The civil rights statute that points to another document

The Americans with Disabilities Act of 1990, Public Law 101-336, 104 Stat. 327, signed July 26, 1990 by the 101st Congress after passing as S. 933, codified at 42 U.S.C. sections 12101 and following and amended substantially in 2008, is the rare civil rights law whose operative content lives mostly outside its own text. The ADA Standards for Accessible Design are not a chapter of the statute. They are technical rules, measured in inches and pounds of force, adopted by the Department of Justice through notice and comment rulemaking and developed from guidelines issued by an independent federal board that most readers have never heard of. The statute says that places of public accommodation and public entities must not discriminate; the standards say how wide the door must be, how steep the ramp may run, and how high the counter may sit. A reader who wants to know what any given building must actually do will learn almost nothing from the statute alone and almost everything from the standards.

That split, between the promise in the statute and the numbers in the standards, is the subject of this article. The design rules are where Congress delegated the hard decisions: how much access existing buildings owe, how far a renovation must reach, what counts as too costly, and which defenses excuse noncompliance. Five mechanisms do the work. A three-body structure divides guideline writing, standard setting, and enforcement among different actors. A three-tier obligation sorts every building into existing, altered, or new and assigns each a different duty. A safe harbor protects elements that satisfied earlier rules, within strict limits that are constantly overstated. Three defenses, constantly conflated, excuse different duties under different titles and are measured against different resources. And on websites, the highest volume compliance question in the field, the rules never arrived at all, leaving the courts to divide over first principles. This is the practitioner article for architects, facility managers, small business owners, and the lawyers who advise them. Its answer to the single test question, what any given facility must actually do, is that the answer depends on the tier, the defense, and the date, and that anyone offering a simpler answer has not read the regulations.

The five mechanisms are worth naming once more before the detail begins, because they are the article’s load-bearing structure. The three-body structure explains who decides. The three-tier obligation explains what is owed. The safe harbor explains what past compliance protects. The three defenses explain what excuses noncompliance. And the website question explains where the framework runs out, leaving courts to divide over first principles. A reader who can state all five, with the citations that carry each, has the whole subject in hand.

The article’s reference date is September 15, 2015, and every proposition stated as the law is the law as it stood on that date. Developments after that date carry explicit dates and are framed as later developments, never as the law at the reference date.

Three bodies, one set of enforceable rules

The technical content of the disability design rules comes from three distinct actors, and the division among them is deliberate. The Architectural and Transportation Barriers Compliance Board, usually called the Access Board, is an independent federal agency whose statutory job is to develop minimum accessibility guidelines. The Department of Justice takes those guidelines and converts them into enforceable design standards through notice and comment rulemaking. Enforcement then runs through the Department, which can investigate and sue, and through private plaintiffs, who can bring their own suits for injunctive relief. The body that writes the technical content does not enforce it, and the body that enforces it did not write the underlying measurements.

The statute builds this division into its own text. Title III directs the Attorney General to issue regulations whose standards “shall be consistent with the minimum guidelines and requirements issued by the Architectural and Transportation Barriers Compliance Board in accordance with section 12204 of this title,” 42 U.S.C. 12186(c), while the Department’s own section-by-section analysis of the 2010 rules insists that “although the ADA requires the enforceable standards issued by the Department under title II and title III to be consistent with the minimum guidelines published by the Access Board, it is the sole responsibility of the Attorney General to promulgate standards and to interpret and enforce those standards.” Consistency is required; independence is preserved. The Board supplies the floor, and the Department decides what the law commands above it. Remedies run through 42 U.S.C. 12188(b) for Department enforcement under Title III, 42 U.S.C. 12133 for Title II remedies, and 42 U.S.C. 12188(a) for the private right of action under Title III, which makes private plaintiffs genuine co-enforcers rather than mere complainants.

This three-body design has a parallel in the workplace safety statute, though the parallel requires care. The Occupational Safety and Health Act divides its labor differently: the Occupational Safety and Health Administration inside the Department of Labor both sets and enforces standards, an independent three-member Occupational Safety and Health Review Commission handles adjudication, and the National Institute for Occupational Safety and Health inside the health department conducts research and makes recommendations. The roles are not the same as the disability statute’s, where the guideline writer does not adjudicate and the enforcer adopts rather than invents the technical content. What the two statutes share is the structural instinct: divide the work among distinct bodies so that no single actor writes the rules, judges the disputes, and polices compliance all at once. Readers who want the full anatomy of the safety statute’s arrangement can consult the history of the Occupational Safety and Health Act of 1970, which shows the same instinct expressed through different institutions.

The reason for splitting authorship from enforcement is the same reason Congress created the Access Board in the first place. Writing accessibility guidelines is a technical undertaking that requires sustained attention to the details of the built environment: the turning space a wheelchair needs, the reach ranges of a seated person, the detectable warnings that make a curb ramp safe for a blind pedestrian. A law enforcement agency is not the natural author of that material, and a technical board is not the natural enforcer of civil rights. The statute therefore assigns the technical drafting to the specialists and the legal force to the department, with the consistency requirement binding them together.

Who actually writes the technical rules a business must follow?

The Access Board writes the minimum accessibility guidelines under 42 U.S.C. 12204, and the Department of Justice converts those guidelines into enforceable design standards through notice and comment rulemaking. The Board therefore supplies the technical content while the Department decides what becomes law, and the Department together with private plaintiffs enforces the result.

The practical consequence of the three-body structure is that compliance advice has three possible points of failure. A business can misunderstand the guidelines, which are only the minimum. It can misunderstand the enforceable standards, which are what a court applies. Or it can misunderstand who will hold it to account, since a private plaintiff needs no agency permission to file. The Department’s technical assistance materials, which explain the rules in plain language, are useful for all three purposes but are not themselves law; only the standards as adopted in the Code of Federal Regulations bind. Architects who design to the guidelines without checking the adopted standards, and owners who wait for the Department to contact them before acting, are each making the same category of error: treating one body as if it were all three.

The long gap between the 1991 rules and the 2010 Standards is itself instructive. The Access Board’s revision process ran for years, with the 2004 guidelines representing the Board’s considered view of what two decades of experience had taught. The Department’s subsequent rulemaking then took further years, as the proposal drew extensive comment from the disability community, industry groups, and state and local governments. The six-year span from the Board’s 2004 guidelines to the Department’s 2010 final rule, and the additional eighteen months to the March 15, 2012 compliance date, meant that the regulated community had nearly eight years of notice that new specifications were coming. The transition rules, the either-or provision, and the safe harbor were the Department’s answer to the fairness question that such a long rulemaking raises: those who built to the old rules in good faith would not be punished for the government’s slow revision, but neither would the old rules govern forever.

The Board’s independence also protects the guidelines from the enforcement pressures that might otherwise distort them. Because the Board does not litigate, its guidelines are written as technical judgments about what accessibility requires, not as litigation positions about what the Department can defend in court. The Department then makes its own litigation judgments when it adopts, interprets, and enforces the standards. The separation keeps each body’s work honest to its own function: the Board answers to expertise, the Department answers to law, and the courts answer to both as presented. A system that merged the functions would get guidelines written with an eye on the next case rather than on the next building.

The rulemaking path from guideline to enforceable rule

A guideline becomes a binding standard through notice and comment rulemaking, and the distinction between the two stages explains why the Access Board’s documents and the Department of Justice’s regulations are not interchangeable. The Board develops its guidelines through its own public process and publishes them as the minimum that accessibility requires. The Department then proposes a rule adopting those guidelines as enforceable standards, takes public comment on the proposal, and issues a final rule that responds to the comments and sets the binding text. Only the final rule, as codified in the Code of Federal Regulations, is law that a court applies. Everything before it is input, and everything the Department publishes afterward in the way of technical assistance is explanation.

The 1991 cycle shows the path in its simplest form. The Department adopted the original guidelines as Appendix A to 28 CFR Part 36 on July 26, 1991, publishing the adoption at 56 FR 35592, and amended the appendix in 1993 and 1994 as early experience revealed gaps and ambiguities. The 2010 cycle shows the path at full length. The Access Board issued revised guidelines in 2004, published at 36 CFR Part 1191, Appendices B and D. The Department then conducted its own rulemaking and published the final rule adopting the 2010 Standards on September 15, 2010, incorporating the Board’s 2004 guidelines rather than rewriting them. The incorporation matters for readers who compare the two documents: the technical provisions live in the Board’s text, while the scoping provisions, the effective dates, and the transition rules live in the Department’s regulatory text around them.

Two features of this rulemaking relationship deserve emphasis because practitioners get them wrong. First, the consistency requirement runs in one direction. The statute directs that the Department’s standards “shall be consistent with the minimum guidelines and requirements issued by the Architectural and Transportation Barriers Compliance Board in accordance with section 12204 of this title,” 42 U.S.C. 12186(c). The Department may go beyond the Board’s minimum, but it may not fall below it. The Board sets the floor, and the Department decides how far above the floor the enforceable law sits. Second, the Department’s own analysis of the 2010 rules insists that “although the ADA requires the enforceable standards issued by the Department under title II and title III to be consistent with the minimum guidelines published by the Access Board, it is the sole responsibility of the Attorney General to promulgate standards and to interpret and enforce those standards.” That sentence allocates authority among the three bodies with precision: the Board originates, the Department decides, and no one else’s document binds.

The practical upshot is a hierarchy of documents that every compliance file should respect. At the top stands the regulation as adopted, the only text a court enforces. Below it sit the Board’s guidelines, which explain the technical content and set the minimum the regulation must meet. Below those sit the Department’s technical assistance materials, which explain the regulation in plain language and are genuinely useful, but which create no obligations the regulation does not contain. An architect who designs to the guidelines without checking the adopted regulation may miss scoping provisions the Department added. An owner who treats a technical assistance answer as a safe harbor misunderstands what the document is. And a litigant who cites the Board’s guidelines against the Department’s regulation has the hierarchy upside down. The three-body structure only works when each body’s product is given its proper weight.

Who enforces and how

The standards would be advisory opinions without enforcement, and the statute assigns enforcement to two different kinds of plaintiffs. The Department of Justice enforces Title III through the authority granted at 42 U.S.C. 12188(b) and enforces Title II through the remedies provision at 42 U.S.C. 12133. Private plaintiffs enforce Title III through the private right of action at 42 U.S.C. 12188(a). The Access Board, the body that wrote the technical content, enforces nothing. That allocation is the point of the three-body design: the measurements come from the technical experts, the legal obligation comes from the Department’s rulemaking, and the policing comes from the Department together with private parties who need no agency permission to file.

The Department has described the overall arrangement as a three-level system that balances the access needs of people with disabilities against the constraints of operating a business, and the enforcement pattern reflects that balance. Department enforcement tends to address systemic or high-impact noncompliance, the cases where an investigation can move an entire chain or a whole municipality. Private enforcement tends to address the individual facility, the single restaurant or shop where a plaintiff encountered a barrier. The two channels are complements, not substitutes: the Department cannot inspect every building in the country, and private plaintiffs cannot rewrite national policy. A compliance program that watches only for Department investigations, or that dismisses private suits as a lesser threat, misunderstands the structure it operates under.

Enforcement also interacts with the tier system in ways that shape litigation strategy. A private plaintiff challenging an existing facility must engage the readily achievable analysis, showing what removal would have cost and why it was achievable. A plaintiff challenging an alteration engages the maximum-extent-feasible standard, a harder test for the defendant because cost alone does not excuse. A plaintiff challenging new construction faces the simplest case, since the design either met the specifications or did not. Defendants, for their part, reach for the defenses this article separates: undue hardship in employment cases, undue burden for auxiliary aids and program access, fundamental alteration where the requested change would transform the offering. The pleadings in a design case are, in miniature, the whole framework, with each side assigning the building to a tier and the duty to a defense.

From 1991 to 2010: how the technical rules grew up

The first generation of enforceable design rules arrived quickly after the statute. On July 26, 1991, the Department adopted the original ADA Accessibility Guidelines as Appendix A to 28 CFR Part 36, published at 56 FR 35592, and amended them in 1993 and 1994. Those 1991 rules governed design and construction for nearly two decades and became the reference point against which every later dispute about grandfathering would be argued. They were detailed, dimensional, and finite: a designer who satisfied them had a defensible position, and a litigant who ignored them had a problem.

