Abstract

Ensuring the accessibility of web content is key to ensuring that users with disabilities have equal access to online information and services. However, as a review of the literature demonstrates, even in the face of legal requirements, accessibility problems persist across the web, including in the online content created and shared by libraries. This article examines the new success criteria in the recently released WCAG 2.1, considers the opportunity they present for libraries to improve the user experience for users with a broad range of disabilities, and proposes steps to improve compliance with WCAG and online accessibility more broadly.

Introduction

Web accessibility is a key to ensuring that web-based content is available and usable for all users. Beyond this, under certain circumstances, it is also a legal requirement in the United States and many other countries in the world. Though this topic has been considered by those creating web content for decades now, web accessibility standards are dynamic, and frequently have failed to evolve quickly enough to stay up to date with growing online capabilities, new use cases, and new technologies. In addition, web accessibility standards have yet to meet the needs of all individuals with all types of disabilities, meaning that additional work is continually needed to ensure that all users have access to web content.

The latest example of this continuing process is the release of Web Content Accessibility Guidelines (WCAG) 2.1 in June of 2018. These new standards further develop and supplement the existing WCAG 2.0 guidelines to improve accessibility for users with certain types of disabilities that have previously been underserved and to take into account changing technology usage patterns. While compliance with WCAG 2.1 will not ensure that web-based content is accessible to all users, it is an important standard against which content can be measured and provides guidance on improving accessibility for individuals with various types of disabilities. However, to ensure greater accessibility of online content, libraries should not simply strive towards compliance with WCAG 2.1, but should establish best practices and workflows that make functional accessibility a priority at all stages of the online content development process. This article considers WCAG 2.1 and presents options that will acknowledge the importance of compliance while also ensuring broader accessibility of online content.

The Development of the Web Content Accessibility Guidelines

As the internet has developed, so too have web accessibility standards. One of the key players in web accessibility has long been the World Wide Web Consortium (W3C). This international group is responsible for developing the standards for the web, and in 1996, several members of the W3C turned their attention towards web accessibility, ultimately founding the Web Accessibility Initiative (WAI) (Dardailler, 2009). Since that time, this group has worked on developing guidelines intended to ensure that content on the web is accessible to users with disabilities. The first set of web content accessibility guidelines was released as WCAG 1.0 on May 5, 1999 (Chisholm, Vanderheiden, & Jacobs, 1999). These standards proved important in their reach, but, as Power, Freire, Petrie, and Swallow note, by all metrics “the impact of WCAG 1.0 on improving the accessibility of the Web remained quite low throughout the period of its use” (2012, p. 433).

Almost ten years later, WCAG 2.0 was released on December 11, 2008, having been designed to “make content accessible to a wider range of people with disabilities, including blindness and low vision, deafness and hearing loss, learning disabilities, cognitive limitations, limited movement, speech disabilities, photosensitivity and combinations of these” (Caldwell, Cooper, Guarino, & Vanderheiden, 2008). WCAG 2.0 aimed to apply “more broadly to different types of Web technologies and to more advanced technologies” and to be “more precisely testable with automated testing and human evaluation” (WAI, 2009). This approach represented a large scale reimagination of the guidelines and it was at this point that the “success criteria” were added to the guidelines to simplify the testing process. The release of WCAG 2.0 also saw the introduction of the three conformance levels of A, AA, and AAA, where Level A represents “the minimum level of conformance” and Level AAA represents conformance with all success criteria (W3C, 2016). This approach offers web content creators the ability to set conformance level goals that meet their own specific needs or the institutional or governmental requirements set forth for their content.

WCAG and the Law

Since the publication of WCAG 1.0, WCAG has grown to be a dominant standard for web accessibility. In the United States, WCAG standards have long played a role in the enforcement of web accessibility. In July of 2010, the U.S. Department of Justice went as far as to publish an Advance Notice of Proposed Rulemaking, setting forth the idea of “adopt[ing] the WCAG 2.0’s ‘Level AA Success Criteria’ as its standard for Web site accessibility for entities covered by titles II and III of the ADA [Americans with Disabilities Act of 1990]” (Nondiscrimination on the Basis of Disability, 2010). This approach was not ultimately adopted, however, and a recent letter from the Department of Justice regarding web accessibility made it clear that “absent the adoption of specific technical requirements for websites through rulemaking, public accommodations have flexibility in how to comply with the ADA’s general requirements of nondiscrimination and effective communication” (Boyd, 2018).

