Retrofitting Accessibility: The Legal Inequality of After-the-Fact Online Access

Retrofitting Accessibility: The Legal Inequality of After-the-Fact Online Access for Persons with Disabilities in the United States

by Brian Wentz, Paul T. Jaeger, and Jonathan Lazar

From the Editor: The following article is reprinted from the online journal First Monday, Volume 16, Number 11--7 November, 2011. The authors are all distinguished researchers. Dr. Brian Wentz is an assistant professor in the Department of Computer Science and Information Technologies at Frostburg State University. His research interests include human computer interaction, accessibility, user-centered design, social computing, policy implications of accessibility and usability, and making business practices more accessible. His doctorate is in information technology from Towson University. Dr. Paul Jaeger is assistant professor and co-director of the Information Policy and Access Center in the College of Information Studies at the University of Maryland. His research focuses on the ways in which law and public policy shape information behavior. He is the author of more than one hundred journal articles and book chapters, along with seven books. His research has been funded by the Institute of Museum and Library Services, National Science Foundation, American Library Association, Smithsonian Institution, and Bill and Melinda Gates Foundation. Dr. Jaeger is the Associate Editor of Library Quarterly and co-editor of the Information Policy Book Series from MIT Press. Dr. Jonathan Lazar is a professor of computer and information sciences, director of the Undergraduate Program in Information Systems, and director of the Universal Usability Laboratory at Towson University. He teaches and does research in human-computer interaction—specifically, web usability, web accessibility, user-centered design methods, assistive technology, and public policy in human-computer interaction. He has published five books. He serves on the editorial boards of Interacting with Computers, Universal Access in the Information Society, and ACM Interactions Magazine and serves on the executive board of the Friends of the Maryland Library for the Blind and Physically Handicapped. Dr. Lazar was a winner of the 2011 University System of Maryland Regents Award for Public Service, and a winner of the 2010 Dr. Jacob Bolotin Award from the National Federation of the Blind. For general readability we have removed the bibliographic citations and the notes and references. You can read the complete text at <http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/3666/3077> by using the archive search box.

A warning to readers seems in order: this article is long and in parts somewhat technical. Understanding that articles must first and foremost be interesting, that technology issues seem hardly relevant to some, and that understanding the law can be almost overwhelming to those not trained in the profession, we nevertheless print this article and urge that all advocates who are serious about confronting the civil rights challenges of this decade and perhaps this century read carefully what is here. Take breaks when you must; reread things that don't at first make sense, but please consider the warning here that, if we fail to address the technological challenges facing us and the underlying attitudinal problems that allow them to persist, we risk becoming a permanent underclass. The best insurance against this catastrophe is the National Federation of the Blind, you and me. Here is the article:

Introduction

For people with disabilities equal access to information, communication, and related technologies defines many of the civil rights concerns in the age of the World Wide Web. With Internet and web-based content now central to education, employment, entertainment, socialization, and civic engagement, unequal access to the content and technologies necessary for social participation results in very significant virtual segregation. Despite the fact that the Obama administration began the federal government’s first concerted effort to enforce online accessibility in 2010, those with physical, sensory, and cognitive disabilities face significant differences in levels of access to and on the Internet due to the widespread inaccessibility of its content and of the software and hardware necessary for online access.

Issues of electronic accessibility for people with disabilities have meaning to a large part of the population. In the United States 54.4 million people have a disability (18.7 percent of the overall population in 2005), while the number of persons with disabilities worldwide is more than 550 million. Disability does increase with age—13 percent of people age twenty-one to sixty-four have a disability, but 53 percent of those over seventy-five have a disability. The number of Americans age fifty-five or older is increasing rapidly as a percentage of the total population; as a result the number of persons with disabilities will grow significantly in the next few years as the baby boom generation ages.

Technologies that are inherently designed to be inclusive of all users regardless of ability are known as accessible technologies or universally usable technologies. For a technology to be accessible, it needs to be usable in an equal manner by all users regardless of specific senses or abilities. It should also be compatible with assistive technologies, such as narrators, scanners, enlargement, voice-activated technologies, refreshable Braille, and other devices. The use of different assistive technology varies widely between individuals based not only on type and scope of impairment, but also a range of factors related to personality, environment, support, and nature of the technology itself. However, even when assistive technology is an available and functional solution, it is an extra cost faced by people with disabilities that other users do not face.

Nothing about technology makes it inherently accessible or inaccessible. Most of today’s technologies are digital, meaning that they are made up of zeros and ones, and there is nothing inherently visual or auditory about zeros and ones. Digital information is not inherently accessible or inaccessible, but the choices made by those developing and implementing technology determine whether a technology will ultimately be accessible or inaccessible. This is particularly true online, given the rapid pace of technological change and introduction of new web-enabled technologies, since online technologies are often obsolete before they are made accessible.

Online accessibility has been most commonly studied in websites since they have been around far longer than social media or mobile devices. However, the growing use of these newer technologies by governments, schools, and corporations makes the focus on born-accessible technologies more important, as the online environment becomes more communication-based and more central to everyday information behavior. While most of the research has been focused thus far on websites, this paper examines the range of technologies and tools related to online environment because policy changes will need to address all of these technologies to create an inclusive and accessible online environment.