The second generation took the better part of a decade to produce. The Access Board issued revised guidelines in 2004, and the Department published its final rule adopting the 2010 Standards on September 15, 2010. The new rules became effective March 15, 2011, and compliance was required beginning March 15, 2012, with a transition rule that allowed either the 1991 or the 2010 Standards to be used in full between publication and the compliance date. Compliance for newly constructed and altered swimming pools, wading pools, and spas was later extended to January 31, 2013 by a rule published May 21, 2012 at 77 FR 30171. The 2010 Standards consist of the regulatory text plus the 2004 guidelines issued by the Access Board at 36 CFR Part 1191, Appendices B and D, which the Department incorporated rather than rewrote.

The duties these design rules implement belong to the three titles of the statute, and the tiers described in this article only make sense against that background. Title I governs employment, Title II governs state and local government, and Title III governs places of public accommodation operated by private entities, and the design rules attach differently to each. The explanation of the three titles sets out who owes what to whom before any measurement is taken, and it is the necessary companion to everything that follows here.

Understanding the two generations matters because the transition between them is where most compliance mistakes live. The safe harbor discussed later in this article is a creature of that transition: it protects elements that satisfied the 1991 rules from immediate reworking under the 2010 rules, but only for elements that truly satisfied the earlier specifications, only until alteration, and not at all for elements the earlier rules never addressed. A facility manager who treats the 2010 Standards as a fresh start for every element, and one who treats the 1991 rules as a permanent shield, are both wrong in opposite directions.

Why the Board does not enforce

The separation between the Access Board’s guideline writing and everyone else’s enforcement is worth examining on its own, because it is the least intuitive feature of the three-body structure. In most regulatory schemes, the technical experts who write the rules also police them, or at least advise the policing. Here the Board writes the minimum guidelines and then steps aside entirely. It brings no enforcement actions, issues no binding interpretations of the Department’s regulations, and appears in litigation, if at all, as the author of the incorporated text rather than as a party. The Department’s insistence that promulgation, interpretation, and enforcement are “the sole responsibility of the Attorney General” is the legal expression of that distance.

The logic of the separation is institutional rather than personal. A body that both writes the technical content and enforces it would face a standing temptation to resolve ambiguities in favor of its own prior choices, expanding its guidelines through enforcement rather than through the public process that produced them. Separating the functions forces the technical content to survive two independent judgments: the Board’s expert judgment that the guidelines state the right minimum, and the Department’s legal judgment, tested through notice and comment, that the guidelines should become binding standards. The regulated community gets two bites at influence, commenting to the Board on the guidelines and to the Department on the adoption, and the resulting standards carry the legitimacy of both processes.

The workplace safety parallel sharpens the point by contrast. Under the Occupational Safety and Health Act, the standard setter and the enforcer are the same agency, the Occupational Safety and Health Administration, while independence is supplied by a separate adjudicator, the Occupational Safety and Health Review Commission, and by a research body, the National Institute for Occupational Safety and Health. The disability framework inverts the emphasis: the technical writer is the independent body, while setting and enforcement are combined in the Department. The two statutes thus solve the same problem, preventing any single actor from writing, judging, and policing everything, with different institutional geometries. What they share is the underlying distrust of concentrated regulatory power, expressed in each case through a division of labor that no single participant can override.

For practitioners, the separation has a concrete consequence: the Board’s documents are never the final word. A designer who finds a Board publication suggesting a particular approach must still check what the Department adopted, because the Department may have gone beyond the Board’s minimum or framed the requirement differently in the binding text. Conversely, a litigant who wants to argue that the standards mean something other than what the Department says they mean cannot enlist the Board as an ally against the Department’s interpretation. The Board originated the content; the Department owns the law.

The 2004 guidelines: the Board’s revision

The Access Board’s 2004 guidelines are the technical heart of the 2010 Standards, and they deserve a brief account of their own. The Board undertook the revision to incorporate what had been learned since 1991: which specifications had proved workable, which had proved ambiguous, and which facility types the earlier guidelines had never addressed. The resulting document, published at 36 CFR Part 1191, Appendices B and D, supplied the dimensional and technical provisions that the Department later incorporated into the enforceable standards. The Department’s regulatory text around the incorporation handles the matters only the Department can decide: applicability, effective dates, the transition rule, and the safe harbor.

The division of labor in the 2010 cycle thus repeated the three-body structure at the level of documents. The Board’s expertise produced the technical content; the Department’s rulemaking produced the legal obligation; and the two were joined by incorporation rather than by rewriting. A reader comparing the Board’s 2004 text with the Department’s 2010 regulation will find the measurements in the former and the law in the latter, which is exactly where the statute puts them. The arrangement also explains why challenges to the technical content and challenges to the transition rules travel to different places: the former implicate the Board’s expert judgment, the latter the Department’s regulatory choices.

What changed between 1991 and 2010

The 2010 Standards were not a restatement of the 1991 rules with new dates. They were a larger, more detailed, and more ambitious document, built on the Access Board’s 2004 guidelines and incorporating two decades of experience with what the earlier rules had missed. Understanding the delta between the generations is what makes the safe harbor intelligible, because the harbor exists precisely to manage that delta for elements that satisfied the old rules.

The most visible change was coverage. The 1991 rules addressed the common building elements of their era: entrances, routes, ramps, restrooms, signage, and the other features of offices, stores, and restaurants. The 2010 Standards added entire categories the earlier rules had never specified, the supplemental elements listed in the safe harbor provision itself: residential facilities and dwelling units, amusement rides, recreational boating facilities, exercise machines and equipment, fishing piers and platforms, golf courses and miniature golf, play areas, saunas and steam rooms, swimming pools and spas, and shooting facilities with firing positions. For these categories there was no 1991 baseline, which is why the safe harbor cannot apply to them: one cannot be protected for having satisfied a rule that did not exist. A water park, a golf course, or a fitness center that complied perfectly with the 1991 rules still faced new obligations under the 2010 rules for the elements the earlier generation never reached.

Beyond the new categories, the 2010 Standards revised technical specifications and scoping provisions across the board, reflecting the Board’s 2004 guidelines. The regulation’s structure stayed familiar, with the Department’s scoping and administrative text wrapped around the Board’s technical provisions, but the content grew more precise. That growth is the reason the transition rules mattered so much: the regulated community needed time to learn the new specifications, and the Department gave it to them through the delayed compliance date and the either-or transition rule described in the next section.

The transition rule and the compliance calendar

The move from the 1991 rules to the 2010 Standards was managed by a compliance calendar with four dates that every practitioner memorized. The final rule was published September 15, 2010. It became effective March 15, 2011. Compliance was required beginning March 15, 2012. Between publication and the compliance date, a builder could use either the 1991 Standards or the 2010 Standards in full, but could not mix and match provisions from each. The either-or rule prevented cherry-picking: no designer could take the lenient scoping of one generation and the lenient technical provisions of the other. Compliance for newly constructed and altered swimming pools, wading pools, and spas received its own extension to January 31, 2013, set by a rule published May 21, 2012 at 77 FR 30171, after the industry demonstrated that the pool-entry provisions needed more lead time.

March 15, 2012 does triple duty in the framework, which is why it recurs throughout this article. It is the date compliance with the 2010 Standards became mandatory. It is the trigger date for the safe harbor: only elements unaltered on or after that date keep the protection of 28 CFR 36.304(d)(2)(i). And it is the reference point for Title II’s path-of-travel harbor, which protects work built or altered to the earlier rules before that date, 28 CFR 35.151(b)(4)(ii)(C). A facility manager reconstructing the applicable rules for any element asks, in order: was it built or altered on or after March 15, 2012, and if so the 2010 rules apply; did it satisfy the corresponding 1991 specifications and remain unaltered since, and if so the safe harbor applies; is it a supplemental element the 1991 rules never addressed, and if so the harbor never applied. That three-question sequence resolves most transition disputes before they reach a lawyer.

Scoping and technical: how to read the ADA Standards for Accessible Design

The standards are organized around a distinction the safe harbor provision names explicitly: “technical and scoping specifications.” Scoping provisions say what must be accessible and how many: how many parking spaces, how many restrooms, which entrances. Technical provisions say how: the dimensions, slopes, clearances, and operating forces that make those elements usable. A designer who satisfies every technical specification but scopes the wrong elements has not complied, and a designer who scopes correctly but builds the ramp too steep has not complied either. The two halves are equally binding, and the safe harbor’s element-by-element test examines both: the element must have satisfied the corresponding technical and scoping specifications for that element in the 1991 rules.

Reading the standards therefore requires two passes. The first pass is scoping: for the facility type and size at issue, which elements must be accessible and in what numbers. The second pass is technical: for each scoped element, what the measurements require. Practitioners who read only the technical provisions produce beautifully dimensioned features in the wrong quantities, and practitioners who read only the scoping provisions produce the right number of features built wrong. The regulation’s organization reinforces the habit: the Department’s text around the incorporated guidelines handles applicability, dates, and transitions, while the incorporated technical provisions handle the measurements. The most common reading error is treating the document as a single list of numbers, when it is really a set of instructions for deciding what to measure followed by the measurements themselves.

Practitioners cite the 2010 Standards as a unit, but the unit has two layers, and knowing which layer a requirement comes from occasionally matters. The first layer is the department’s regulatory text, which contains the scoping provisions, the definitions, the safe harbor, the disproportionality rule, and the other provisions the department wrote as regulation. The second layer is the Access Board’s 2004 guidelines, adopted as appendices, which contain the bulk of the technical specifications: the dimensions, the slopes, the clearances, the maneuvering spaces, the signage rules. When a practitioner looks up the required width of an accessible route or the slope of a ramp, the answer comes from the guidelines layer. When a practitioner asks whether the safe harbor protects an element or how the twenty percent cap is calculated, the answer comes from the regulatory layer.

A practitioner opening the 2010 Standards for the first time benefits from knowing how the document is organized, because the organization mirrors the scoping-plus-technical logic described above. The working method the organization implies is a two-step lookup. First, the scoping provisions answer whether the element in question must be accessible at all and how many of them the facility must provide. Second, the technical provisions answer what accessible means for that element, down to the inch and the degree. A designer who skips the first step may build a fully compliant element the facility was never required to provide, while omitting one the scoping required. A designer who skips the second step knows what is required but not how to build it. The discipline of the document is the discipline of always asking both questions in order.

Tier one: the existing building and the achievable removal

The first tier answers the question every owner of an older building asks first. Title III treats the failure to remove architectural barriers, and structural communication barriers, in existing facilities as discrimination where removal is “readily achievable,” 42 U.S.C. 12182(b)(2)(A)(iv), and where removal is not readily achievable the entity must still make its goods and services available through alternative methods if those methods are readily achievable, 42 U.S.C. 12182(b)(2)(A)(v). The duty reaches backward in time: a building constructed before 1990 is not exempt from the statute, and the question is never whether the building predates the law but whether the barrier can be removed without much difficulty or expense.

“Readily achievable” has a regulatory definition, and the definition is deliberately modest. The regulation describes it as “easily accomplishable and able to be carried out without much difficulty or expense,” 28 CFR 36.304(a). The Department has long characterized the arrangement as a three-level system that balances the access needs of people with disabilities against the constraints of operating a business. The assessment turns on the circumstances of the particular business: the resources available to the enterprise enter the analysis, so a large national chain and a corner shop face the same legal test but will reach different answers about the same ramp. That relativity is not a loophole; it is the test working as designed.