Nonetheless, despite the fact that it is not the official standard under the ADA, WCAG 2.0 has remained important to the enforcement of web accessibility in the United States. Pursuant to a rule released in January of 2017 by the Access Board under Section 508 of the Rehabilitation Act, “the Revised 508 Standards and 255 Guidelines incorporate by reference the Web Content Accessibility Guidelines (WCAG) 2.0” which they describe as “a globally-recognized and technologically-neutral set of accessibility guidelines for Web content” (Information and Communication Technology Standards and Guidelines, 2017). Practically, they go on to explain that this means that “all covered Web and non-Web content and software—including, for example, Web sites, intranets, word processing documents, portable document format documents, and project management software—is required, with a few specific exceptions, to conform to WCAG 2.0’s Level A and Level AA Success Criteria and Conformance Requirements” (Information and Communication Technology Standards and Guidelines, 2017).

While Section 508 of the Rehabilitation Act, which applies specifically to federal agencies, does not have the same reach as the Americans with Disabilities Act, WCAG has also been used as the standard for web accessibility in a number of settlement agreements, resolution agreements, and consent decrees under the ADA and the Rehabilitation Act of 1973 (United States of America v. Teachers Test Prep, 2018; United States of America v. National Museum of Crime and Punishment, 2015; United States of America v. Louisiana Tech University and the Board of Supervisors for the University of Louisiana, 2013), including a few related to online content provided by libraries (Lanzilotti, Cossaboon, and the National Federation of the Blind v. Atlantic Cape Community College, 2015; University of Montana v. the U.S. Department of Justice, 2014; Penn State University v. National Federation of the Blind, 2011).

Internationally, many countries look to WCAG for their own legal standards of web accessibility. The European Union has set forth accessibility standards that apply to all member countries in their Directive on the Accessibility of the Websites and Mobile Applications of Public Sector Bodies (European Parliament and the Council of the European Union, 2016). This directive set the minimum requirement as standard EN 301 549, which incorporated a number of WCAG 2.0 success criteria in setting accessibility standards for use within Europe (European Telecommunications Standards Institute, 2014). Approaches in other countries have been mixed with many, including Australia, Ireland, Israel, Italy, the Netherlands, Switzerland, and the United Kingdom, integrating WCAG into their laws and others, such as France, Germany, and the Republic of Korea, creating their own adaptations of WCAG for use in their country’s laws (Education & Outreach Working Group, 2018). Though China, India, New Zealand, and Taiwan have not integrated WCAG into their laws, they all rely on WCAG for voluntary or mandatory policies (Education & Outreach Working Group, 2018). While it is true that there are other countries that have taken completely different approaches to online accessibility or have not enacted such laws at all, it is clear that WCAG has had a significant impact on web development around the world.

How Accessible is the Web?

Despite the widespread adoption of WCAG 2.0, compliance is far from universal, and even content that complies with the guidelines may not be fully accessible to all users. In the wake of earlier data that found a significant majority of popular websites were not accessible, Lazar, Beere, Greenidge, and Nagappa (2003) studied a mixture of governmental, nonprofit, and commercial websites in the Mid-Atlantic region. They found that only one site was accessible under the standards of WCAG and Section 508, with forty-six sites being marginally inaccessible and three sites being moderately inaccessible under the WCAG standards. They also found thirty-four sites being marginally inaccessible and fifteen sites being moderately inaccessible under the standards of Section 508. A follow up study of the sites found that despite some improvements, such as increased presence of alternative text, “on average, the 50 web sites increased the number of accessibility flaws over this time period. Web sites had actually been made more inaccessible” (Lazar & Greenidge, 2006, p. 289).

The problems identified in these studies are hardly regional ones. A 2004 study of forty-five websites selected from top online product or service providers found that none of the websites evaluated met the requirements for Level AA or Level AAA conformance to the guidelines and only 9 percent met the Section 508 criteria (Loiacono & McCoy). A more recent review of websites from the private sector found that “in total 12 percent of the analyzed Web sites passed the accessibility evaluation, while the vast majority failed” (Leitner, Strauss, and Stummer, 2016, p. 251).