Already at a Disadvantage

Over the brief history of the web, access to and use of the Internet by people with disabilities has been consistently approximately half that of the rest of the population. In 2011 54 percent of adults with disabilities used the Internet, while 81 percent of other adults did. People with disabilities who do regularly use the Internet also lag behind in quality of access, with 41 percent of adults with disabilities living in homes with broadband access, in contrast to 69 percent of the rest of the population. A 2010 study similarly found that broadband adoption by persons with disabilities was two-thirds the national average and that people with disabilities who have broadband engage in a much smaller range of online activities as a result of accessibility issues. People with disabilities who live in nonmetropolitan areas have the lowest Internet use of any population in the United States.

Given the importance of the Internet in education and employment, such differences in access have serious ramifications for the opportunities available to people with disabilities. A 2011 study found that 46 percent of adults with disabilities live in a household with U.S. $30,000 or less in annual income, in contrast to 26 percent of the rest of the population. Similarly, 61 percent of adults with a disability had only high school education or less, while only 40 percent of other adults did. Those with disabilities also face unemployment at levels three times higher than the rest of the population and suffer similar gaps in educational attainment.

In 2005 81 percent of individuals without disabilities were employed full or part-time, compared to only 32 percent of those with disabilities. A 2008 study found that the employment rate of working-age people with disabilities was 39.5 percent, and the full-time employment rate was 25.4 percent. For some types of disability the gaps in employment are even higher—for people considered to have a severe disability by the Census Bureau, 69.3 percent are unemployed and 27.1 percent live in poverty, more than three times the national average. Yet 75 percent of people with disabilities who are not employed want to work. Only 30 percent of high school graduates with disabilities enroll in college, compared with 40 percent of the general population; one year after high school graduation, only 10 percent of students with disabilities are still enrolled in two-year colleges, while only 5 percent are still enrolled in four-year colleges. While accessibility is well served by designers who are motivated to include or focus on accessibility in their product designs (whether physical or electronic), this article primarily focuses on the impact that laws and regulations can have on encouraging or even discouraging accessible design.

An accessible web offers potential opportunities to shift these disparities significantly because the promise of telecommuting and telework, online courses, and social networking includes the ability to get an education, be employed, and engage in social activities, regardless of the barriers of the physical world. In light of these considerable potential benefits, the federal government has passed a series of laws and regulations—most notably the Americans with Disabilities Act (ADA), the Telecommunications Act of 1996, Sections 504 and 508 of the Rehabilitation Act, the E-government Act, and the Individuals with Disabilities Education Act (IDEA)—that promote online accessibility for people with disabilities.

However, since the advent of the World Wide Web in the mid-1990s, equal access to information and communication technologies has grown increasingly untenable for people with disabilities because the introduction and evolution of technologies has accelerated to the point that most new technologies introduced are obsolete before they become accessible. Ironically, the reasons that most technologies are born inaccessible to people with disabilities can be found in the very legislation designed to promote access. Key components of disability rights laws—levels of proof of disability, means of pursuing claims of inaccessibility, options for parallel versions, and exceptions for undue burden—all perversely perpetuate barriers to online accessibility, making the web a separate but unequal environment for people with disabilities.

Due to these components of the laws, existing disability laws empower a culture of retrofitting rather than early planning or even long-range planning, much in the way long-term zoning is done with building and land development in order to reach a goal that may seem to be in the distant future. As a result the accepted approach seems to be to satisfy the minimum requirements only after attention to the inaccessibility is noted, usually in the context of active discrimination against people with disabilities. If the Internet is to fulfill its promise of providing new levels of inclusion for people with disabilities, the barriers to equal access need to be eradicated. Otherwise the opportunities for social inclusion that people with disabilities have fought so hard to win over the past half century will recede as participation in education, employment, government, and society as a whole become less possible due to technological barriers. The failure to address issues of accessibility for those with physical, sensory, and cognitive disabilities ultimately threatens to segregate people with disabilities as the permanent second-class citizens of the information age.

This article will examine the components of disability rights laws that perpetuate a separate but unequal environment online. Against the backdrop of the current enforcement activities of the Obama administration, the paper will examine the underlying access assumptions of disability rights law in contrast with the access assumptions of other forms of civil rights laws. The paper will then examine the legal mechanisms that result in after-the-fact accessibility, retrofitting accessibility—not always successfully—into products that would have been better for all users had they been accessible from the outset. Finally, the article will offer a set of policy recommendations that should be considered in light of the Obama administration’s engagement with online accessibility.

Current Efforts to Improve Accessibility

In many ways 2010 stands as a watershed year in the struggle for equal online access for people with disabilities. The Departments of Education and Justice took the unique step of issuing a joint statement to educational institutions to say that the use of inaccessible e-book readers and similar devices by elementary, secondary, and post-secondary institutions (without providing any accommodations) was a violation of the ADA and Section 504. This statement followed a series of settlement agreements in which universities agreed to desist from the use of inaccessible e-book readers for course readings, such as the agreement that ended the lawsuit filed by the National Federation of the Blind (NFB) and the American Council of the Blind (ACB) against Arizona State University (ASU), which had been using the inaccessible Kindle e-book reader in courses.