Two features of this tier surprise even experienced practitioners. First, the duty is continuing. Readily achievable is not a budget test taken once and filed away; it is reassessed as circumstances change. A business that declined a modification as too costly in one year may owe it in another, because revenues rose, because the cost of the modification fell, or because a related renovation made the work cheaper to perform. The compliance question is never finally closed. That is the namable claim of this article, and it is the feature that most frustrates owners who want a definitive answer: the answer can change without any change in the law. Second, the duty is Title III’s alone in this form. Title II, which governs state and local government, does not use “readily achievable” for existing facilities at all. Instead it applies the program accessibility standard: each service, program, or activity, “when viewed in its entirety,” must be accessible to people with disabilities, 28 CFR 35.150(a), which means a public entity need not make every existing facility accessible so long as the program as a whole is. A city can comply by moving a program to an accessible building rather than by rebuilding the old one, an option no private place of public accommodation enjoys. Conflating the two titles here, by telling a city it must remove barriers where readily achievable or telling a restaurant it need only make its program accessible in the aggregate, is one of the commonest analytical errors in the field.

The three-level character of the system becomes clearer when the tiers are compared across time. The existing-facility duty is the most forgiving of the three, because it asks only what is easily accomplishable, and it is also the most persistent, because it never expires and is reassessed as circumstances change. The alteration duty is stricter but episodic, triggered only when the owner chooses to renovate. The new-construction duty is strictest but applies once, at the design stage. The Department’s three-level description captures this architecture: a lenient but endless duty for what exists, a demanding duty for what is changed, and an uncompromising duty for what is built new. The balance between access and business constraint is struck differently at each level, which is why the framework can be simultaneously criticized as too weak by advocates and too burdensome by owners, with each side pointing to a different tier.

The alternative-methods fallback deserves a final emphasis because it is the provision most often overlooked in compliance planning. Where barrier removal is not readily achievable, the statute does not release the entity; it redirects the duty to other means of providing access where those means are readily achievable. The classic illustrations are operational rather than architectural: taking orders or providing service at the curb, installing a buzzer or intercom at an inaccessible entrance, retrieving merchandise for customers who cannot navigate the aisles, or offering home delivery where the storefront cannot be made usable. These measures are not consolation prizes for failed compliance; they are compliance, of a different kind, and a business that provides neither removal nor alternative risks liability on both fronts. The fallback also has its own readily achievable analysis, which means the question is asked twice, once for the physical change and once for the operational one, with the resources of the business measured anew each time.

Do older buildings have to comply with the disability design rules?

Yes, within limits. A building constructed before the 1990 Act must remove barriers where removal is readily achievable, a continuing duty reassessed as finances and technology change. A building that satisfied the 1991 design rules gains a limited safe harbor against the 2010 rules, but the removal duty still reaches every element that never satisfied any earlier rule.

The regulation gives the abstract definition concrete illustrations. The readily achievable standard is illustrated by examples of steps that are often, though not always, readily achievable: installing ramps, making curb cuts in sidewalks and entrances, repositioning shelves, rearranging tables, chairs, vending machines, and other furniture, widening doors, installing offset hinges, insulating lavatory pipes, repositioning telephones, adding raised markings on elevator control buttons, installing flashing alarm lights, and similar modest measures. The list is illustrative rather than exhaustive, and no item on it is automatically required in every case; each is assessed against the entity’s resources and circumstances. But the list communicates the tier’s character: the inexpensive fix, the rearrangement, the modest retrofit, the change that a functioning business can absorb.

The statute’s inclusion of communication barriers alongside architectural barriers in the first tier is frequently overlooked. The duty covers barriers to communication that are structural in nature: the fixed, physical features of a building that impede communication, such as the absence of visual alarms for deaf occupants or the absence of adequate signage for blind visitors. It does not convert every communication difficulty into a barrier removal question; the auxiliary aids and services duty, governed by the undue burden defense rather than by readily achievable, handles the nonstructural side. But the structural communication barrier is squarely within the first tier, which means the restaurant whose only fire alarm is audible, or the office building whose permanent room signs lack raised characters, faces the same continuing, resource-sensitive duty as the restaurant with the stepped entrance.

The fallback provision deserves attention because it is the tier’s quiet workhorse. Where barrier removal itself is not readily achievable, 42 U.S.C. 12182(b)(2)(A)(v) requires alternative methods of making goods and services available where those methods are readily achievable: delivering goods to the curb, providing a buzzer at an inaccessible entrance, offering assistance with merchandise retrieval. These alternatives are not second-class compliance; they are the statute’s own answer to the hard case. A business that cannot afford the ramp but can afford the buzzer, and installs neither, has violated the law twice over. The hierarchy matters for litigation too, since courts expect a plaintiff to show what removal would have cost and a defendant to show why it was not achievable, with the alternative-methods question sitting in reserve.

Barrier removal in practice: the sequence of analysis

A facility manager facing the existing-building tier needs a sequence, not just a standard, and the regulation supplies one. The analysis runs in order: identify the architectural barrier, ask whether its removal is readily achievable, and if it is, remove it; if removal is not readily achievable, ask whether an alternative method of making the goods or services available is readily achievable, and if so, provide it. Each step is measured by the same definition, “easily accomplishable and able to be carried out without much difficulty or expense,” 28 CFR 36.304(a), but each step asks the question about a different action. The ramp that is not readily achievable does not end the inquiry; the buzzer, the curbside service, or the staff assistance that is readily achievable begins the next one.

The Department’s characterization of the arrangement as a three-level system captures what this sequence does. The first level is the removal duty itself, which pushes toward physical change. The second level is the alternative-methods fallback, which ensures that the inability to rebuild does not become an excuse to do nothing. The third level, implicit in the continuing nature of the duty, is reassessment: the analysis is redone as circumstances change, so that a barrier left standing this year is reconsidered next year. The system balances the access needs of people with disabilities against the constraints of operating a business not by splitting the difference once, but by requiring the question to be asked repeatedly, with each answer documented.

Documentation is where the tier is won or lost in practice. A business that can show its work, the barriers identified, the cost estimates obtained, the resources considered, the alternatives evaluated, presents a record of compliance even where barriers remain. A business that kept no record presents only an assertion. The regulation does not prescribe a particular form for the analysis, but the structure of disputes rewards the party with the paper: plaintiffs arrive with photographs and measurements, and defendants who arrive with nothing but testimony about tight budgets find the testimony weighed against the absence of any contemporaneous assessment. The annual facilities review that re-asks the readily achievable question is therefore not bureaucratic caution; it is the evidentiary foundation of the defense.

The regulation’s treatment of the readily achievable inquiry gives the practitioner a framework for the reassessment the statute requires. The assessment weighs the nature and cost of the barrier removal against the resources of the entity, considered in light of factors the regulation identifies: the overall financial resources of the site, the number of persons employed there, the effect on expenses and resources, legitimate safety requirements, and the impact on the operation of the site. No single factor controls. A costly modification may still be readily achievable for a prosperous enterprise, and a modest modification may not be for a struggling one. The multifactor character of the test is what makes the duty continuing rather than static: when any material factor changes, the balance recomputes, and the answer may change with it.

The barrier removal survey deserves a fuller description, because it is the compliance activity most entities perform worst. The survey catalogs the architectural barriers and the structural communication barriers in the facility, measured against the 2010 Standards’ technical provisions, which supply the yardstick even though the first tier does not demand full compliance with them. The catalog is then prioritized, and the prioritization follows the logic the department’s guidance has long suggested: the entrance first, because a person who cannot enter cannot use anything inside; then the route to the goods and services; then the restrooms; then the remaining elements. The documentation records the assessment and its date, which matters because the duty is continuing and the record of a past assessment is the evidence of a genuine achievability judgment rather than inattention.

The sequence also clarifies the relationship between the removal duty and the safe harbor. The two are sometimes described as if the harbor replaced the duty for older buildings, but they operate on different questions. The removal duty asks whether a barrier can be removed without much difficulty or expense. The harbor asks whether an element that satisfied the 1991 specifications must be upgraded to the 2010 specifications. An element can be harbored against the upgrade and still be subject to removal analysis if it presents a barrier the earlier rules never cured, and more importantly, an element that never satisfied any rule is subject to the removal duty with no harbor at all. The facility manager’s sequence handles the harbor as a preliminary filter: harbored elements are set aside from the upgrade question, and every element, harbored or not, remains in the removal analysis where it presents a barrier.

New construction: designing from the blank page

The new-construction tier rewards a different professional habit than the other two: designing to the specifications from the first sketch rather than verifying compliance at the end. Because the standard is full compliance with no readily achievable qualifier and no proportionality cap, the cost of a late-discovered violation is the cost of rebuilding finished work. Architects who integrate the design rules into schematic design, checking scoping as the program develops and technical provisions as details are drawn, rarely face the retrofit problem. Architects who treat accessibility as a final review item discover violations in concrete.

The trigger for the tier is the date of first occupancy. Facilities “designed and constructed for first occupancy after January 26, 1993” must be “readily accessible to and usable by individuals with disabilities,” 42 U.S.C. 12183(a)(1), implemented at 28 CFR 36.401(a). The date matters for buildings on the boundary: a facility first occupied before it is measured against the tier for existing buildings and alterations, while one first occupied after it must have been built to the specifications. The phrase “designed and constructed” also matters, because it reaches the design professionals as well as the owners; the drawings themselves must specify compliant work, not merely the finished building.

The structural impracticability exception, the tier’s sole escape, is narrow by design and turns on terrain rather than economics. Section 36.401(c) excuses compliance only where the unique characteristics of the land prevent incorporating the accessibility features: the hillside site where a level entrance is physically impossible, the constrained urban lot where the topography defeats the required route. What the exception never excuses is expense, inconvenience, or aesthetic preference. A developer who finds compliance costly has not shown impracticability; a builder who preferred a monumental staircase has not shown it either. The exception’s rarity is a feature of the framework, not a gap in it: the new-construction tier is where the law insists that a blank page produces an accessible building, and the only recognized excuse is that the page was not blank but sloped.

Tier two: the renovation that must reach as far as possible

The second tier governs alterations, and its standard is stricter than the first. Altered portions of a facility must comply with the design rules “to the maximum extent feasible,” 42 U.S.C. 12183(a)(2). The regulation sharpens the phrase: the obligation applies where full compliance is “virtually impossible,” and in that circumstance the entity must provide “the maximum physical accessibility feasible,” 28 CFR 36.402(c). Where the existing-facility tier asks what is easy, the alteration tier asks what is possible, and the difference between those two questions is the difference between a duty to try and a duty to succeed. A renovation that could have met the standards but did not, because the owner chose cheaper finishes or a faster schedule, fails the alteration test even if full compliance would have cost more than the owner wished to spend.

The alteration tier carries a second obligation that the other tiers do not. When an alteration affects the usability of, or access to, a primary function area, the entity must also make the path of travel to the altered area accessible, 28 CFR 36.403. A primary function area is the space where the main activity occurs: the dining room rather than the kitchen, the sales floor rather than the stockroom, the office where services are provided rather than the mechanical room. The path of travel includes the entrance, the route to the area, and the restrooms, telephones, and drinking fountains serving it. The logic is anti-evasion. Without this rule, an owner could renovate the showroom to perfection while leaving the only accessible entrance three blocks around the back, and the renovation would have made the facility less rather than more usable in practice.

The path-of-travel obligation is subject to a proportionality limit, and the limit is precise. Alterations made to provide an accessible path of travel to the altered area “will be deemed disproportionate to the overall alteration when the cost exceeds 20% of the cost of the alteration to the primary function area,” 28 CFR 36.403(f)(1). The statute itself frames the duty as applying where the path-of-travel work is “not disproportionate to the overall alterations in terms of cost and scope,” 42 U.S.C. 12183(a)(2), and the regulation converts that standard into arithmetic. When the 20 percent cap binds, the regulation sets a priority order for spending what is available: the entrance first, then the route to the altered area, then at least one restroom for each sex or a unisex restroom, then telephones, then drinking fountains, then other elements, 28 CFR 36.403(f)(2). The cap is Title III’s rule. Title II has its own path-of-travel disproportionality provision for public entities at 28 CFR 35.151(b)(4), and the two citations must not be swapped; an article that presents 36.403(f) as the rule for city hall has cited the wrong title.