The field of education is also struggling with web accessibility. A recent study of a sample of K-12 schools’ websites found that “nearly two-thirds of schools failed at least one of the measurable WCAG guidelines. 89.3 percent of schools had contrast issues, which typically represent a WCAG failure. 95.5 percent of school home pages had a detectable error of some kind, with the average site having over 24 errors” (Kimmons & Smith, 2019). Similar studies around the world have found ongoing accessibility issues with higher education websites as well (Alahmadi & Drew, 2017; Ahmi & Mohamad, 2015; Ringlaben, Bray & Packard, 2014). Overall, research suggests websites across all sectors have significant accessibility issues that present barriers for users with disabilities as they navigate the web.

Web Accessibility in Libraries

Given the general state of web accessibility, it is perhaps unsurprising that library web content is not consistently accessible or that web accessibility is not always an institutional priority. In fact, a 2006 survey of web developers at libraries randomly selected from the Carnegie classification list found that less than 40 percent of respondents’ institutions required or recommended that web content comply with ADA web accessibility requirements (Connell, 2008). Despite this lack of institutional requirement, many libraries have turned to WCAG 2.0 to set their standards for accessibility. A recent survey of the members of the Association of Research Libraries found that 67 percent of respondents used Level AA conformance to WCAG 2.0 as criteria for their web accessibility testing, and 12 percent of respondents went further to use Level AAA conformance as their criteria (Spina & Cohen, 2018).

Unfortunately, the general lack of focus on accessibility at the institutional level at many libraries appears to be accompanied by a lack of actual accessibility as well. A 2012 study of sixty-four library websites in Ontario, Canada, found that none of the websites were compliant with WCAG 2.0 and, in fact, the sites “on average, contained 14.75 WCAG 2.0 accessibility errors” with a particular predominance of WCAG issues on library homepages (Oud, 2012, p. 14). Other research has found similar results. In a longitudinal study of library and library school websites in North America, Schmetzke and Comeaux found that, as of the Summer of 2006, “despite improvements over the past 4 years, library and (even more so) LIS school Web sites (particularly those in the U.S.) still tend to contain many accessibility barriers, and thus they continue to exclude, unnecessarily, some potential users” (2009, p. 150). When they revisited the same websites in 2010 and 2012, they found that “unfortunately, since 2006 the overall percentage of approved pages has plateaued; it remained at 61 percent both in 2010 and 2012,” though they did note that “the average number of errors per page has decreased steadily and considerably. From 4.8 errors per page in 2002, to 3.6 in 2006, to 2.3 in 2010, it has dropped all the way down to 1.7 in 2012” (Comeaux & Schmetzke, 2013, p. 22). Despite finding some areas of improvement, the authors noted that “the fact that two out of five library web pages are still plagued with at least one Priority 1 error, and that this percentage has not improved in the last six years, is worrisome” (2013, p. 25). A more recent study of public library homepages found that of 122 members of the Urban Library Council, only 7 library’s homepages had no Section 508 compliance issues (Liu, Bielefield, & McKay, 2019).

While library websites are attempting to make at least some improvements to their accessibility, they are far from the only stumbling blocks that individuals with disabilities encounter when it comes to web-based library resources. In their 2010 study of thirty-two databases, Tatomir and Durrance found that “nine sites (or 28 percent of all databases) were rated as moderately accessible, nine sites (or 28 percent of all databases) were rated as marginally accessible, and 14 sites (or 44 percent of all databases) were rated as inaccessible... No database included in the study was rated as largely accessible” (2010, p. 588). In a study of Voluntary Product Accessibility Templates (VPAT) provided by database vendors to document their products’ accessibility, DeLancey found that “vendors reported being fully compliant with 64 percent of the checkpoints they deemed applicable, and partially compliant with a further 24 percent of applicable checkpoints” (2015). Though this imperfect level of accessibility compliance might already be a cause for concern, a further investigation of the databases’ actual accessibility found that “in 37 cases, the information provided on the VPAT was incorrect: the scan found compliance issues where the VPAT stated “n/a” or full compliance. This indicates a VPAT inaccuracy rate of 19.6 percent,” which the author went on to note likely would have been higher if additional testing had been done (DeLancey, 2015).