In March 2010 the U.S. Access Board released a draft version of the new Section 508 guidelines and made them available for public comment. The intent was to harmonize, update, clarify, and refocus the requirements by the functionality of technologies, instead of by product type, to account for the range of features in many products. The guidelines will also expand the ADA requirements to include self-service machines used for retail transactions and will incorporate the principles of the WCAG 2.0. If these guidelines are implemented as intended, the principles of accessibility will be strengthened considerably under the law, though they continue to focus primarily on sensory and mobility impairments.

In the fall of 2010 President Obama signed into law the Twenty-First Century Communications and Video Accessibility Act of 2010. Among other qualities the law requires:

Televised emergency information to be accessible to individuals who are blind or have low vision.
Accessible user controls for televisions and set-top boxes, and easy access to closed captioning and video description.
Captioned television programs to be captioned when delivered over the Internet.
Accessible advanced communications equipment and services, such as text messaging and email.
Access to Internet services that are built into mobile telephone devices, like smartphones, if achievable.
Devices of any size to be capable of displaying closed captioning, delivering available video description, and making emergency information accessible.

In addition, the law:
Authorizes the FCC to require seven hours per week of video description on the top four network channels and top five cable channels nationwide.
Allocates up to U.S. $10 million per year for equipment used by individuals who are deaf-blind.

At the same time the U.S. Department of Justice (DOJ) announced that it would be engaging in efforts to promote Internet access for persons with disabilities, including the very worthy goals of increasing accessibility of government and educational websites. The DOJ also began pursuing a series of revisions to the ADA to account for changes in technology and society since the passage of the law. These updates include accessibility of movie theaters, design of furniture, access to 911, and website accessibility. The last is the most significant change, because it clarifies ADA coverage including the websites of all entities covered by the ADA: local governments, state governments, and places of public accommodation. As far back as the 1990s, the DOJ had stated that the ADA applied to websites of state and local government and public accommodations. However, no technical standards or guidance were provided within the ADA regulations. These strengthened regulations will be of value only if they are actually complied with, monitored, and enforced.

All of these efforts to effect more accessible content, software, and hardware are important steps in advancing equal access online. And such advances are needed. In studies of the top 100 companies and the top 100 nonprofit groups in the United States, only 6 percent of corporations and only 10 percent of nonprofit organizations were found to have accessible home pages. In 2009 fewer than 10 percent of leading corporate and e-commerce sites were found to be free of accessibility barriers, while similarly low percentages of government sites are accessible for people with disabilities. However, these efforts do not challenge the underlying issues of the construction of disability rights laws that perpetuate unequal access online. As a result, even if all of these efforts are successful, they are not likely to correct the larger problems that promote and enable an unequal online environment. Schmetzke has compiled a noncomprehensive listing of other similar studies in the U.S. and throughout the world. Table 1 includes a sample of studies in the U.S. concerning the lack of website accessibility that are referenced throughout this paper.

Table 1: Sample of Studies on Website Accessibility Referenced in this Paper.

Study	Scope	Date	Percentage of inaccessible sites found
Ellison	50 Government website home pages	2004	78%
Jaeger	Government websites	2006	100%
Loiacono, et al.	Government, nonprofit organizations, and corporate websites	2009	77%, 89%, 94% respectively
Olalere and Lazar	Government websites	2011	92%
Rubaii–Barrett and Wise	State government websites	2008	63%

The Accessibility Dichotomy

People with disabilities are treated with greater differences under the law than any other minority group in the United States. These differences are the primary driver of the levels of inaccessibility that individuals with disabilities face online, creating challenges in the law to enforcement of equality that no other minority populations must attempt to navigate. As a result of these differences, civil rights laws have been considerably more effective for other disadvantaged populations in creating enforceable rights and, perhaps more significantly, in cultivating belief in the validity and importance of those rights. Conversely, the legal approach taken thus far toward accessibility for people with disabilities reinforces the notion that equality of people with disabilities is less important than equality for other traditionally disadvantaged populations.

For much of the law establishing the rights of various minority populations, the rejection of the separate-but-equal mentality, articulated in the U.S. Supreme Court’s 1954 decision in Brown v. Board of Education, serves as the core foundation of protection for being included. As a result, under most civil rights laws, the expressed goal is the creation of environments, services, and technologies that equally include of all populations. All civil rights laws in the United States—with the exception of disability rights laws—are based on an anti-differentiation approach, meaning that anyone has protections under the law if he or she is being discriminated against. Thus laws prohibiting discrimination based on race, ethnicity, gender, national origin, sexual orientation, or age protect all citizens from any of these forms of discrimination and give all citizens standing to identify and challenge such discrimination, regardless of whether they are the direct victims of the discrimination.

In contrast, disability rights laws in the United States have been built on an anti-subordination approach, meaning that rights are available only if one is a member of the legally defined class of people protected and can prove that one is a member of the class. Thus people with disabilities are the only group that has active responsibility under the law to enforce their own rights and petition for equality when it is not already available. This difference means that disability rights laws are much harder to enforce, since people with disabilities must first prove that they have standing under the law, something no other population must do under civil rights laws.