The alteration tier is also where the safe harbor ends. The transition protection for elements that satisfied the 1991 rules applies only to elements “that have not been altered in existing facilities on or after March 15, 2012,” and once an element is altered past that date the safe harbor for that element is gone. A building can therefore sit comfortably under the old rules for decades and lose that comfort one renovation at a time, element by element, as each altered piece comes under the 2010 rules. Facility managers planning phased renovations should understand that sequencing is strategy: each phase resets the applicable rules for the elements it touches.

Alterations: the details that decide cases

The alteration tier generates more disputes than the other two combined, because renovations are where money is spent, where the safe harbor ends, and where the path-of-travel duty attaches. Three details decide most alteration cases, and each rewards close attention to the regulation’s text.

The first is the primary function area. The path-of-travel obligation attaches when an alteration affects the usability of or access to a primary function area, the space where the main activity of the facility occurs. The dining room is a primary function area; the kitchen is not. The sales floor qualifies; the stockroom does not. The office where the public receives services qualifies; the mechanical room does not. The classification determines whether the path-of-travel duty arises at all, so the opening move in any alteration dispute is characterizing the space. Owners sometimes argue that a renovation touched only back-of-house areas, and plaintiffs sometimes argue that the public’s experience of the space makes it primary; the regulation’s concept, not the parties’ labels, controls. The regulation’s illustrations extend the same logic to the patient examination rooms of a medical office and the classrooms of a school on the primary side, with boiler rooms, employee-only storage, and janitorial closets on the excluded side.

The second is the content of the path of travel itself. The duty reaches the entrance, the route to the altered area, and the restrooms, telephones, and drinking fountains serving it. These are the elements a person needs to arrive, to reach the renovated space, and to use the building’s basic amenities along the way. When the 20 percent cap binds, the priority order in 28 CFR 36.403(f)(2) sequences the spending: entrance first, then the route, then at least one restroom for each sex or a unisex restroom, then telephones, then drinking fountains, then other elements. Advisory guidance accompanying the regulation elaborates the same ordering for designers working within a constrained budget. The priority order embodies a judgment about what matters most: getting into the building comes before everything else.

The third is the relationship between alteration and the safe harbor. The harbor protects elements “that have not been altered in existing facilities on or after March 15, 2012,” which means each alteration ends the harbor for the elements it touches. A building can keep its 1991-compliant entrance under the harbor for decades while its restrooms, altered in a 2014 renovation, move to the 2010 rules. Facility managers planning phased work should treat sequencing as strategy: the order of renovations determines the order in which elements migrate to the new rules, and a phase that touches an element unnecessarily accelerates that migration. The alteration tier is also stricter than the existing-facility tier in a way that surprises owners: cost alone does not make compliance infeasible, so work that would have been excused as not readily achievable in an unaltered building must be performed to the maximum extent feasible once the renovation begins.

Title II runs its own alteration regime at 28 CFR 35.151, including its own path-of-travel disproportionality provision at 28 CFR 35.151(b)(4). The concepts parallel Title III’s, but the citation does not transfer, and a brief that cites 36.403(f) for a public entity’s renovation has cited the wrong title.

The public entity side of the alteration and new construction tiers runs through 28 CFR 35.151 rather than the Title III provisions this section has emphasized. Under the public entity title, new construction must comply with the design standards, alterations must comply to the maximum extent feasible, and alterations affecting a primary function area trigger a path-of-travel obligation subject to a disproportionality limit, at 28 CFR 35.151(b)(4). The structure mirrors the public accommodations tiers deliberately, but the citations differ, the enforcement mechanisms differ, and the defenses differ.

Planning a compliant renovation

The alteration tier rewards planning and punishes improvisation, because the duties it triggers are determined by decisions made before construction begins. A renovation planned with the framework in mind moves through a predictable sequence, and each step corresponds to a rule this article has already stated.

The first step is characterizing the work. The project team identifies which areas are primary function areas and which alterations affect their usability or access, because that classification determines whether the path-of-travel duty arises. A lobby reconfiguration that serves the building’s main entrance raises the question; a mechanical system replacement that the public never sees generally does not. The characterization should be documented, since it is the opening move in any later dispute about whether the duty attached.

The second step is budgeting the path of travel before the finishes. Because the 20 percent cap is measured against the cost of the alteration to the primary function area, the project needs a credible estimate of that cost early, and a separate estimate of the path-of-travel work the regulation requires: the entrance, the route, the restrooms, telephones, and drinking fountains serving the altered area. Where the path-of-travel estimate exceeds the cap, the priority order in 28 CFR 36.403(f)(2) determines what gets built, and the design should reflect that sequence rather than discovering it after bids come in. The common planning error is pricing the primary renovation to the last dollar and treating the path of travel as a contingency; the regulation treats it as an obligation with its own arithmetic.

The third step is sequencing for the safe harbor. Because alteration on or after March 15, 2012 ends the harbor element by element, the order of phases determines which elements migrate to the 2010 Standards and when. A phase that unnecessarily touches a harbored element accelerates its migration for no programmatic gain. Project teams sometimes assume the harbor protects the building as a whole and plan phases without regard to which elements each phase disturbs; the regulation’s element-by-element test makes that assumption expensive. The phasing plan should identify harbored elements and route work around them where the program permits.

The fourth step is applying the maximum-extent-feasible standard to the altered portions themselves. Unlike the existing-facility tier, this standard does not ask what is easily accomplishable; it asks what is possible, excusing full compliance only where it is virtually impossible and requiring then the maximum physical accessibility feasible. Value engineering that trades compliant details for cheaper substitutes fails this test, because cost is not the measure. The construction documents should specify compliant work throughout the altered areas, with any departures justified by genuine physical impossibility rather than by budget pressure.

The fifth step is reassessing the existing-facility tier for everything the renovation did not touch. The alteration does not suspend the readily achievable removal duty for the rest of the building; barriers outside the project area remain subject to the continuing analysis, possibly cheaper to address while contractors are mobilized. The renovation is therefore also an occasion to update the facility’s barrier removal documentation, closing the loop back to the first tier. A project that treats the alteration as the whole of its accessibility obligation has read the framework as a single tier; the framework is three, and the renovation implicates all of them at once.

The proportionality arithmetic

The 20 percent path-of-travel cap is the framework’s most mathematical rule, and working through its arithmetic shows how the tiers interact in a real project. The regulation provides that path-of-travel spending “will be deemed disproportionate to the overall alteration when the cost exceeds 20% of the cost of the alteration to the primary function area,” 28 CFR 36.403(f)(1). The denominator is the cost of the alteration to the primary function area itself, not the cost of the entire project; the numerator is the cost of making the path of travel accessible. Only the path-of-travel elements required by 28 CFR 36.403 count toward the numerator: the entrance, the route, the restrooms, telephones, and drinking fountains serving the altered area.

Consider the illustrations the regulation’s structure invites, stated here as illustrations rather than as reported cases. A restaurant renovates its dining room, a primary function area, at a cost of 200,000 dollars. The path-of-travel work, a new accessible entrance and an accessible route through the building, is estimated at 30,000 dollars. Fifteen percent of the primary-function-area cost falls below the cap, so the full path-of-travel work is required. Change the estimates: the same dining room renovation at 200,000 dollars, but the path-of-travel work, expanded to include restroom alterations, is estimated at 60,000 dollars. Thirty percent exceeds the cap, so the obligation is limited to 40,000 dollars of path-of-travel work, spent in the priority order of 28 CFR 36.403(f)(2): entrance first, then route, then restroom, then telephones, then drinking fountains, then other elements. The owner does not choose which elements to fund; the regulation sequences them.

The arithmetic also reveals the cap’s relationship to the alteration tier’s otherwise strict standard. Without the cap, the maximum-extent-feasible duty could require path-of-travel spending that dwarfs the renovation itself, turning a modest dining room refresh into a building-wide accessibility project. The cap prevents that outcome by deeming the excess disproportionate as a matter of law, not as a matter of argument. But the cap is also what makes the priority order necessary: because the available funds are limited, the regulation must decide which elements matter most, and it answers that getting into the building and reaching the altered area come before the supporting amenities. The statute’s underlying standard, that the path-of-travel work apply where “not disproportionate to the overall alterations in terms of cost and scope,” 42 U.S.C. 12183(a)(2), is thus implemented as a rule of arithmetic plus a rule of sequence, the kind of precise implementation the framework favors wherever it can provide one.

Tier three: the new building with no excuses

The third tier is the simplest and the strictest. Facilities “designed and constructed for first occupancy after January 26, 1993” must be “readily accessible to and usable by individuals with disabilities,” 42 U.S.C. 12183(a)(1), implemented at 28 CFR 36.401(a). There is no readily achievable qualifier, no proportionality cap, and no balancing test. The designer starts from a blank page, and the law expects the blank page to produce a compliant building. The standard is not perfection in some abstract sense; it is conformance with the technical specifications in the adopted design rules, measured the same way for every new building in the country.

The sole exception proves the strictness of the rule. New construction may depart from the specifications only where compliance is structurally impracticable, 28 CFR 36.401(c), and the regulation construes that exception narrowly: it applies where the unique characteristics of terrain prevent the incorporation of accessibility features, not where the builder finds compliance expensive or inconvenient. A hillside site that makes a level entrance physically impossible may qualify; a flat site where the developer preferred a grand staircase does not. The exception is about the land, not the budget. In practice it is invoked rarely and succeeds rarer still, which is exactly what the drafters intended: the new-construction tier is the one place in the framework where the law refuses to negotiate.

The three-tier obligation table

The three tiers can be held in a single table. Each row states the duty, the limiting principle that keeps the duty from becoming absolute, the event that triggers it, and the practical question a facility manager or architect should ask first. The table states the Title III tiers; the Title II existing-facility analog is program accessibility rather than readily achievable barrier removal, as the preceding sections explain.

Tier	Standard applied	Limiting principle	Trigger	Practical compliance question
Existing facility	Remove architectural and structural communication barriers where readily achievable, 42 U.S.C. 12182(b)(2)(A)(iv)	Easily accomplishable without much difficulty or expense, assessed against the circumstances of the business, 28 CFR 36.304(a)	Any barrier in a building that predates compliance; duty reassessed as circumstances change	Is removal of this barrier easily accomplishable without much difficulty or expense
Alteration	Altered portions comply to the maximum extent feasible, 42 U.S.C. 12183(a)(2)	Full compliance unless virtually impossible, then maximum physical accessibility feasible, 28 CFR 36.402(c); path of travel capped at 20 percent of primary-function-area alteration cost, 28 CFR 36.403(f)	A renovation affecting usability, with path-of-travel duties when a primary function area is touched	What is the most access this renovation can deliver, and does the path of travel stay within the cap
New construction	Comply fully with the design rules, 42 U.S.C. 12183(a)(1)	No readily achievable qualifier and no proportionality cap; structural impracticability the sole exception, 28 CFR 36.401(c)	Design and construction for first occupancy after January 26, 1993	Does the design meet every applicable specification from the start

The table’s practical value is diagnostic. When a client describes a project, the practitioner locates its row first and reads the limiting principle second, because the row determines which legal test applies and the principle determines how that test is argued. Misdiagnosis is the expensive error: analyzing a renovation under the readily achievable standard understates the duty, while analyzing an existing building under the maximum-extent-feasible standard overstates it. The table prevents both mistakes by forcing the tier question before the measurement question.

Used correctly, the table sorts any project into its tier before any measurement is taken. Used incorrectly, as a substitute for reading the regulations, it becomes the checklist the law refuses to be. The rows are starting points for analysis, not answers, because two of the three tiers turn on judgments about achievability, feasibility, and resources that no table can make in advance.

Program accessibility in practice

Title II’s program accessibility standard deserves its own exposition, because it is the most misunderstood duty in the framework and the one most often confused with Title III’s readily achievable test. The regulation provides that each service, program, or activity of a public entity, “when viewed in its entirety,” must be accessible to individuals with disabilities, 28 CFR 35.150(a). The unit of analysis is the program, not the building. A city need not make every existing facility accessible; it must make each program accessible in the aggregate.