Multimedia content from libraries and library vendors has proved similarly problematic. A study that evaluated vendor webinars and tutorials based on a WCAG 2.0-based checklist found that none of the webinars were accessible, “none of the screencast or video tutorials tested were completely accessible, and all failed in at least one checklist item” (Oud, 2016, p.11). Though some tutorials were found to be at least somewhat accessible, “59 percent of vendors omitt[ed] usable captions and 50 percent present[ed] tutorials that couldn’t be found on the page or played by people using screen-reading software” (Oud, 2016, p. 13). Libraries have tended to do little better when creating multimedia tutorials in-house. Clossen and Proces reviewed tutorials created by seventy-one libraries at public, R1 institutions and found “serious accessibility problems in the areas of captioning videos and appropriate heading levels” as well as “concerns in many other elements of video and Web tutorial creation” (2017, p. 822–823).

Additions in WCAG 2.1

Released in June of 2018, almost ten years after the publication of WCAG 2.0, WCAG 2.1 represents the latest stage in the development of WCAG. Rather than replacing WCAG 2.0, this document serves to expand the existing guidelines to better serve those who were not adequately considered in the existing guidelines and to take into account new use patterns, particularly related to mobile devices. According to WAI, “WCAG 2.1 was initiated with the goal to improve accessibility guidance for three major groups: users with cognitive or learning disabilities, users with low vision, and users with disabilities on mobile devices” (Kirkpatrick, A., O Connor, J., Campbell, A., & Cooper, 2018). It is important to note that, as is frequently the case with web accessibility fixes, many of the changes required to comply with WCAG 2.1 will increase the flexibility and customizability of the user experience and will therefore also improve the user experience for all users.

Rather than replacing WCAG 2.0, WCAG 2.1 supplements it by reproducing the entirety of WCAG 2.0 and then “adding new success criteria, definitions to support them, guidelines to organize the additions, and a couple additions to the conformance section” (Kirkpatrick et al., 2018). This approach means that “if you want to meet both WCAG 2.0 and WCAG 2.1, you can use the 2.1 resources and you don’t need to bother looking at 2.0” (WAI, 2018a). This backwards compatibility means that existing web content can simply be further refined to meet the new success criteria. In addition to maintaining all existing elements of WCAG 2.0, it is important to note that this new content is based on the same four underlying principles as WCAG 2.0, namely that all content must be perceivable, operable, understandable, and robust.

In total, seventeen new success criteria have been introduced in WCAG 2.1. Of these, five are Level A, seven are Level AA, and five are Level AAA. All of the new success criteria fit into the existing WCAG 2.0 structure, having been added to the end of the relevant guideline for clarity. Many of the new success criteria cannot be checked through automated tools alone, particularly since several include exceptions for essential functionality. This makes it even more essential for content creators to design with WCAG 2.1 in mind rather than simply relying on automated tools to catch and correct mistakes after the fact.

New Level A Success Criteria

Two of the Level A success criteria aim to improve usability for those using speech recognition software. Criterion 2.1.4 covers keyboard shortcuts that could be inadvertently triggered while speaking to a device, such as when an unknown shortcut deletes content if the user says a particular phrase. To address this issue, any shortcut that is “implemented in content using only letter (including upper- and lower-case letters), punctuation, number, or symbol characters” must be able to be turned off, reconfigured to work with other keys, or activated only when the relevant section of the code has focus (WAI, 2018b). Criterion 2.5.3 addresses the fact that the visual labels on user interface components often have text that does not match their underlying, invisible programmatic name, which can then become “an unknown hidden command for speech input users that can be accidentally activated without the user knowing what has happened” (Campbell, Cooper, & Kirkpatrick, 2018). Given that they offer user experience improvements for anyone who uses speech inputs or voice-to-text tools, these success criteria will not only improve accessibility, but will also have a wider impact on user experience.