In most cases the main recourse for people with disabilities is to file a discrimination claim in the court system. For discrimination claims against private entities, the cases enter the civil court system, while discrimination claims against federal government agencies enter the administrative law courts. Claims against state and local governments can go straight to civil court or can be filed with an enforcement agency for investigation. This approach places a curious burden on people with disabilities to be able to afford legal counsel and have the time to file a case to assert their rights. People with disabilities are also the only population required to prove their qualifications under the law to file a disability discrimination claim. Many cases filed under disability rights law are actually dismissed because the courts have taken a very narrow perspective as to who should be covered by the law, frequently not allowing cases to be heard because the court decides that the plaintiff is not “sufficiently disabled” or something similar. Although “discrimination law has a weak history of enforcement” due to budgetary, bureaucratic, and administrative constraints, these issues are particularly prominent in disability rights. Another barrier to enforcement is the fact that people with disabilities cannot receive damages against private entities under the ADA.

Given these requirements, courts have not been generally receptive to disability discrimination cases. Cases filed under the Americans with Disabilities Act are lost up to 96 percent of the time. The only litigants less successful in court than people claiming disability discrimination are prisoner plaintiffs, who rarely have representation by counsel. Perhaps more significantly, the U.S. Supreme Court has spent the last two decades significantly limiting the disability rights laws passed by the U.S. Congress, and many state courts and state legislatures have followed the lead of the Supreme Court. Many states have changed state disability rights laws to mirror the limitations imposed by the Supreme Court on federal laws, while other states have strengthened state laws to counter the rulings of the Supreme Court. Most distressingly, the majority opinion in a 2001 Supreme Court decision on disability dismisses the history of discrimination against people with disabilities as exaggerated and inconsequential.

This anti-subordination approach to disability rights also results in many laws allowing or directly encouraging the creation of separate versions in many contexts. Under various disability rights laws, buildings, public services, educational settings, and technologies—along with many other examples—can occur in separate versions for people with disabilities and for everyone else. The result of this approach is the creation of two separate versions of ways to do the same thing, such as stairs and ramps or the creation of an after-the-fact accessible version available perhaps years after the inaccessible version. This approach serves as an acceptance of separate but equal, a concept Brown seemingly rejected for all populations. While this situation raises numerous concerns for disability rights, the example of technology provides an example in which the significant problems of this approach are made extremely clear.

Over the years higher education has provided many examples of users with disabilities being forced to lobby for changes that are needed. In 1996 a complaint was filed at the Office for Civil Rights at the U.S. Department of Education against California State University (Los Angeles) for the failure to provide blind and low-vision students access to library, publications, computer labs, and accessible testing. This was, of course, a result of these students bringing these issues to the surface and in this case even then not immediately receiving proper satisfaction. Another example, in 1999, is the class-action lawsuit brought against the University of California Berkeley and Davis campuses concerning the lack of accommodations for students with hearing impairments. More recently Pennsylvania State University made headlines when the National Federation of the Blind filed a complaint with the U.S. Department of Education alleging that blind students and faculty were unable to access resources, including the library website, many departmental websites, the course management system, “smart” podiums in the classrooms, and student banking through PNC Bank. Once again these known accessibility problems would have continued without correction if individuals with disabilities were not proactively advocating for their own rights. Problems like those seen at Penn State are likely to be common today at many other colleges and universities as well.

E-commerce websites and other providers of commercial technology products often take the same approach to accessibility. An example from the late 1990s is that of AOL and the lawsuit brought by blind individuals concerning problems with the accessibility of its products. In 2004 Priceline.com and Ramada.com made a settlement with the State of New York regarding the lack of accessibility of their websites. The settlement between the National Federation of the Blind and Target is another well-known example. In this case the commercial website Target.com was noted to be inaccessible to individuals with visual disabilities.

The transportation industry has been a notorious example of this problem. As early as 2002 Southwest Airlines was involved in a lawsuit over the accessibility of their website in a case which was eventually dismissed. Also in 2002 the Metropolitan Atlanta Rapid Transit Authority was sued because its website was inaccessible to individuals with visual disabilities. This case, rather than being dismissed like the Southwest Airlines case, became early proof that the ADA applies to website access for individuals with disabilities. More recently (in 2010) United Airlines was sued over the inaccessibility of its airport kiosks (Standen, 2010), and this was again because of advocacy by individuals with disabilities. Even after this series of cases, however, many airlines do not comply with regulations related to disability, and some apparently remain unaware of the relevant regulations.

The notion of parallel versions and after-the-fact fixes can be found in the first major disability rights law. Having a tremendous positive impact on the accessibility of construction in the United States, the Architectural Barriers Act in 1968 set standards for accessibility in future construction and retrofitting requirements for existing buildings. However, it also established the basic approach in which changes to existing structures are generally made only after complaints have been filed.

