voting for with anybody but myself...And I didn’t want to feel like I had to share my vote with my husband.” (P9)

“It bothers me, because I can’t really, I’m not filling out myself, I can’t really control what’s going on... like it’s my ballot, and I would like to fill it out.” (P8)

13 of 20

ATTITUDES TOWARDS ASSISTANCE IN ABSENTEE VOTING:

13

Resigned with a sense that it was their best or only means to vote.

“There are no other options for me to have viable thoughts, I guess.” (P1)

And I would have preferred a totally secret ballot, like the rest of the fully sighted population, but I understand why the absentee ballot was necessary.” (P4)

Felt as though their voting rights had been taken away.

“I felt violated. Like I lost my rights...For confidential voting.” (P9)

“I no longer had the right to vote independently and privately. And I felt in some ways my rights had been, had been abrogated.” (P7)

Assistance taken as a routine part of life.

“...Me I trust my wife. And so, I’m not terribly worried that she’s going to vote some way that I would not want her to vote.” (P3)

“And they said, how...can this be in the 20th century? That there is no way for you to vote independently and privately? And I said, oh, well, you know, most of life is like that.” (P7)

The attitudes on human assistance ranged significantly. Some participants were not bothered by the assistance. One participant, when asked how he felt about getting assistance from a family member replied “...me I trust my wife. And so, I’m not terribly worried that she’s going to vote some way that I would not want her to vote (P3).”

Another participant recalled voting in the 1970’s and having her accompanying friends express a sense of shock, “And they said, how...can this be in the 20th century? That there is no way for you to vote independently and privately? And I said, oh, well, you know, most of life is like that(P7).”

Another group of participants felt as though their voting rights had been taken away. An interviewee who had lost their vision in the last 5 years described the experience of human assistance with voting as a violation. “I felt violated. Like I lost my rights...for confidential voting (P9).”

One interviewee habitually voted in person with accessible machines and felt that, after voting absentee for the COVID-19 pandemic, “I no longer had the right to vote independently and privately. And I felt in some ways my rights had been, had been abrogated (P7)."

Another group of participants voiced feelings of resignation and a sense that, while they did not relish the thought of assistance, it was their best or only means to vote. “There are no other options for me to have viable thoughts, I guess (P1).”

Another participant expressed concern about the COVID pandemic, feeling that voting absentee was an unpleasant but necessary option, “And I would have preferred a totally secret ballot, like the rest of the fully sighted population, but I understand why the absentee ballot was necessary (P4).”

14 of 20

KEY FINDING #4

Despite boundaries, voters were confident that their votes had been counted. However, 9 out of 12 participants did not check or consider checking whether their ballots had been submitted successfully.

14

“I think I just trusted...that it would get counted.” (P11)

“I mean, I just didn’t do it, just because basically, I just didn’t want to do it...I jumped through all the hoops, and I filled out all the forms and everything...I can’t change anything.” (P8)

We asked participants whether they checked or attempted to check whether their absentee ballots had been submitted successfully. The majority (9 out of 12) did not check or consider checking. One interviewee didn’t confirm because they implicitly trusted their election officials. “I think I just trusted...that it would get counted (P11).” Another participant felt that after the work and effort of submission, knowing their ballot hadn’t been successfully submitted would have added a layer of helplessness. “I mean, I just didn’t do it, just because basically, I just didn’t want to do it...I jumped through all the hoops, and I fled out all the forms and everything...I can’t change anything (P8).”

The lack of submission confirmation pointed to both trust in and exhaustion of the voting process. After facing a complex, inaccessible process some participants did not want to know if their ballot wasn’t successful. Most participants seemed to expect an inaccessible experience and felt resigned. “I know voting is a right. But honestly, sometimes it feels like a privilege at which I need to work harder. And a responsibility that I sometimes cannot meet (P1).”

15 of 20

KEY FINDING #5

Despite boundaries and a fundamentally inaccessible experience, all voters preferred having sighted assistance to not voting.