The remaining three Level A success criteria are focused on making motions, gestures, and movements more forgiving. This ensures that the content is accessible to users whose fine motor skills are limited as well as those who may use assistive tools that limit their ability to interact with their device in these ways. Criterion 2.5.1 ensures that all actions that can be done with a gesture, such as pinching to zoom or swiping to navigate, can also be done without complex motions. It is important to point out that this success criterion is limited in scope and only “applies to author-created gestures, as opposed to gestures defined on the level of operating system or user agent” (Campbell et al., 2018). Criterion 2.5.2 seeks to minimize the impact of accidental or mistaken clicks or touch events by requiring that either nothing happen until the up-event (for example, when the user lifts their finger from the touchscreen) or that the event can be cancelled or reversed with ease and without significant impact. Criterion 2.5.4 focuses on offering alternatives for actions that require users to move their device in specific ways, such as the feature on some mobile devices that allows users to shake their device to undo their last action, or requires the user to move in specific ways to activate features. Offering alternatives can make these features accessible to a wider range of users and in more settings, such as when the device in question is mounted in a stationary position. All three of these success criteria do acknowledge that noncompliance may sometimes be essential and therefore allow for exceptions. However, in many cases, creative design solutions can offer alternatives that will offer more flexible options for users in a wider range of situations and with diverse needs.

New Level AA Success Criteria

The largest number of new success criteria fall within Level AA. The seven new success criteria at this level cover a range of different types of use cases and will be useful to a wide range of users. Criterion 1.3.4 is an excellent example of the range of possible user experiences that can be improved by focusing on these success criteria. It requires that content not require a particular screen orientation (such as landscape) to display properly unless there is an integral reason that it is required. One of the primary use cases driving this success criterion is that of users who operate a mobile device that is mounted to an assistive device, such as a mobile phone mounted to a wheelchair. However, designing to allow this type of flexibility will also benefit those who are using a device temporarily mounted in a stand in their vehicle and those who mount devices in a public space for use by visitors. Similarly, 1.3.5, which deals with input fields, not only benefits those with disabilities that may make it difficult for them to input information due to learning disabilities such as dyslexia or dyscalculia, but also those who are in a hurry or prefer not to memorize this type of information. This success criterion ensures that form fields that request information have the necessary attributes or metadata associated with them so that they can be filled out automatically.

WCAG 2.1 also includes several Level AA success criteria that are designed to improve the user experience for individuals with low vision. Criterion 1.4.10 addresses reflow, which is the adjustment of content to remain visible in the width of the window even after the user zooms in on content. To achieve this, the success criterion requires that content reflow for “vertical scrolling content at a width equivalent to 320 CSS pixels [and] horizontal scrolling content at a height equivalent to 256 CSS pixels” (WAI, 2018b). Criterion 1.4.11 adds a contrast requirement for content that is not text, including both user interface components and graphics. This means that elements in forms, buttons, visual effects, and similar features must have at least a 3:1 contrast ratio compared to the adjoining colors and elements. Criterion 1.4.12 is another guideline focused on allowing users to customize their own experience to improve usability. This success criterion requires that web content that is created using tools that allow for style overrides must allow users to change the line spacing, post-paragraph spacing, letter spacing, and word spacing of textual content. It includes minimum standards for each of these elements, but great flexibility would, of course, meet the needs of a wider range of users. Criterion 1.4.13 ensures that any content that becomes visible on hover or keyboard focus should be able to be dismissed on demand, persist if the user moves the pointer within the newly visible content, and remain visible until the user takes an action unless it is no longer valid. Taken together, these success criteria will offer users with low vision a much-improved user experience as they navigate compliant web content and will also offer features that many other users will likely appreciate as well.

The final Level AA success criterion, 4.1.3, pertains to status messages and requires that they can be programmatically determined by assistive technologies so that they can convey the information to the user even if the user is not able to navigate to the status message on the page. Currently, many status messages remain invisible to assistive technologies such as screen readers, unless the user brings the focus on to them. In cases when the user may not even be aware that messages have appeared on the page, it is difficult or impossible to find the message to focus on it. This success criterion will give users of assistive technology the ability to customize how they receive status messages and to ensure that they do not miss key information.

New Level AAA Success Criteria

Level AAA is the most advanced compliance level for WCAG and is the most difficult to achieve. In fact, W3C’s documentation notes that “it is not recommended that Level AAA conformance be required as a general policy for entire sites because it is not possible to satisfy all Level AAA Success Criteria for some content” (W3C, 2016). Nevertheless, the five new Level AAA success criteria in WCAG 2.1 should not be dismissed out of hand. They offer important features that can be key to building truly inclusive web content and even where they cannot be entirely met, designing with them in mind can help to improve the overall accessibility of web content.