The disability rights laws that have followed over time, including Section 504, IDEA, ADA, the Telecommunications Act, and Section 508, all take a similar approach to accessibility, allowing for parallel versions and after-the-fact accessibility in different contexts, from classroom education to mobile phones. IDEA promotes integration of students with disabilities but allows for segregation of many students with disabilities into separate classes or schools. Some interpretations of Section 508 seem to allow government agencies to have a separate, accessible website for people with disabilities and a primary website for everybody else, but Title II of the ADA specifically mandates integration of people with disabilities. Though there are exceptions—for example, the Library Services Construction Act (LSCA) tied receipt of funds to making library buildings and services more accessible—the overall canon of disability rights law is built on the notion that making things accessible is a process of parallelism. Even if this approach successfully results in parallel versions for persons with disabilities and for everyone else, the goal is one of separate but equal. The oblique but inherent truth of disability rights law is that it allows, even endorses, a separate but equal approach to accessibility.

Separate, However, Is Still Not Equal

Separate, however, is still not equal. In practice this parallelism does not always occur, much less occur successfully. In the case of information and communication technologies, the speed of change in the creation of new technologies and in the change between versions of existing technologies demonstrates the inherent problem in a parallelism approach. “The introduction of new technologies sees people with disabilities overlooked, omitted, neglected, and not considered.” The distance between writing and writing systems for people with visual impairments can be measured in millennia. The gap between typeset printed books and Braille and Talking Books was nearly half a millennium. More recent developments, like TTD/TTY services and closed captioning, came decades after mass production of the telephone and television. Even when separate websites are created, the version for people with disabilities is frequently far inferior in content and functionality to the standard version of the site.

Section 508 and the Americans with Disabilities Act

The federal government itself is a prime example of these delays. In 1998, then President Clinton signed into law Section 508 of the Rehabilitation Act, which mandates accessible online materials and the use of accessible technologies by federal government agencies. The law was to have been fully implemented by 2001. A wide range of studies has shown low levels of compliance with the law, often with single-digit percentages of the websites studied compliant. Both the U.S. Federal Communications Commission’s “National Broadband Plan” and the National Council on Disability’s “National Disability Policy: A Progress Report” noted the continuing widespread failure of the federal government to comply with Section 508 requirements. The report equates noncompliance with Section 508 and the common inaccessibility of commercial websites. As noted above, in the summer of 2010, DOJ announced it planned to query the government agencies as to the accessibility of their sites. They did not plan actually to evaluate the accessibility of the sites, but instead to ask the agencies to tell them how accessible the sites are.

While this is a bit surprising, this is the same method used to evaluate Section 508 compliance when compliance activities were actually being done in 2001. The regulations require the DOJ to evaluate compliance with Section 508 every two years, starting in 2001. While the 2001 compliance survey was completed and data posted on the Web, the DOJ indicated that a compliance report was done in 2003, but never posted that data. No further compliance activities have been done since then. The summer 2010 memo on 508 compliance noted that the compliance activities required by the law will commence again in September 2010, but the survey was actually sent out in February 2011. These efforts are the first attention given to Section 508 compliance since 2003, many years after agencies were supposed to have been in compliance.

In the age of the Internet, the average time between the introduction of a new information technology and the availability of a version that is accessible to people with disabilities is three years. This may sound like a tremendous improvement over previous technologies, but technological change happens so rapidly that the gap might as well be measured in millennia. The accessible versions of information and communication technologies are often made available long after a technology is current or are never made at all, since the cycle of change is faster than the cycle of creation of accessible versions. As a result, retrofitting accessibility into technologies frequently never occurs. When it does happen, it often does not work as well as it would have if the technology had been born accessible.

Some products are intentionally made inaccessible when introduced as a commercial decision. For example, though portable e-book readers can easily be built with the capability to verbalize the text of the e-books, often they are not. Amazon’s Kindle reader has the capability, but, when it was launched, the speech function was blocked in most of the titles available for the reader, and the navigation options were limited for users with visual and mobility impairments. And, as newer versions of the Kindle were introduced, they included text-to-speech capabilities for book content, but the menus and controls themselves were not accessible. A number of major universities planned to start using Kindle for textbooks, and that number continues to grow without consideration of the implications for students and faculty with visual and mobility impairments. Threats of lawsuit from a number of disability rights groups were required to change the attitudes of Amazon and the universities, though people with disabilities were mocked on technology blogs and websites—and on Amazon’s own website—for fighting for equal access to the Kindle. Educational institutions were even more enthusiastic to adopt the Blackboard online course software, which was predominantly inaccessible when it was launched and only became disability-friendly ten years later. The e-reader barriers stubbornly persist. Even after the letter was sent out by the U.S. Departments of Education and Justice in 2010, the newest version of the Kindle, the Kindle Fire, is not accessible for people with visual impairments.

Two Separate Applications, or One Application at Two Different Times, Will Be Inherently Unequal

The launch of an inaccessible product and then a later accessible one, even if it ultimately results in a universally usable product, is still a glaring form of inequality, since people with disabilities are excluded during the time gap between inaccessibility and accessibility. Another approach is the creation of parallel versions—one accessible and the other inaccessible. This approach can be seen in the creation of two versions of the same website or in the creation of a fully functional product and a light version of the product with fewer capacities for people with disabilities to use.