In practice this gives public entities compliance methods that private businesses lack. A city can move a popular evening class from an inaccessible historic community center to an accessible school building across town. A county can offer home visits as an alternative to an inaccessible office. A parks department can concentrate its accessible programming at the facilities that can be made usable, provided the program as a whole, viewed in its entirety, serves people with disabilities on equal terms. None of these methods would satisfy Title III, where each facility must remove its own barriers where readily achievable. The difference reflects the titles’ different subjects: Title II regulates governments, which operate portfolios of facilities and can shift programs among them, while Title III regulates individual places of public accommodation, which cannot.

The program accessibility duty is limited by the undue financial and administrative burdens defense at 28 CFR 35.150(a)(3), and the regulation surrounds that defense with procedure. The burden of proof rests on the public entity. The decision must be made by the head of the entity or a designee, after considering all resources available for the program. It must be accompanied by a written statement of reasons. And even where the defense succeeds, the entity must take some other action to ensure that individuals with disabilities receive the program’s benefits. These requirements convert the defense from a conclusion into a process: a city that announces burden without the designee’s decision, the written statement, and the alternative action has not invoked the regulation. The procedural detail is also what distinguishes the Title II defense from Title III’s undue burden for auxiliary aids, which carries no equivalent written-statement machinery.

The safe harbor that is not a grandfather clause

No provision of the design rules is more cited and more misunderstood than the safe harbor. The Title III rule provides, in exact regulatory text, that “elements that have not been altered in existing facilities on or after March 15, 2012 and that comply with the corresponding technical and scoping specifications for those elements in the 1991 Standards are not required to be modified in order to comply with the requirements set forth in the 2010 Standards,” 28 CFR 36.304(d)(2)(i). Title II carries its own version at 28 CFR 35.150(b)(2), which also reaches elements that conformed to the Uniform Federal Accessibility Standards. The protection is real, but every word of it is a limitation.

First, the trigger date matters. Only elements unaltered on or after March 15, 2012 keep the protection; an element altered after that date must meet the 2010 rules. The brief shorthand that elements complying with the 1991 rules “need not be modified until altered” is imprecise precisely here: an alteration in 2005 did not end the safe harbor, and an alteration in 2013 did. Second, the protection is element-by-element, not building-wide. The element must satisfy the corresponding technical and scoping specifications for that element in the 1991 rules, which means a compliant entrance does not protect a noncompliant restroom in the same building. A facility cannot wave its 1991 certificate of occupancy as a blanket shield; the regulation asks about each element separately. Third, and most consequentially, the safe harbor does not apply at all to supplemental elements, the categories the 2010 rules addressed for the first time and the 1991 rules never specified. Section 36.304(d)(2)(iii) lists them: residential facilities and dwelling units, amusement rides, recreational boating facilities, exercise machines and equipment, fishing piers and platforms, golf courses and miniature golf, play areas, saunas and steam rooms, swimming pools and spas, and shooting facilities with firing positions. For these elements there is no earlier rule to have satisfied, so there is nothing to be harbored; they must be modified to the extent readily achievable to meet the 2010 rules. An amusement park that complied perfectly with the 1991 rules still owes the 2010 rules for its rides, because the 1991 rules said nothing about rides.

A worked illustration shows how the element-by-element character operates in practice. Consider a hotel built in 1995 whose guest rooms, lobby restrooms, and entrance complied with the 1991 Standards, and whose swimming pool, added in the same year, complied with everything the 1991 edition required of pools, which was little, since the 1991 edition contained no pool accessibility specifications. In 2013, the owner renovates the lobby restrooms. The restrooms, being altered on or after March 15, 2012, must meet the 2010 Standards, and their safe harbor ends. The guest rooms and entrance, unaltered, keep their protection: their 1991 compliance continues to excuse them from the 2010 edition. The pool, however, was never protected at all, because pools are among the supplemental elements the 1991 Standards did not address; the pool must be modified to the extent readily achievable to meet the 2010 Standards regardless of the alteration schedule. Three elements of one building, three different answers, all flowing from the same regulatory text. The illustration is hypothetical, but the structure it demonstrates is the regulation’s.

The meaning of alteration for safe harbor purposes tracks the ordinary sense of the word as the regulation uses it elsewhere: a change to the element that affects its usability or its compliance characteristics. Routine maintenance, such as repainting a compliant handrail or replacing a worn carpet with an identical one, does not end the protection, because it does not alter the element in the relevant sense. Reconfiguring the restroom, moving the fixtures, or changing the dimensions does. The line between maintenance and alteration is occasionally contested, and the contest matters because crossing it ends the safe harbor for the element crossed.

The element-by-element character of the harbor rewards careful record-keeping and punishes assumption. Because the protection asks whether each element satisfied the corresponding 1991 specifications, a facility needs to know, element by element, what the 1991 rules required and whether its construction met them. Original construction documents, certificates of occupancy, and contemporaneous inspection records become the evidence of the harbor, decades after the work was done. A facility that kept those records can establish the harbor element by element; a facility that discarded them may find the protection theoretically available but practically unprovable. The harbor thus favors the organized owner, not as a matter of legal design but as a matter of evidence.

The Title II harbor adds its own wrinkle. Section 35.150(b)(2) reaches elements that conformed either to the 1991 Standards or to the Uniform Federal Accessibility Standards, the federal government’s earlier accessibility code. Public entities whose older facilities were built to UFAS therefore carry a harbor the Title III regulation does not provide, reflecting the federal government’s longer history of building to its own accessibility code before the 1990 statute. The path-of-travel harbor at 28 CFR 35.151(b)(4)(ii)(C) similarly protects path-of-travel elements built or altered to the 1991 Standards or UFAS before March 15, 2012 from being retrofitted solely because a later alteration touches the primary function area they serve. These Title II provisions share the Title III harbor’s logic, protecting past compliance from becoming the trigger for new obligations, while sharing its limits: only what actually complied is protected, and the program accessibility duty continues regardless.

Is there a grandfather clause in the disability design rules?

No. The rules contain a limited safe harbor, not a grandfather clause. Elements that satisfied the corresponding 1991 specifications need not be upgraded to the 2010 rules until they are altered on or after March 15, 2012. But the readily achievable removal duty continues for every element that never satisfied any earlier rule, so age alone never excuses a barrier.

The reason the safe harbor is not a grandfather clause lies in the tier it was built to temper. The readily achievable barrier removal duty applies to every barrier in every existing facility, and it applies whether or not the element ever satisfied any design rule. The safe harbor excuses an element from the upgrade to the 2010 specifications; it does not excuse the building from the removal duty for elements that never complied with anything. A doorway that violated the 1991 rules when they were current, and still violates them, gets no safe harbor against the 2010 rules, because there is no 1991 compliance to protect, and it remains subject to removal where readily achievable. That is why the common claim that old buildings were grandfathered fails twice: it overstates the protection for elements that complied, and it invents protection for elements that did not. Readers who want the full treatment of this and the other persistent misunderstandings should see the examination of disability law myths, which takes up the grandfathering claim directly.

Title II adds one more harbor worth knowing. Elements of a path of travel that were built or altered to the 1991 rules or the Uniform Federal Accessibility Standards before March 15, 2012 need not be retrofitted solely because a later alteration touches the primary function area they serve, 28 CFR 35.151(b)(4)(ii)(C). The logic is the same as the main harbor: do not punish past compliance by making it the trigger for new obligations. But like the main harbor, it protects only what actually complied, and it does not touch the program accessibility duty that is Title II’s real measure for existing facilities.

Three defenses that are not interchangeable

Conflating the three defenses is the most common analytical error in this area, and the error is understandable because the names sound alike. Undue hardship, undue burden, and fundamental alteration are three different doctrines, applying to different duties under different titles, measured against different things. Keeping them separate is not pedantry; it determines who bears the burden of proof and what evidence matters.

Undue hardship belongs to employment. Title I provides that an employer need not provide a reasonable accommodation where doing so would impose “an action requiring significant difficulty or expense, when considered in light of the factors set forth in subparagraph (B),” 42 U.S.C. 12111(10). The factors look to the nature and cost of the accommodation, the financial resources of the facility and its headcount, the effect on the facility’s operations, and the resources, size, and operations of the entity as a whole. The measurement is against the employer’s resources, which means the defense is employer-specific by design: the same accommodation can be reasonable for a large employer and an undue hardship for a small one. The doctrine never leaves the employment title. Invoking undue hardship to excuse a Title III barrier removal, or a Title II program access decision, is citing the wrong law.

Undue burden belongs to two different places, and each must be cited correctly. Under Title III, the failure to provide auxiliary aids and services, such as sign language interpreters or accessible formats, is discrimination “unless the entity can demonstrate that taking such steps would fundamentally alter the nature of the good, service, facility, privilege, advantage, or accommodation being offered or would result in an undue burden,” 42 U.S.C. 12182(b)(2)(A)(iii). Note the structure of that clause: it carries both the fundamental alteration defense and the undue burden defense in the alternative, and the two must not be presented as if they lived in separate provisions. Under Title II, the parallel concept is “undue financial and administrative burdens” in the program accessibility regulation: a public entity need not take an action it “can demonstrate would result in a fundamental alteration in the nature of a service, program, or activity or in undue financial and administrative burdens,” 28 CFR 35.150(a)(3). The regulation adds procedural teeth that the statute’s defenders sometimes overlook. The burden of proof rests on the public entity, not the complainant. The decision must be made by the head of the entity or a designee, after considering all resources available for the program. It must be accompanied by a written statement of reasons. And even where the defense succeeds, the entity must still take some other action that ensures people with disabilities receive the program’s benefits. A city that invokes undue financial burden without the written statement, the designee’s decision, and the alternative action has not invoked the defense; it has announced a conclusion.

The citation that must never be used here deserves explicit warning. Section 12134 of Title 42 is the Title II rulemaking provision: it directs the Attorney General to issue regulations. It is not a defense, and it has never been one. Briefs and memoranda that cite 42 U.S.C. 12134 for undue burden are repeating an error that the regulation’s own numbering invites, since 35.150(a)(3) is where the defense actually lives. A lawyer who cites the rulemaking provision as the defense has told the court, in effect, that the Department’s authority to write rules is the reason the city need not follow them.

What separates undue burden from fundamental alteration?

Undue burden measures cost and administrative difficulty against the resources of the whole entity; a well funded business can owe more than a struggling one for the same accommodation. Fundamental alteration asks whether the requested change would transform the nature of what is offered, regardless of cost, as when a change would rewrite the rules of the program itself.

Fundamental alteration is the third defense, and it is the one that has nothing to do with money. Under Title III, a place of public accommodation must make reasonable modifications in policies, practices, and procedures where necessary to avoid discrimination, unless doing so “would fundamentally alter the nature of the good, service, facility, privilege, advantage, or accommodation being offered,” 42 U.S.C. 12182(b)(2)(A)(ii). Under Title II, a public entity must make reasonable modifications in policies, practices, or procedures when necessary to avoid disability discrimination, “unless the public entity can demonstrate that making the modifications would fundamentally alter the nature of the service, program, or activity,” 28 CFR 35.130(b)(7)(i). The question is not what the change costs but what the change does: whether the requested modification would transform the thing itself into something else. Waiving an essential eligibility requirement, or rewriting the rules of competition, are the classic illustrations; installing a ramp at great expense is not, because expense is the domain of the other two defenses. That is the cleanest way to keep the three apart. Undue hardship and undue burden ask how much the duty costs relative to the resources at hand. Fundamental alteration asks whether the duty, performed, would still be the same duty. A defendant that argues cost under the name of fundamental alteration, or argues transformation under the name of undue burden, has the right facts in the wrong box, and courts notice the mismatch.

Fundamental alteration, finally, is the defense that asks what the thing is rather than what it costs, and its illustrations clarify the boundary. A public swimming program that waived its requirement that participants be able to swim would not be modifying the program but replacing it. A professional licensing examination that eliminated the skill it was designed to test would not be accommodating the applicant but abolishing the examination. A theater asked to admit a fundamentally incompatible use of its stage would not be modifying its seating policy but changing its business. In each illustration, the requested change does not make the existing thing accessible; it makes a different thing.