Criterion 1.3.6 asks that elements of the user interface, such as icons and regions, have metadata associated with them so that they can be automatically customized for the needs of the user. This might mean converting icons with text to symbols or allowing users to apply the same symbols that they are familiar with across sites and applications. This supports people with varying cognitive needs, but also opens the door to personalization and customization more generally. Criterion 2.2.6 is also primarily focused on supporting people with varying cognitive needs, but serves the interests of all users. It requires that users be notified of any periods of inactivity that could lead them to lose their progress or unsaved data. The success criterion does allow an exception if the information is preserved for longer than 20 hours, and the documentation actually recommends retaining information for 20 hours as a best practice (Campbell et al., 2018), but more generally, the expectation is that compliant features will include information about any cutoff periods. It is worth noting that the documentation also acknowledges that the decision on how to handle this data might depend on privacy regulations in the relevant jurisdiction and recommends considering that before making a decision on how to approach this element (Campbell et al., 2018).

Criterion 2.3.3 builds on existing guidelines about content that can cause seizures or physical reactions, to specifically contemplate that animations that start upon interaction should be able to be disabled unless they are essential. Criterion 2.5.5 is designed to address the size of target areas, an issue that impacts both users with and without disabilities, particularly on small screens such as those used for mobile devices. The use case example given is, in fact, one that many users can relate to: “The buttons are so small, I hit ‘Cancel’ when going for ‘Submit’. Then I have to start all over again” (WAI, 2018b). This success criterion offers some exceptions, such as when there are alternatives available on the same page, the author does not have control over the target’s size, or there is an essential need for it, but in general, the target’s size should be at least 44 by 44 CSS pixels. Criterion 2.5.6 states that there should be no restrictions placed on the number or type of inputs used and “users should be able to switch input mechanisms at any point should the user determine that certain tasks and interactions are more easily accomplished by using an alternative input mechanism” (Campbell et al, 2018). This ensures that users have the flexibility to use their preferred approach to enter data or interacting with content and also incorporates the flexibility necessary to work with a variety of combinations of assistive devices since each user has their own preferred suite of tools.

The Current Status of WCAG 2.1

In the recommendation document, W3C acknowledges that “significant challenges were encountered in defining additional criteria to address cognitive, language, and learning disabilities, including a short timeline for development as well as challenges in reaching consensus on testability, implementability, and international considerations of proposals” (Kirkpatrick et al., 2018). It is important to note, however, that these new guidelines do not represent a final answer nor will they ensure complete accessibility for all users. Though they do extend beyond the topics covered in WCAG 2.0, it is clear that future work in these and other areas will be needed to continue to improve accessibility. However, this is not a reason to delay implementation of the recommended changes. In fact, the Accessibility Guidelines Working Group has already stated that they recommend “that sites adopt WCAG 2.1 as their new conformance target, even if formal obligations mention WCAG 2.0, to provide improved accessibility and to anticipate future policy changes” (Kirkpatrick et al., 2018). Though WCAG 2.1 may not achieve total accessibility, it is an important step towards more inclusive and usable web content creation practices.

WCAG 2.1 and the Future of Library Web Accessibility

It will take time to know what the long-term impact of WCAG 2.1 will be, but early signs suggest that it will gain traction. Companies that provide tools and services related to accessibility compliance are already beginning to release tools that incorporate advice about complying with WCAG 2.1, including Level Access (Avila, 2018) and WebAIM (Smith, 2018). By September of 2018, EN 301 549, a European standard entitled "Accessibility requirements suitable for public procurement of ICT products and services in Europe," had already been updated and now “directly references WCAG 2.1,” which represents an even more complete integration of the guidelines than had been seen in previous versions of EN 301 549 (Abou-Zahra, 2018). But, even if it is not ultimately widely adopted as a legal standard, it is clear that work in this area will have to continue. As is evident from past web content accessibility guideline versions, new technologies will likely arise that are not clearly addressed by existing guidelines. Evidence of this can already be seen in W3C’s recent launch of a Pronunciation Task Force that will involve those with expertise in a wide range of technologies, including assistive devices and accessibility, to ensure that text to speech technologies are usable by all (Sajka, 2018). In addition to the issues presented by new technologies, WAI’s Accessibility Guidelines Working Group is already at work on another version that “is expected to be a more substantial restructuring of web accessibility guidance” (Kirkpatrick et al., 2018). And, researchers have noted that “while technical compliance with accessibility guidelines would be a first step, websites may be technically compliant without being particularly usable for screen reader users” (Mulliken, 2017, p. 116).