Yahoo Mail Classic, Outlook Web Access 2007 Light, and the “basic HTML” version of Gmail are good examples of this. Yahoo recommends that users of assistive technology use the Yahoo Mail Classic version of its web-based email interface rather than the current version of the Yahoo Mail interface, even though the two web-based interfaces do not have all of the same features. Due to accessibility problems with its standard web-based interface for accessing corporate email, Microsoft recommends that users who are blind, have low vision, or require screen magnification use a different interface called Outlook Web Access “light.” This version of the web-based email interface is not consistent with the features provided by the standard version of Outlook Web Access. Google recommends its “basic HTML” version of Gmail as the best version that has compatibility for screen readers, yet it is missing features, including spell check and the ability to manage contacts.

Another common example is maintaining a separate, text-only version of a website. Many organizations select this as a way to avoid maintaining a single, accessible interface, and the current version of Section 508, as well as the previous version of the Web Content Accessibility Guidelines from the W3C (WCAG 1.0), seem to permit this. However, Section 508 notes that separate, text-only pages are an alternative only as a last resort and must be updated consistently with the primary webpage. Also the newest version of WCAG (2.0) no longer provides an allowance for separate, text-only interfaces. Some examples of separate, text-only websites include the MTA, New York City Transit, some government websites, and some university websites. Some universities and other organizations rely on products like Usablenet Assistive from the company UsableNet, which attempts to convert current website content dynamically into a text-only format.

Another pathway in separate but unequal can be found in what might be called an accessible-upon-request approach. Once again, higher education provides painfully apt examples. “Colleges that wouldn’t dare put up a new building without wheelchair access now routinely roll out digital services that, for blind people, are the Internet equivalent of impassible stairs.” For example, the new Facebook-based virtual student union at Arizona State University that was made to increase a sense of community among students is inaccessible for students using screen readers, but the developers would apparently be willing to make it accessible if there are sufficient complaints. As mentioned previously, in late 2010 Pennsylvania State University had a complaint filed against it with the U.S. Department of Education, stating that the University engaged in “pervasive and ongoing discrimination” against students and faculty with disabilities due to inaccessible websites, library catalogue, departmental websites, and course management software that were often adopted with the notion that they might be made accessible at some undetermined point in the future.

But other universities have had success in addressing technology accessibility. For instance, the California State University system had an Accessible Technology Initiative, in which it refused to adopt software or hardware products systemwide until companies such as Apple, Google, and Blackboard made their products accessible for students with disabilities. Because of the large procurements involved (the CSU system has an estimated 430,000 students), the system was able to pressure the technology companies to make their products more accessible. Unfortunately, the status of the California Accessible Technology Initiative is unclear now due to state budget cuts. Due to all of the attention being paid to technology accessibility on campus, perhaps it should not be a surprise that the U.S. Department of Justice has recently begun to focus greater attention on the inaccessibility of postsecondary education websites.

Undue Burden? Or Unfair Disadvantage?

The largest driver of inequality, however, is probably the concept known as “undue burden.” Under all of the disability laws, covered entities, both public and private, can claim that the requested accommodation is not financially or practically reasonable and thereby an undue burden under the law. If the accommodation is an undue burden, then the entity does not need to provide the accommodation. Corporate and public entities both tend to make very wide claims about what constitutes an undue burden, and courts tend to believe pretty much any claim of an undue burden. Corporations in particular have employed this exemption liberally as a defense against making any accessible versions of information technologies, including both the means to access the Internet and the websites on the Internet. Yet reliance on industry standards is insufficient to promote accessibility, and disability rights laws cannot be revised fast enough to match technological change at this point. As a result, the passage of Section 508 did not lead to major changes.

A potential reason why government technology did not become accessible, despite the Section 508 law, is that the focus on enforcement tended to be in procurement procedures, which is an indirect approach, and most websites are developed in-house, not acquired through contractual procurements. Furthermore, a piece of hardware that is procured is accessible or inaccessible upon purchase and generally does not change status until someone takes an action, whereas a website is a living, breathing information source that changes on a daily basis and therefore may be accessible one day and inaccessible another day. The procurement approach to monitoring accessibility simply does not work for websites, which need to have regular, ongoing evaluation for compliance with Section 508.

The burden-of-proof statements in IDEA and Section 504 certainly may contribute to the chronic lack of accessibility and proactive measures to address it at colleges and universities. Two examples illustrate this attitude. A quote from the Community College of Allegheny County (CCAC) on a webpage that discusses disability services states the following: “civil rights laws and the reasonable accommodations they call for are in no way intended—nor are they able to guarantee success. At most, a student can expect a more equal chance to do the same work as their peers ... . In higher education—as well as in employment, public services and public accommodations—the individual with a disability bears the burden of proof.” During the legal controversy over the accessibility of the Amazon Kindle and afterward, representatives from Cornell University have also voiced a similar attitude. When complaints were raised by blind students regarding the inaccessibility of websites and email at Cornell, although the University agreed to address the problems, the statement by the associate university counsel implied that Cornell had a responsibility to provide accessibility only upon request, rather than preemptively.