Why the conflation persists is worth a paragraph, because the persistence itself teaches something about legal language. The three resource defenses share a family resemblance: each excuses an entity from an accessibility duty where compliance would cost too much, measured against that entity’s means. The resemblance is real but shallow. The duties they excuse are different, the entities they protect are different, the measures are phrased differently, and the procedural requirements differ, with the public entity defense carrying documentation duties the others do not. A practitioner who treats them as one general cost defense will cite the wrong provision, apply the wrong measure, and miss the procedural requirements, and the error will surface at the worst possible moment, when the defense is tested.

What is the practical difference between undue hardship and undue burden?

Undue hardship is the employment defense under 42 U.S.C. 12111(10), measured by significant difficulty or expense against the employer’s resources. Undue burden is the public accommodations defense under 42 U.S.C. 12182(b)(2)(A)(iii), measured against the entity’s resources for auxiliary aids and services. They share the word undue and nothing else.

Auxiliary aids and services: where Title III’s undue burden lives

The undue burden defense under Title III belongs to a specific duty that is easy to overlook in an article about buildings: the duty to provide auxiliary aids and services. The statute treats the failure to furnish such aids as discrimination “unless the entity can demonstrate that taking such steps would fundamentally alter the nature of the good, service, facility, privilege, advantage, or accommodation being offered or would result in an undue burden,” 42 U.S.C. 12182(b)(2)(A)(iii). Auxiliary aids and services are the communication measures that make goods and services usable: sign language interpreters, assistive listening systems, accessible formats, and similar steps that bridge the gap between the offering and the customer who needs it in a different form.

The placement of the defense matters. Unlike the barrier removal duty, which carries its own cost-sensitive test in the readily achievable standard, the auxiliary aids duty is stated without an internal cost limit, which is why it needs the undue burden defense alongside it. The clause’s structure is also what the verifier flagged as essential to get right: it carries both the fundamental alteration defense and the undue burden defense in the alternative, joined by “or,” and the two must not be presented as if they inhabited separate provisions. A defendant may argue either that providing the aid would transform what is offered or that providing it would impose an undue burden; the clause permits both, and the evidence for each is different.

The undue burden analysis for auxiliary aids compares the cost and administrative difficulty of the step against the resources of the entity, the same resource-relative logic that runs through the framework. A large business may be required to provide an interpreter where a small one is excused, not because the statute favors large businesses but because the burden is measured against what each can bear. And where the defense succeeds, the inquiry does not end: the entity must still consider whether some alternative aid, less costly or less difficult, would serve, because the defense excuses the particular step shown to be an undue burden rather than the duty itself. The structure mirrors the barrier removal tier’s fallback to alternative methods, and it reflects the same legislative judgment: cost can limit the means, but it does not extinguish the obligation to try.

The defenses in litigation: who proves what

The three defenses share a procedural feature that matters more than their substantive differences: in each case, the defendant proves the defense. The Title III auxiliary aids provision excuses the entity only where it “can demonstrate” undue burden or fundamental alteration, 42 U.S.C. 12182(b)(2)(A)(iii). The Title III reasonable modifications provision excuses only where the accommodation “would fundamentally alter” the offering, a showing the defendant must make, 42 U.S.C. 12182(b)(2)(A)(ii). The Title II program accessibility regulation is explicit that the public entity must “demonstrate” undue financial and administrative burdens or fundamental alteration, 28 CFR 35.150(a)(3), and adds the procedural machinery of the designee’s decision and the written statement of reasons. The Title I employment defense similarly requires the employer to show significant difficulty or expense in light of the statutory factors, 42 U.S.C. 12111(10). No defense in this framework is presumed; each is an affirmative case the defendant builds.

That allocation shapes discovery and trial. A Title III defendant invoking undue burden for auxiliary aids must produce evidence of cost and of the resources against which cost is measured, because the defense compares the two. A defendant invoking fundamental alteration must produce evidence of what the good, service, or program essentially is and how the requested change would transform it, because the defense is about identity rather than expense. A Title II defendant invoking undue financial and administrative burdens must produce the written statement, identify the deciding official, and show the alternative action taken, because the regulation makes those elements part of the defense itself. A Title I employer must address the statutory factors, from the nature and cost of the accommodation through the entity’s overall resources and operations. In each forum, the defendant who treats the defense as a label to be asserted finds it treated as a case to be proved.

The burden allocation also explains why conflating the defenses is dangerous in litigation and not merely sloppy in scholarship. A defendant that argues cost under the name of fundamental alteration has invoked a defense that does not consider cost, and the court will evaluate the cost evidence against a standard that ignores it. A defendant that argues transformation under the name of undue burden has invoked a defense measured against resources, and the court will ask for financial evidence the defendant never prepared. An employer that cites undue burden, the Title III and Title II concept, instead of undue hardship, the Title I concept, has cited law from the wrong title. The framework’s separations are load-bearing in the courtroom: each defense carries its own elements, its own evidence, and its own burden, and the boxes do not interchange.

Reading the regulation: where each rule lives

Practitioners need a map of the Code of Federal Regulations provisions that carry the framework, because the citations are numerous and the titles must not be mixed. This section collects them in one place, with each provision stated as the verifier confirmed it.

Title III’s existing-facility duty lives at 42 U.S.C. 12182(b)(2)(A)(iv), with the alternative-methods fallback at clause (v) and the regulatory definition of readily achievable at 28 CFR 36.304(a). The Title III safe harbor lives at 28 CFR 36.304(d)(2)(i), with the supplemental elements exclusion at 36.304(d)(2)(iii). Title III’s alteration duty lives at 42 U.S.C. 12183(a)(2), implemented at 28 CFR 36.402, with the maximum-extent-feasible construction at 36.402(c) and the path-of-travel duty at 28 CFR 36.403, including the 20 percent disproportionality rule at 36.403(f)(1) and the priority order at 36.403(f)(2). Title III’s new-construction duty lives at 42 U.S.C. 12183(a)(1), implemented at 28 CFR 36.401(a), with the structural impracticability exception at 36.401(c). Title III’s defenses live at 42 U.S.C. 12182(b)(2)(A)(ii) for reasonable modifications and fundamental alteration, 42 U.S.C. 12182(b)(2)(A)(iii) for auxiliary aids with its alternative undue burden and fundamental alteration defenses, and 42 U.S.C. 12111(10) for Title I undue hardship. Title III’s rulemaking and enforcement provisions live at 42 U.S.C. 12186(c), 42 U.S.C. 12188(b), and 42 U.S.C. 12188(a).

Title II’s program accessibility duty lives at 28 CFR 35.150(a), with the undue financial and administrative burdens and fundamental alteration defenses at 28 CFR 35.150(a)(3) and the Title II safe harbor at 28 CFR 35.150(b)(2). Title II’s new construction and alteration rules live at 28 CFR 35.151, with the Title II path-of-travel disproportionality provision at 28 CFR 35.151(b)(4) and the path-of-travel harbor at 28 CFR 35.151(b)(4)(ii)(C). Title II’s reasonable modifications duty and fundamental alteration defense live at 28 CFR 35.130(b)(7)(i). Title II’s remedies live at 42 U.S.C. 12133. The Access Board’s guidelines live at 36 CFR Part 1191, Appendices B and D, incorporated into the 2010 Standards. Held together, these citations are the framework’s skeleton: every duty, every defense, and every transition rule in this article hangs on one of them.

The structural point worth carrying out of this section is that the three defenses map onto the three tiers without duplicating them. The readily achievable test of the existing-facility tier already accounts for cost, which is why Title III needs no separate undue burden defense for barrier removal: the tier itself is the cost-sensitive rule. The auxiliary aids duty needs its own undue burden defense because it is not otherwise cost-limited. The alteration tier needs no cost defense because feasibility, not expense, is its measure, which is why cost arguments fail there. And fundamental alteration stands apart from all three tiers, guarding the identity of what is offered rather than the price of offering it. The framework is economical in this way: each cost question is asked exactly once, in the place where it belongs, and conflating the defenses double-counts or misplaces the question the law actually poses.

Websites: the question the rules never answered

Websites are the highest volume compliance question in the field, and the design rules are silent about them. The silence is not an oversight in the drafting; it is a rulemaking that never finished. The Department has long taken the position that Title III applies to the websites of places of public accommodation, treating the statutory guarantee of the full and equal enjoyment of goods and services as reaching the digital storefront the same way it reaches the physical one. On July 26, 2010, the Department issued an advance notice of proposed rulemaking on website accessibility under Titles II and III, the first formal step toward a technical rule for private websites. As of the reference date of this article, September 15, 2015, that rulemaking had produced no final rule, which meant that the Department’s position existed as enforcement policy and litigation posture without the dimensional specificity that the design rules give to doorways and ramps. Two later developments belong in the record with their dates stated plainly: the advance notice was withdrawn on December 26, 2017 without a rule ever issuing, and a final rule for state and local government web content, adopting a recognized technical standard, issued on April 24, 2024 under Title II. Neither development was the law at the reference date, and neither should be read backward into it.

The Department pressed its view in court as well as in rulemaking. In litigation brought by the National Association of the Deaf against the streaming service Netflix, the Department filed Statements of Interest on October 3, 2011 and May 15, 2012, setting out its position that the public accommodations title reaches web-based services. A Statement of Interest is the United States telling a court what the enforcing agency believes the law means; it is not a ruling, and it does not bind the court, but it put the Department’s interpretation on the record in a high-visibility dispute. The episode illustrates the enforcement posture of the period: the agency advancing its reading case by case while the rulemaking that would have settled the technical questions sat unfinished.

The federal courts, meanwhile, divided, and the division is best described by its reasoning rather than by its scorecard. One line of reasoning treats the website as itself a place of public accommodation, or treats the website as the means by which a place of public accommodation offers its goods and services, and concludes that the statute’s nondiscrimination duties apply to it directly. On this view, a business that sells to the public through a website is offering goods and services within the meaning of the title, and the absence of a physical threshold on the transaction does not remove the transaction from the statute. The other line of reasoning requires a nexus between the website and a physical place of business, holding that the title’s list of covered establishments describes physical places and that a website unconnected to such a place falls outside it. On this view, the statute regulates places, and a purely virtual operation is not a place however much commerce flows through it. The two readings cannot both be right about the same website, which is why a business asking what the law requires of its site received different answers in different courtrooms at the reference date.

Must a business website follow the standards?

Through September 15, 2015, the law was unsettled. The department maintained that the public accommodations title reaches the websites of covered entities, stated in filings dated October 3, 2011, and May 15, 2012, but no regulation specified the technical requirements. Federal courts divided on whether a website needs a connection to a physical place of business.

The two lines of judicial reasoning merit a fuller statement, because the division is substantive rather than terminological. The nexus reasoning starts from the statute’s enumeration of twelve categories of public accommodation, observes that every category describes a physical establishment, a restaurant, a theater, a store, a hotel, and concludes that the title’s reach is anchored in physical places. On this reading, a website is covered when it functions as a gateway to the goods and services of a physical place, while businesses with no physical presence present the harder case. The broader reasoning starts from the statute’s prohibition of discrimination in the full and equal enjoyment of goods and services, observes that the categories describe types of businesses rather than types of buildings, and concludes that a covered business is covered in all its operations, including those conducted exclusively online. The two readings cannot be reconciled by clever drafting.

The Netflix litigation that produced the department’s two Statements of Interest illustrates how the website question reached the courts. The National Association of the Deaf sued the streaming service over the absence of captions on its video content, arguing that the service, as a place of public accommodation, discriminated by offering its library on terms that deaf viewers could not equally enjoy. The department’s Statements of Interest, filed October 3, 2011, and May 15, 2012, supported the position that the public accommodations title reaches such services, framing the absence of captions as the digital equivalent of the physical barriers the statute was written to remove. The case settled before producing appellate precedent on the threshold question, which left the department’s position stated but untested at the circuit level in that posture. The department pressed the same position through enforcement agreements as well: a consent decree with H and R Block entered March 25, 2014, a settlement agreement with Peapod dated November 14, 2014, and a settlement with edX dated April 2, 2015.