Without a significant investment in and prioritization of web accessibility, libraries will be unable to offer an equitable experience to individuals with disabilities, both as library patrons and as library employees. The evidence demonstrates that current practices are not sufficient to allow most libraries to achieve even minimal compliance with WCAG 2.0. To meet the standards set by WCAG 2.1 and to be positioned to address new accessibility challenges presented by emerging technologies, libraries must make a clear commitment to accessibility. At a time when competition for resources exists at many libraries, this will necessitate devoting time and money to accessibility as well as working to ensure that it remains a priority on an ongoing basis.

One way that libraries can make a public commitment to accessibility is by sharing an accessibility statement online that provides specific standards to which they will adhere. These can take a variety of formats, as long as they set forth a standard that the library will commit to on an ongoing basis. The Hannon Library at Southern Oregon University specifically states that the library’s website will comply with WCAG 2.0, conformance level AA, as well as detailing the schedule on which accessibility checks will happen and even the tool they use (Southern Oregon University, n.d.). Harvard Library not only specifies that their web content follows WCAG, but also uses the page to recruit participants for their “accessibility testing participant pool” (Harvard Library, n.d.). Other libraries are already beginning to cite WCAG 2.1 in their accessibility statements (Ray Howard Library, 2019; Brookline Public Library, n.d.). The Michigan State University Libraries maintain an entire section of their website which details their accessibility efforts, including the accessibility of online content, collections, and services (MSU Libraries, n.d.). Regardless of the format of the accessibility statement, the important aspect is to have documented compliance practices and standards that can provide accountability about web accessibility.

Even libraries that cannot share a public accessibility policy can develop internal policies, procedures, and timelines for web accessibility that ensure that it remains at the forefront rather than falling to the side. These policies can cover not only the standards that web content will adhere to, but also how accessibility will shape collection development decisions. A recent survey of Association of Research Libraries members found that only 33 percent of respondents had included accessibility considerations in their collection development policies (Spina & Cohen, 2018), but this is an area that is ripe for growth. Considering accessibility at the point of purchase or subscription has the potential to have a huge impact on how accessible library collections are, which is particularly important for online vendor content, which generally cannot be edited by the purchasing or subscribing library to improve accessibility. The first step towards considering accessibility as part of the collection development process can be collecting VPATs from all vendors, but, as previously discussed, VPATs are not always accurate (DeLancey, 2015), which has motivated some libraries to move beyond this. While some libraries do internal accessibility testing for databases and other electronic resources, the libraries of the Big Ten Academic Alliance have come together to “fund third-party accessibility evaluations for select vendor e-resources based upon recommendations from the member libraries and the platforms that the majority of the consortium owns or is considering for purchase” and have also developed and publicly shared standardized language about accessibility that libraries can negotiate to insert into vendor agreements (Big Ten Academic Alliance, n.d.).

From a user’s perspective, it is also important to have a way to get in touch with the appropriate person at a library when they are not meeting their obligations. All of the policies discussed above provide contact information so that individuals can reach out to the library if they encounter accessibility issues. Giving users a sense of what to expect from the library and a way to communicate any issues that they have is a key to ensuring that individuals with disabilities will have the information that they need to plan how they will make use of the library, but it is also an important part of ensuring public accountability around web accessibility. These documents are a useful first step towards making accessibility a permanent commitment and giving notice to the library community of this obligation so that they know that this is an area that will be a continued priority even in the face of other competing demands. This ensures that patrons, staff, and other stakeholders all have an avenue for holding the library accountable for meaningful accessibility of all web content.

Policy statements are only the first step towards meaningful progress on accessibility and, in fact, do very little to improve accessibility without a simultaneous commitment of resources. Though a growing number of libraries are hiring individuals with accessibility expertise or tasking specific staff with accessibility compliance, far too often accessibility falls outside the scope of everyone’s job description or is hastily tacked on without budgeting a realistic amount of time to the ongoing accessibility auditing and improvement that is necessary for a library to offer a truly equitable and accessible experience for users with disabilities. Libraries need to move towards a model of clearly assigning accessibility tasks and ensuring that the individual or individuals asked to do this vital work have the necessary training, support, authority, and resources to make meaningful strides towards accessibility. Without this dedication of resources, libraries will never manage to achieve ongoing compliance with basic accessibility standards, much less offer equitable access for individuals with disabilities.