Essentially individuals with disabilities are expected to request accessibility, wait for it to happen, and then potentially be granted access to resources at a later time, which would inherently provide an unequal advantage to individuals without disabilities. This often happens to students when it comes to access to textbooks. Students with print disabilities need to have access to course textbooks at the same time as their fellow students. However, they often receive an accessible version of their course textbooks (e.g., an electronic version of their textbook or a large-print or Braille version), in the middle of the academic semester (such as seven weeks into a 15-week semester) or sometimes after the semester ends. Clearly students with disabilities are then at a disadvantage compared to their classmates who do not have disabilities. A delay in access to information is automatically discrimination.

Retrofitting Technology Access

There are many examples of the expense involved in retrofitting physical structures to address accessibility. The cost of accessibility when carefully planned and designed is almost zero. However, extreme challenges and significant expenses are often involved in the process of retrofitting an existing website for accessibility. To illustrate the challenges and potential costs, retrofitting a simple Tic-Tac-Toe computer game for accessibility resulted in the lines of code growing from 192 to 1,412. This type of retrofit would be time-consuming and costly, compared to a project started with universal access in mind. If the application were designed with accessibility in mind, the design could have done it with fewer statements of code.

A good example of the problems raised by retrofitting is the London subway system. While newer stations on the system are designed with accessibility in mind, older stations that were not so designed are not yet all accessible due to the technical difficulties and costs involved. Another example is the Singapore metro system. A report in 2004 noted the growing population of individuals with disabilities and the need to retrofit the metro system to accommodate them. The report noted that the cost of incorporating accessibility into new construction was minimal compared to the astronomical costs associated with retrofitting the system years later. For technology the pattern is very similar. The cost of retrofitting inaccessible design can sometimes be difficult, resource-intensive, and costly compared to the cost involved in creating accessible designs from the onset. Accessibility and usability have also been shown to add other value to products—as much as a 100-fold return on investment, according to early research.

People with disabilities, in fact, can serve as an excellent customer base for accessible products. They control discretionary funds that are more than twice those of teens and more than seventeen times greater than those of tweens, the two demographic groups currently most coveted by businesses. E-commerce use demonstrates the impact of accessible shopping options. While 53 percent of Internet users with disabilities engage in e-commerce activities—far lower than the 68 percent of other Internet users who shop online—online shoppers with disabilities are more loyal to a smaller number of sites.

Generally, industries have reacted to online accessibility requirements with “outright opposition, passive ignorance, acts of omission, or unwillingness to embrace required change.” In many cases corporations view people with disabilities as a niche market that they are not interested in or do not think it worth the effort to design for. Sadly, few businesses view people with disabilities as an important market or see public relations benefits from meeting the needs of people with disabilities. Developers and business owners are therefore unlikely to make changes unless they are legally bound to do so. With a minority of exceptions, providing accessibility is ignored in favor of faster market entrance for a product or simply because of long-standing assumptions regarding the need for accessibility and the cost of creating accessible products.

The most frequently cited reasons for not providing accessible technologies are that incorporating accessibility into the technology will:

increase costs of the technology;
lengthen time of development;
serve only a small market;
necessitate special design requirements;
result in low-tech and uncool products;
sacrifice aesthetics;
create difficulties in supporting accessible products;
never meet the needs of each different disability;
make the product worse for all other users.

While none of these claims are accurate, these stances are made possible, if not indirectly encouraged, by the federal laws that provide exemptions for undue burden, promote the creation of parallel versions of products, and embrace the anti-subordination approach to disability rights. Because born-accessibility is not required, retrofitting for accessibility has become the default approach for most in the private, nonprofit, and public organizations.

While the discussion above reveals many of the gaps in online access—equal or at all—for persons with disabilities that have resulted from the current approach of retrofitting, a more tangible example may help to illustrate the resistance to accessibility that individuals with disabilities often encounter. Shopping centers built before a certain time did not have the curb cuts that facilitate access for wheelchair users as well as people with carts or luggage, parents with strollers, bicyclists, and many other unintended beneficiaries. The placement of curb cuts added to older shopping centers often reflects a lack of interest in making them effective, such as placing them far from the designated parking spaces that wheelchair users would likely use, reflecting the fact that retail associations consistently oppose disability rights legislation.

This same obstructionist spirit is often reflected in the resistance to accessibility online. Retrofitting for accessibility often occurs only begrudgingly, after sufficient complaints are made, normally many years after the accessible version is made available. If a parallel version is created to provide for accessibility, it often has fewer features and capabilities, offers far less content, and is frequently out of date. The combination of these attitudes and the approach of the law has resulted in disastrous consequences in equal access online for persons with disabilities.