The July 26, 2010, Advance Notice of Proposed Rulemaking deserves a closer description, because its questions reveal what the department itself considered unresolved. The advance notice asked whether the department should regulate website accessibility under Titles II and III, what technical standard should measure compliance, how the requirements should apply to different kinds of entities, and what the costs and benefits of regulation would be. Those are the questions a completed rulemaking would have answered, and their persistence as questions, rather than answers, through the reference date is the regulatory gap in its entirety. An advance notice imposes no duties. It solicits information for a rulemaking that may or may not follow.

Both sides of the dispute have arguments that deserve equal care, and this article takes no position between them. The accessibility rationale holds that the statute’s promise of full and equal enjoyment is empty if it stops at the building’s edge while commerce, education, and civic life move online; exclusion from a website that is the sole channel for a good or service is exclusion from the good or service, and a civil rights statute that cannot see the screen cannot see the discrimination. The notice and uncertainty objection holds that liability without a technical standard is unfair: the design rules tell a builder exactly how wide the door must be, but no adopted rule told a website operator which technical standard to meet, at what level, or by when, and fair notice requires the government to state the rule before punishing its breach. Each argument answers the other only partially. The accessibility rationale does not supply the missing measurements, and the notice objection does not explain why the statute’s broad language should halt at the network cable. The dispute remained live at the reference date, and readers should understand that it was live: heated, consequential, and unresolved.

The two positions can be stated at their strongest without endorsing either. The accessibility rationale begins from the statute’s text, which promises the full and equal enjoyment of the goods, services, facilities, privileges, advantages, and accommodations of places of public accommodation, and asks what that promise means when the place offers its goods through a website. If the website is the store, excluding a blind customer from the website is excluding the customer from the store’s goods, and the statute’s application to the physical store while ignoring its digital counterpart draws a line the statute’s own language does not draw. The fact that Congress in 1990 did not foresee commercial websites does not, on this view, limit a broadly worded guarantee; the guarantee was written to outlast the technologies of its enactment.

The notice and uncertainty objection begins from the design rules themselves, which give builders exact measurements and give website operators, at the reference date, nothing comparable. A builder knows how wide the door must be because the regulation says so; a website operator knew only that the Department favored coverage, without any adopted rule specifying which technical standard applied, at what conformance level, or on what timeline. Liability under such conditions, the objection runs, punishes businesses for failing to guess the right answer, and the advance notice of July 26, 2010 compounded the problem by signaling that a rule was coming without ever delivering one. Fair notice, on this view, requires the government to state the technical rule before enforcing it, and the Department’s enforcement posture asked courts to do the stating instead.

Each side’s weakness is the other’s strength, which is why the dispute proved so durable. The accessibility rationale supplies the moral and textual force but not the measurements; the notice objection supplies the procedural critique but not an account of why the statute’s broad language should stop at the network cable. The later developments did not resolve the underlying interpretive question so much as change its context: the withdrawal of the advance notice on December 26, 2017 closed the rulemaking path for private websites, while the Title II final rule of April 24, 2024 settled the technical standard for state and local government web content without touching the Title III question. At the reference date, and for years after, the division by reasoning described above was the state of the law.

Two cautions close this section, and both are matters of professional hygiene. First, no federal appellate decision on the website nexus question, and no district court website-accessibility decision, should be cited as in-horizon authority for either side of the division; the reasoning described above stands without them. Second, this article asserts no statistics about the volume of website accessibility filings, because no named source for a stated period supports such a figure here, and the heated character of the dispute is no substitute for counted cases. A practitioner advising a client on website exposure at the reference date could describe the Department’s stated positions with their dates, describe the divided reasoning, and lay out the two arguments with the care shown above. Anything more specific than that, any prediction about which reasoning would govern a particular website in a particular courtroom, would have been speculation dressed as advice, and this article offers none of it.

The federal technology rules that run alongside

The design rules for buildings are not the only federal accessibility standards, and the other set comes from the earlier statute. Section 508 of the Rehabilitation Act of 1973 addresses the accessibility of electronic and information technology developed, procured, maintained, or used by federal agencies. It is a technology procurement and deployment duty, not a building code: it asks whether the software, websites, and electronic systems the government buys and operates can be used by federal employees and members of the public with disabilities. The duty binds federal agencies and, through procurement, the vendors who sell technology to them; it does not bind the private places of public accommodation that Title III governs, and it does not measure doorways.

The comparison with the ADA design rules is instructive because the two regimes share a family resemblance and a different enforcement reality. Both translate a civil rights promise into technical specifications. But where the ADA design rules are enforced by the Justice Department and by private plaintiffs in court, the Section 508 duty runs primarily through agency procurement and administrative complaint processes, a quieter channel that produces less case law and less public drama. The standards themselves are maintained separately from the ADA Standards for Accessible Design, and a product that satisfies one regime has not necessarily satisfied the other. Readers who want the full account of the earlier statute’s technology requirements, and of how Section 504’s program accessibility concept became the template for Title II, should consult the guide to the Rehabilitation Act of 1973, which owns that territory in this series.

Compliance under Section 508 is driven through the federal acquisition process: agencies must procure technology that meets the standards, vendors must warrant that their products do, and the remedy for noncompliance runs through procurement law and administrative complaint processes rather than through the private damages and injunctive actions that enforce the design standards.

The 1998 amendments that gave Section 508 its modern form are worth understanding because they explain the procurement mechanism’s logic. Congress recognized that the federal government is the largest single purchaser of information technology in the country, and that conditioning that purchasing power on accessibility would move the commercial market without regulating private parties directly. The market effect was the point: a vendor that wants to sell to the government builds accessibility into the product, and the accessible product is then available to every other customer.

The Access Board’s role under Section 508 differs structurally from its role under the 1990 statute, and the difference illuminates both. Under the 1990 statute, the board writes minimum guidelines and the Justice Department translates them into enforceable standards through rulemaking, with the consistency requirement binding the two. Under Section 508, the board writes the standards themselves, published as 36 CFR Part 1194, and the standards go directly to the federal agencies that must apply them in procurement. There is no intermediate adoption because there is no intermediate regulator: the board is the standard-setter, the agencies are the regulated community, and the Federal Acquisition Regulation carries the standards into the contracts. The board thus plays two different institutional parts in the two regimes, advisor in one and author in the other, which is another reason practitioners should keep the regimes distinct.

What the standards built

The design rules are easy to discuss as abstractions and harder to picture as outcomes, but their product is visible in every American downtown. Curb ramps at intersections, accessible entrances on new commercial buildings, elevators in mid-rise construction, accessible restrooms in restaurants and theaters, and the gradual retrofitting of older properties as barrier removal duties were reassessed year after year: these are the physical residue of the three tiers doing their work over a quarter century. The existing-facility tier produced the slow accumulation of small changes, one buzzer and one ramp at a time. The alteration tier produced the pattern in which every renovation dragged its path of travel partway toward compliance. The new-construction tier produced the generation of buildings, those first occupied after January 26, 1993, that were simply built accessible from the foundation. The account of what the standards produced measures that residue in detail, and it is the right next stop for readers who want effects rather than mechanisms.

The product is incomplete by design, and saying so is not a criticism of the framework. A duty measured by achievability will always leave some barriers standing, because some barriers genuinely cannot be removed without much difficulty or expense. A proportionality cap will always leave some paths of travel unfinished, because the regulation chose 20 percent rather than 100. The framework’s defenders call this realism: a rule that demanded everything at once would have been repealed or ignored, while a rule that demands what is achievable, what is feasible, and what a new building can bear has survived for decades and rebuilt the country’s physical plant one tier at a time. The framework’s critics call it permission to delay. Both descriptions fit the same facts, which is why the measurement of outcomes belongs to a different article than the explanation of the rules.

Common errors and how the framework corrects them

The brief for this article identified three recurring errors, and each has its correction built into the framework itself. They are worth collecting in one place, because they are the errors that even careful readers make.

The first is treating the safe harbor as grandfathering. The correction is the harbor’s own text: it protects elements that complied with the corresponding 1991 specifications and were not altered on or after March 15, 2012, element by element, and it never protected elements that never complied or the supplemental categories the 1991 rules never addressed. A building is not grandfathered; at most, some of its elements are harbored, and the readily achievable removal duty continues for everything else. The grandfathering claim also appears in the series’ myth-busting article, which takes it up directly.

The second is using undue hardship and undue burden interchangeably. The correction is the titles: undue hardship is the Title I employment defense at 42 U.S.C. 12111(10), measured against the employer’s resources, while undue burden is the Title III auxiliary aids defense at 42 U.S.C. 12182(b)(2)(A)(iii) and the Title II program access defense at 28 CFR 35.150(a)(3), each measured against the resources of the whole entity. Fundamental alteration is the third box, asking about transformation rather than cost. A memorandum that argues cost under the name of fundamental alteration, or that cites 42 U.S.C. 12134, the Title II rulemaking provision, as if it were a defense, has the right facts in the wrong boxes.

The third is assuming a single national answer on websites. The correction is the divided record: the Department’s stated positions with their dates, the advance notice of July 26, 2010 that produced no Title III rule, the withdrawal of December 26, 2017, the Title II final rule of April 24, 2024, and the courts’ division over the physical nexus question, described by reasoning rather than by scorecard. At the reference date there was no single answer to give, and a compliance memorandum that offered one was overstating what the law contained.

Two further errors deserve mention because they follow from the framework’s complexity. One is citing 28 CFR 36.403(f), the Title III path-of-travel provision with its 20 percent cap, for a public entity’s renovation, when Title II’s analog is 28 CFR 35.151(b)(4). The other is designing to the Access Board’s guidelines or the Department’s technical assistance materials as if they were the binding regulation, when only the adopted standards in the Code of Federal Regulations bind. Both errors come from treating one body’s product as another’s, and both are corrected by the three-body structure this article opened with: guidelines originate, the Department decides, and the regulation governs.

The tiers also explain the pattern of what was built when. The first decade after the statute produced the new-construction generation, buildings designed from the start to the 1991 specifications, because developers building new had no alternative. The same decade produced relatively little alteration-driven access, because the renovation cycle had not yet turned over. The second decade, after the 2010 Standards and their transition rules, produced the wave of alteration-triggered upgrades, as renovations began dragging their paths of travel toward the new specifications element by element. And throughout both decades, the existing-facility tier produced the steady background accumulation of barrier removals and alternative methods, reassessed year after year as the continuing duty required. The built environment’s accessibility is thus a layered record of the three tiers operating on different clocks.

Why no checklist can settle it

The belief that compliance is a checklist dies hard, because checklists are what architects and facility managers are trained to use, and the design rules themselves look like a checklist: hundreds of specifications, each with a number and a measurement. The belief is wrong in a way that matters. Two of the three tiers turn on contextual judgments about achievability, feasibility, and resources, which means identical buildings can owe different duties. The profitable national chain and the struggling independent restaurant face the same legal test for the same entrance ramp and will lawfully reach different answers. The renovation that is the maximum extent feasible for one owner, given the structure as it stands, is not the maximum for another. Only the new-construction tier behaves like a checklist, and even there the structural impracticability exception requires judgment about terrain rather than arithmetic about budgets.

The contextual character of the duties has a further consequence that practitioners should internalize: compliance in this field is a program rather than a project. A project has an end date. A program has a cycle. The continuing nature of the readily achievable duty means the barrier removal assessment recurs. The alteration tier means every renovation is a compliance event that must be analyzed before the design is final. The safe harbor means the alteration schedule determines when protected elements migrate to the new edition, which makes the capital planning calendar a compliance document. The defenses mean the entity’s resource picture and its documentation practices are part of its legal posture, not merely its accounting. The facility manager who treats accessibility as a program, with recurring assessments, documented decisions, and renovation-triggered reviews, is aligned with the structure of the duties.