15

“I know voting is a right. But honestly, sometimes it feels like a privilege at which I need to work harder. And a responsibility that I sometimes cannot meet.” (P1)

16 of 20

DESIGN INTERVENTIONS

Reduce intermediaries:

BLV voters should also be able to submit a ballot electronically without needing a printer. If ballots must be printed, processes should provide tactile means to way find. For example, add two holes on either side of the signature field to indicate where to sign.

Make supplementary election information available.
Allow for immediate feedback:

BLV voters should be confident that sensitive information revealed in an exchange was used in accordance with their wishes.

16

REDUCE INTERMEDIARIES: BLV people should be able to exchange sensitive information with a minimum of assistance, intermediaries, or additional logistical steps that might expose their information in high-risk contexts. For example, give voters the option to sign up for an absentee voter list that does not require them to send an application for every ballot. On paper ballots, add tactile hints around any signature sections (holes on either side of the section for example)

MAKE ELECTION INFORMATION ACCESSIBLE: BLV people should be able to make informed decisions in high-risk privacy contexts using accessible supplementary information. In the context of voting, BLV voters need access to all pertinent information about the voting process in order to correctly navigate it. This includes information about candidates, ballot proposals, sample ballots and the absentee ballot application process.

ALLOW FOR IMMEDIATE FEEDBACK: Short-term processes like retrieving money from an ATM or buying a train ticket provide immediate feedback through actions (e.g., being able to retrieve money, providing access to transportation). Yet, feedback for longer-term processes like voting, where information must be mailed, is not routinely provided. I recommend designers and developers partner with local and state governments to create systems that, at least, confirms when the completed ballot was received. I helped implement a system somewhat like this (working with the Michigan Secretary of State). It was an automated system that sent an email to voters once their absentee ballots had been sent out to them and another when their ballots were received. Surveyed voters gave it overwhelmingly positive feedback (another instance where accessible design is just better for everyone).

17 of 20

IMPROVEMENTS IN PROCESS

17

13 states now offer electronic ballot return options.
Michigan recently passed a ballot initiative that allowed for early in-person voting (up to 9 days before an election) , a single application to request absentee ballots for all future elections (for any reason) and put current, less stringent photo ID laws for in-person voters into the constitution.
States are increasingly offering accessible electronic ballots (Michigan and New Hampshire).

13 total states offer electronic ballot return options (Alaska WHICH JUST OFFERS FAX which I feel medium about, Colorado, Delaware, Hawaii, Indiana, Louisiana, Maine, Massachusetts, Nevada, North Carolina, Rhode Island, Utah, West Virginia). Nine of these states allow voters with disabilities to return a voted absentee ballot by email. Four states allow voters with disabilities to return a voted absentee l ballot through an online portal (Colorado, Massachusetts, North Carolina and West Virgina) HOW accessible these options are is a bit of mystery. Omniballot is one portal solution (two of our participants had used these at the time of the study and found the experience passable-ish with some barriers) but I don’t really have any insight into how accessible it is.

In the Michigan voter ID model, a voter can go to the polls without a photo ID, sign an affidavit saying they didn’t present a photo ID, and then vote normally. You do not need a photo ID to register to vote.

Many states are offering accessible documents but some of them do require hoops. (Ohio and New Hampshire, for example) require applications for this ballot.