A key piece of allocating sufficient resources to accessibility work is determining the nature of the work to be done. Currently, automated accessibility testing using free or subscription-based tools is a central element of accessibility work. However, these tools can produce incorrect results whether over-reporting errors (Ng, 2017) or missing major problems entirely (Vigo, Brown, & Conway, 2013). These significant limitations taken together with the fact that the existing WCAG standards do not take into account every possible aspect of accessibility clearly demonstrates the need for a more integrated approach to meaningful web accessibility. Rather than relying only on automated tools that attempt to test against WCAG, those tasked with maintaining the accessibility of web content must develop workflows that integrate accessibility testing into overall user testing work. One of these new approaches is to develop expertise in manual testing with available assistive technologies. While no one person will be equally proficient at all possible assistive technologies, the ability to navigate web content with these tools for testing purposes is key to ensuring equitable access for those who use these tools. Moreover, an important part of serving the library’s community is having an understanding of which of these tools is most commonly used by your users, which can often be done by partnering with users who use assistive technologies and the organizations that serve them. This takes accessibility work beyond basic conformance to minimal standards and towards a true focus on accessibility as one aspect of user experience.

User testing that involves individuals who regularly use assistive technologies is a vital complementary piece of this process. A key tenet of user experience work is that “even the most well thought out designs are assumptions until they are tested by real users” (Loranger, 2014). It is vital to involve individuals with disabilities and who use assistive devices in any usability testing processes. Generally speaking, best practices suggest that such individuals be offered the opportunity to use their own technology configured the way that they typically use it to determine how the web content in question works for them under their normal circumstances. Some libraries, such as the Harvard Libraries, maintain an “Accessibility Testing Participant Pool” and use institutional computers but allow participants “to bring their own keyboard, refreshable Braille display, or other input device to use during test sessions” (User Research Center at Harvard Library, n.d.). Either of these approaches can change the dynamics of testing, but it is vital that UX practitioners develop tests that involve individuals with disabilities and become familiar with these best practices so that they can gain an understanding of the actual accessibility issues presented by their web content.

Libraries also need to consider accessibility when considering the adoption of new technologies. Too often libraries make decisions about acquiring and using new technologies and consider accessibility of those technologies only afterwards. This can already be seen in the case of topics that have gained interest in recent years such as immersive technologies (including augmented reality and virtual reality), data visualization, and voice-user interfaces. Libraries cannot wait for groups like WAI to make pronouncements or develop guidelines. Instead, they must work to develop their own accessibility testing policies and procedures and integrate this work into the process of evaluating new technologies. A model of this approach is Temple Libraries’ Digital Scholarship Center, where Jasmine Clark and Jordan Hample are working “to adapt Web Content Accessibility Guidelines (WCAG) to fit VR” (Clark, 2018). In the future, libraries must make accessibility a central component of their decision-making processes when selecting technologies to adopt and tools to invest in and they must be willing to contribute to the ongoing work of developing accessibility standards and practices for these new technologies.

Conclusion

Despite the ongoing work needed in this area to ensure usability for individuals with disabilities, the release of WCAG 2.1 offers a perfect opportunity for libraries to expand their efforts to achieve and even exceed compliance with these standards, which are now designed with a specific aim of making web content accessible to previously underserved users. It is imperative that web developers continue to improve their own accessibility practices, involve individuals with disabilities in their user experience testing processes to ensure that they achieve their accessibility goals, and strive to meet the highest levels of success criteria found in WCAG 2.1 or even exceed them where the success criteria do not adequately address the institution’s technologies or use cases.

Libraries should take this opportunity to examine their own accessibility practices and make a serious and ongoing commitment to improving the accessibility of their existing web content, involving individuals with disabilities in their UX work, and centering accessibility from the beginning in conversations about newly emerging technologies. Only through this continuing effort will it be possible for patrons and employees with disabilities to have equitable access to online library content and will libraries be well-positioned to achieve their goal of offering accessible content and services as new technologies, use cases, guidelines, and standards emerge.

References