Social Responsibility

Despite the fact that the focus of this discussion is on legal and policy approaches and adjustments necessary to prioritize equitable accessibility, we would be remiss in not mentioning that regardless of the policy framework and legal environment, it is still incumbent on our society as a whole to educate developers to incorporate and understand accessibility better. Lack of training and tools has been a rationale of developers for excluding accessibility, so the computing and educational communities must continue both to publicize training and tools and make them widely available. We also need to encourage a culture of such social responsibility that corporations no longer feel rewarded for prioritizing early technology advancements over innovative yet accessible technology. It has also been suggested that a corporation’s attitudes and actions towards web accessibility may be a strong indicator of its commitment to other areas of corporate social responsibility and that a lack of commitment to accessibility can actually overshadow other positive efforts toward social responsibility. The dream of born-accessible technology motivated by social responsibility is, however, dependent upon the virtues of both developers and corporations, which may not always have social responsibility as part of their bottom line, and this is where strong public policies are of utmost importance.

Social responsibility can be fostered by other groups as well. Educators that work with students training to become developers—such as computer science faculty—could work to incorporate accessibility into curricula, so that all developers are prepared to make accessible products and understand that accessibility is the socially responsible approach to development. Researchers in fields related to accessibility—such as computer science, information science, sociology, public policy, and communication—could also support a culture of online accessibility by producing more research to show the impact of inaccessibility on people with disabilities, to support the development of accessible products, and to study policy options related to accessibility.

The greatest driver of increasing focus on and attention to the social responsibility of accessibility would likely be the federal government. Significant conceptual changes to policy approaches to accessibility would demonstrate a greater commitment to online accessibility that would be a major statement to developers and corporations. The next section proposes policy changes that could help foster a culture of social responsibility in accessibility.

Policy Recommendations and Conclusions

It has been suggested that the most effective Internet accessibility policy might focus on one of the core principles of the online environment--equity. For people with disabilities the policy approaches thus far have served more to reinforce inequality than to create equal access online. Because the social impacts of disability are closely tied to the choices society makes about disability, equal access online will be achievable only when federal government policy unequivocally asserts guarantees for accessible websites, social media tools, mobile devices, and all other aspects of the online environment. Better efforts to mitigate controllable risks can reduce the impacts of disability, but the failure to address systematic barriers to people with disabilities will result in ever-increasing social costs.

Unfortunately, the way that the disability rights law currently stands allows the practices of private, nonprofit, and public entities to undermine the overarching goals of the law concerning accessible technology. By permitting the general approach of retrofitting for accessibility through the combination of the anti-subordination approach, the exemptions for undue burden, and the acceptance of parallel versions of products, the law encourages the creation of inaccessible information and communication technologies that may eventually become accessible, but often do not. The current state of the law allows for separate but equal, but usually results in simply unequal.

The resulting current levels of Internet accessibility for people with disabilities are the direct result of policy choices made in the levels of accessibility expected, the ways in which the policies can be enforced, and the levels of effort put into enforcement. If people with disabilities are to move from being the most disadvantaged population online to equal residents of cyberspace, the philosophical approach of disability rights law needs to evolve. This evolution hinges on a rejection of the mentality of retrofitting and separate but equal, incentivizing instead a philosophy that emphasizes born-accessible technologies that are designed to be inherently inclusive of people with disabilities.

Such a change would not be a minor policy adjustment, however; it would necessitate a comprehensive, across-the-board reevaluation of the language of all disability rights laws. To alter the philosophical approach to accessibility, the changes to the laws should be based on a strong and clear legislative mandate that discourages retrofitting and instead shifts significantly more weight toward planning for accessibility and implementing accessibility early in the design and beyond only the minimum that will satisfy the law for today.

Specific changes would need to include:

Eliminating the ability to avoid compliance by invoking undue burden clauses or creating separate and unequal versions of technologies;
Creating a mandate for creating born-accessible technologies capable of providing the same features to all users, preventing a culture of retrofitting for accessibility;
Providing guidelines and best practices that promote the useful consideration of accessibility throughout design and implementation;
Ensuring that new requirements are broad enough to encompass the accessibility of current and future technologies, while also accounting for the full range of disabilities that can be affected by issues of accessibility;
Reducing the burden that an anti-subordination approach places on people with disabilities by converting the laws protecting them to provide the anti-differentiation approach that protects all other populations.

These would clearly be significant changes to existing laws and regulations. However, as the current state of accessibility demonstrates, major changes are needed if people with disabilities are to have an equal role in the age of the Internet.

Such policy changes would likely face resistance in the quarters that have traditionally fought against increases in rights for people with disabilities. Industry organizations, government agencies, and certain conservative members of Congress previously argued against the Rehabilitation Act and the ADA, primarily on economic grounds. Similar concerns were raised against online accessibility when the requirements of Section 508 were supposed to have been implemented. However, people with disabilities have overcome resistance to previous increases in their legal rights through organized support of changes to the law, and there is no reason to believe that they would be unable to support further changes to the law actively. Given that online inaccessibility disadvantages most people with disabilities in one way or another, large numbers of people with disabilities would have strong incentives to support these policy changes actively.

With the enormous potential benefits that an accessible Internet could provide to people with disabilities—online education offering greater opportunities for higher education and better jobs, telecommuting and telework offering meaningful employment options, and social networking offering new ways to engage in social activities regardless of the barriers of the physical world—it is possible that no population could benefit more from the Internet than people with disabilities. These benefits will remain in the realm of the potential, however, until disability rights laws shift from a philosophy of retrofitting to a philosophy of born-accessibility.