18 of 20

THANK YOU

Lauren Trimble

Senior Accessibility Specialist, Thomson Reuters

Lauren.trimble@thomsonreuters.com

19 of 20

REFERENCES

19

Akter, T., Ahmed, T., Kapadia, A., & Swaminathan, S. M. (2020). Privacy considerations of the visually impaired with camera based assistive technologies: Misrepresentation, impropriety, and fairness. The 22nd International ACM SIGACCESS Conference on Computers and Accessibility. https://doi.org/10.1145/3373625.3417003
Azenkot, S., Rector, K., Ladner, R., & Wobbrock, J. (2012). PassChords. Proceedings of the 14th International ACM SIGACCESS Conference on Computers and Accessibility - ASSETS '12. https://doi.org/10.1145/2384916.2384945
Bennett, C. L., Brady, E., & Branham, S. M. (2018). Interdependence as a frame for assistive technology research and design. Proceedings of the 20th International ACM SIGACCESS Conference on Computers and Accessibility. https://doi.org/10.1145/3234695.3236348
Bezyak, J. L., Sabella, S., Hammel, J., McDonald, K., Jones, R. A., & Barton, D. (2019). Community participation and public transportation barriers experienced by people with disabilities. Disability and Rehabilitation, 42(23), 3275–3283. https://doi.org/10.1080/09638288.2019.1590469
Danielsen, C. (2020). Blind Michigan Voters Win Groundbreaking Lawsuit. https://nfb.org/about-us/press-room/blind-michigan-voters-win- groundbreaking-lawsuit
Dill, D. L., Schneier, B., & Simons, B. (2003). Voting and technology: Who gets to count your vote? Communications of the ACM, 46(8), 29–31. https://doi.org/10.1145/859670.859692
Felchner, M. E. (2008). Voting in America. Praeger.
Feiz, S., Billah, S. M., Ashok, V., Shilkrot, R., & Ramakrishnan, I. V. (2019). Towards enabling blind people to independently write on printed forms. Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems. https://doi.org/10.1145/3290605.3300530
Harris, S. P., Owen, R., & De Ruiter, C. (2012, August 22). Civic engagement and people with disabilities: The role of advocacy and technology. JCES. http://jces.ua.edu/civic-engagement-and-people-with-disabilities-the-role-of-advocacy-and-technology/.
Hinkle, B., & Underhill, W. (2020). Voting Outside the Polling Place: Absentee, All-Mail and other Voting at Home Options. https://www.ncsl.org/research/elections-and-campaigns/absentee-and-early-voting.aspx.

20 of 20

REFERENCES CONT.

20

Hinkle, B., & Zoch, M. (2020). Absentee and Mail Voting Policies in Effect for the 2020 Election. https://www.ncsl.org/research/elections- and-campaigns/absentee-and-mail-voting-policies-in-efect-for-the-2020-election.aspx
United States. Congress. House. (2002). Help America Vote Act of 2002: conference report (to accompany H.R. 3295). Vol. 107. U.S. G.P.O.
Lazar, J., Jordan, J. B., & Vanderheiden, G. (2019). Toward unified guidelines for kiosk accessibility. Interactions, 26(4), 74–77. https://doi.org/10.1145/3337779
Meehan, S. (2019, June 30). Blind Marylanders Sue Walmart, saying self-serve checkouts VIOLATE ADA. baltimoresun.com. https://www.baltimoresun.com/maryland/baltimore-county/bs-md-walmart-blind-lawsuit-20181026-story.html
O'Neil, C. (2017). Weapons of math destruction: How big data increases inequality and threatens democracy. Broadway Books.
Schur, L., & Kruse, D. (2021). U.S. election Assistance Commission study on disability and Voting accessibility in the 2020 ELECTIONS: U.S. Election Assistance Commission. U.S. Election Assistance Commission Study on Disability and Voting Accessibility in the 2020 Elections | U.S. Election Assistance Commission. https://www.eac.gov/election-officials/us-election-assistance-commission-study-disability-and-voting-accessibility-2020
Singanamalla, S., Potluri, V., Scott, C., & Medhi-Thies, I. (2019). PocketATM: understanding and improving ATM accessibility in India. Proceedings of the Tenth International Conference on Information and Communication Technologies and Development. https://doi.org/10.1145/3287098.3287106
Spagnuolo, N., & Shanouda, F. (2017). Who counts and who is counted? Conversations around voting, access, and divisions in the disability community. Disability & Society, 32(5), 701–719. https://doi.org/10.1080/09687599.2017.1324765
Stangl, A., Shiroma, K., Xie, B., Fleischmann, K. R., & Gurari, D. (2020). Visual content considered private by people who are blind. The 22nd International ACM SIGACCESS Conference on Computers and Accessibility. https://doi.org/10.1145/3373625.3417014
Stuart , C. (2020). Survey of the performance of american elections. Survey of the Performance of American Elections | MIT Election Lab. https://electionlab.mit.edu/research/projects/survey-performance-american-elections