This frustrates readers seeking a definitive answer, and the frustration is legitimate. A business owner wants to know, once and for all, whether the law requires the ramp. The honest answer is that the law requires the removal of the barrier where removal is readily achievable, that achievability is measured against the business as it stands, that the measurement can change, and that the alternative-methods duty sits in reserve. That answer is less satisfying than yes or no, and it is nonetheless the law. The practitioners who thrive under this framework are the ones who build reassessment into their routines: the annual facilities review that re-asks the readily achievable question, the renovation plan that prices the path of travel before the finishes, the new project that treats the design rules as the starting point rather than the punch list. The framework rewards the habit of asking, not the comfort of having asked once.

Studying the framework

The three-tier obligation, the safe harbor with its limits, and the three defenses are standard examinable material in disability law and administrative law coursework, which is why this article has kept the doctrines separated the way an examination would. Two companion tools accompany this material: a legislation study notebook and a civics study guide. The doctrines repay the kind of study that distinguishes rather than memorizes: which defense belongs to which title, which citation belongs to which defense, and which tier asks about achievability and which asks about feasibility.

The gap between text and practice

The series thesis follows this article everywhere: the gap between statutory text and delivered practice, filled here by technical standards written by a body most readers have never heard of. Congress wrote a sentence about nondiscrimination. The Access Board wrote guidelines measured in inches. The Department of Justice wrote enforceable rules through rulemaking. Private plaintiffs and the Department wrote the enforcement record case by case. At no point in that chain did the statute’s text, standing alone, tell a builder what to build. The law as delivered is the product of all five mechanisms this article has described, and the law as written is only the first of them.

That is worth stating plainly because it reverses the usual assumption about how legislation works. The assumption is that Congress decides and agencies implement, with the real choices made on the floor of the House and Senate. In this corner of disability law, Congress decided the principle and delegated the measurements, and the measurements are where the compliance questions live. The three-tier obligation, the safe harbor, the three defenses, and the unfinished website rulemaking are all creatures of regulation and interpretation rather than of statutory text. A reader who can explain the three tiers, distinguish the three defenses, state the safe harbor’s limits, and describe the website division with its dates has understood not just this article’s subject but the series’ recurring lesson: the statute is the beginning of the law, and the standards are where the law happens.

The framework’s durability is the final evidence for the thesis. Three decades after the statute, the three-tier obligation still sorts every project, the safe harbor still manages the transition between generations of rules, and the three defenses still excuse the duties they were written to excuse. The website question remains the conspicuous exception, the place where the rulemaking never finished and the courts divided. That exception proves the rule about the rest: where the standards exist, they settle the compliance questions the statute left open, and where they do not exist, the questions stay open. The gap between text and practice is filled, in this corner of the law, by technical standards written by a body most readers have never heard of, adopted by a Department acting through rulemaking, and enforced by the Department together with private plaintiffs. That is the whole of the mechanism, and it has held.

The framework’s treatment of existing buildings also reveals a legislative judgment about the pace of change. Congress could have required immediate full compliance from every existing facility, and the resulting economic shock would have made the statute politically impossible. It could have exempted existing buildings entirely, and the statute would have changed nothing for a generation. The readily achievable standard is the middle path: a duty that bites immediately where the bite is small and grows as circumstances allow. The continuing reassessment is what makes the middle path honest rather than merely convenient, because a duty fixed forever at the first year’s assessment would decay into an exemption. The framework thus encodes a theory of progress as well as a set of rules: accessibility expands as achievability expands, and the law’s reach grows with the economy’s capacity to bear it.

That theory also explains why the framework has survived without major structural revision. The tiers are self-adjusting: the readily achievable duty tightens as costs fall and resources rise, the alteration duty captures the renovation cycle’s natural turnover, and the new-construction duty ensures that the building stock of the future is born compliant. A critic who wants faster progress can argue that the adjustment is too slow; a critic who wants less burden can argue that the reassessment never ends. Both criticisms are really arguments about the speed of the self-adjustment, not about its direction. The framework’s answer to both is the same: the standards state the destination, the tiers set the pace, and the defenses mark the boundaries.

Frequently Asked Questions

Q: What are the ADA Standards for Accessible Design?

They are the enforceable technical rules that translate the disability statute’s nondiscrimination promise into measurements: door widths, ramp slopes, counter heights, signage specifications, and hundreds of similar requirements. The Department of Justice adopts them through notice and comment rulemaking for Title II, which governs state and local government, and Title III, which governs places of public accommodation. The technical content originates as minimum guidelines from the Access Board, an independent federal agency, and becomes binding only when the Department adopts it. Two generations exist: the 1991 rules, adopted July 26, 1991, and the 2010 Standards, published September 15, 2010 with compliance required from March 15, 2012. The statute states the principle; the standards state the numbers, and a court applies the numbers.

Q: What is readily achievable barrier removal under the ADA standards?

It is the Title III duty of places of public accommodation to remove architectural barriers, and structural communication barriers, in existing facilities where removal is “easily accomplishable and able to be carried out without much difficulty or expense,” 28 CFR 36.304(a). The test turns on the circumstances of the particular business, so the same ramp can be readily achievable for a large national chain and not for an independent shop. The duty is continuing rather than one-time: it is reassessed as finances and technology change, so a modification declined as too costly in one year may be owed in another. Where removal itself is not readily achievable, the entity must still use alternative methods of providing access where those methods are readily achievable, such as curbside service or a call buzzer at an inaccessible entrance.

Q: Do older buildings have to meet the ADA standards?

Yes, within limits, because the statute contains no exemption for buildings constructed before 1990. An older building that is a place of public accommodation must remove barriers where removal is readily achievable, a continuing duty measured against the business’s resources and reassessed over time. An older building that satisfied the 1991 design rules gains a limited safe harbor against the 2010 rules for the elements that actually complied, until those elements are altered on or after March 15, 2012. But the safe harbor never covers elements that never satisfied any earlier rule, and it never covers the supplemental categories the 1991 rules did not address, so age alone excuses nothing. State and local government buildings are measured instead by Title II’s program accessibility standard, which asks whether each program viewed in its entirety is accessible.

Q: Is there a safe harbor in the ADA standards?

Yes, but it is a limited transition rule rather than a grandfather clause. Under 28 CFR 36.304(d)(2)(i), elements in existing facilities that have not been altered on or after March 15, 2012, and that comply with the corresponding technical and scoping specifications for those elements in the 1991 Standards, need not be modified to meet the 2010 Standards. The protection is element-by-element: a compliant entrance does not protect a noncompliant restroom in the same building. It does not apply to supplemental elements the 1991 rules never specified, such as amusement rides, play areas, and pools, which must be modified to the extent readily achievable. Title II has its own version at 28 CFR 35.150(b)(2). Because the readily achievable removal duty continues for every element that never complied with anything, the safe harbor excuses upgrades, not barriers.

Q: Do websites have to follow the ADA standards?

At the reference date of September 15, 2015, no adopted design rule answered that question for private websites. The Department of Justice had long taken the position that Title III applies to the websites of places of public accommodation, and it issued an advance notice of proposed rulemaking on website accessibility on July 26, 2010, but no final Title III rule had issued. Two later developments carry explicit dates: the advance notice was withdrawn on December 26, 2017 without a rule, and a Title II final rule adopting a recognized technical standard for state and local government web content issued on April 24, 2024. The federal courts were divided on whether a website must have a nexus to a physical place of business, with one line of reasoning treating the website as itself covered and the other requiring the physical connection. The Department filed Statements of Interest in the National Association of the Deaf litigation against Netflix on October 3, 2011 and May 15, 2012, stating its interpretation. No individualized compliance guidance follows from this divided record.

Q: Who writes the ADA standards?

Three actors share the work by statutory design. The Architectural and Transportation Barriers Compliance Board, an independent federal agency usually called the Access Board, develops the minimum accessibility guidelines under 42 U.S.C. 12204. The Department of Justice converts those guidelines into enforceable design standards through notice and comment rulemaking, and the statute requires the Department’s standards to be consistent with the Board’s minimum guidelines, 42 U.S.C. 12186(c), while leaving the Attorney General solely responsible for promulgating, interpreting, and enforcing them. Enforcement then runs through the Department, which may investigate and sue, and through private plaintiffs, who hold their own right of action for injunctive relief. The body that writes the technical content therefore does not enforce it, and the body that enforces it did not originate the measurements.

Q: What is the difference between undue burden and fundamental alteration in the ADA standards?

Undue burden is about cost and administrative difficulty measured against resources. Under Title III it excuses the provision of auxiliary aids and services where the entity demonstrates that taking the steps “would result in an undue burden,” 42 U.S.C. 12182(b)(2)(A)(iii), and under Title II a public entity need not take a program access step that would cause “undue financial and administrative burdens,” 28 CFR 35.150(a)(3). A well funded entity can owe more than a struggling one for the same step. Fundamental alteration is about identity rather than expense: it asks whether the requested change would transform the nature of the good, service, or program itself, regardless of cost, under 42 U.S.C. 12182(b)(2)(A)(ii) for Title III and 28 CFR 35.130(b)(7)(i) for Title II. Installing a ramp at great expense raises undue burden; rewriting an essential eligibility rule raises fundamental alteration. Cost arguments belong in the first box and transformation arguments in the second.

Q: What are Section 508 standards compared to the ADA standards?

Section 508 of the Rehabilitation Act of 1973 governs the accessibility of electronic and information technology developed, procured, maintained, or used by federal agencies, a technology duty that reaches software, systems, and agency websites rather than buildings. The ADA Standards for Accessible Design govern the physical design of places of public accommodation under Title III and public entity facilities under Title II, specifying doorways, ramps, restrooms, and similar elements. The two regimes share a family resemblance, since both convert a civil rights promise into technical specifications, but they bind different actors, cover different things, and are enforced through different channels: agency procurement and administrative processes for Section 508, Department enforcement plus private lawsuits for the ADA design rules. Satisfying one regime does not satisfy the other.

Q: How does the 20 percent path-of-travel limit work in practice?

When an alteration affects the usability of or access to a primary function area, such as a restaurant dining room or a store sales floor, the entity must also make the path of travel to that area accessible, including the entrance, the route, and the restrooms, telephones, and drinking fountains serving it, 28 CFR 36.403. Path-of-travel spending is capped: it “will be deemed disproportionate to the overall alteration when the cost exceeds 20% of the cost of the alteration to the primary function area,” 28 CFR 36.403(f)(1). A 100,000 dollar dining room renovation therefore carries up to 20,000 dollars of path-of-travel obligation before the cap binds. The figure is a ceiling on the supplementary obligation, not a budget for the whole project, and spending up to the cap remains mandatory where the trigger is met and the funds can buy access improvements. The statute frames the duty as applying where the work is “not disproportionate to the overall alterations in terms of cost and scope,” 42 U.S.C. 12183(a)(2). The cap is Title III’s rule; public entities answer instead to 28 CFR 35.151(b)(4).

Q: What does maximum extent feasible mean for alterations?

It means the altered portions of a facility must meet the design rules as fully as possible, a stricter test than the readily achievable standard for existing buildings. The statute commands compliance “to the maximum extent feasible,” 42 U.S.C. 12183(a)(2), and the regulation applies the phrase where full compliance is “virtually impossible,” requiring then “the maximum physical accessibility feasible,” 28 CFR 36.402(c). Cost alone does not make compliance infeasible; the question is what the structure as it stands permits, not what the owner prefers to spend. A renovation that could have met the specifications but used cheaper noncompliant finishes fails this test even though the same owner might lawfully decline the identical work in an unaltered building under the readily achievable standard. The alteration tier is a duty to succeed where the existing-facility tier is a duty to try.

Q: What is the structural impracticability exception for new construction?

It is the sole exception to the new-construction tier’s demand for full compliance, and it is construed narrowly. New facilities designed and constructed for first occupancy after January 26, 1993 must be readily accessible to and usable by individuals with disabilities, 42 U.S.C. 12183(a)(1), with no readily achievable qualifier and no proportionality cap. The regulation permits departure from the specifications only where compliance is structurally impracticable, 28 CFR 36.401(c), meaning the unique characteristics of the terrain prevent incorporating the accessibility features. A steep hillside that makes a level entrance physically impossible may qualify; a flat site where the developer preferred a grand staircase, or where compliance cost more than budgeted, does not. The exception is about the land, not the budget, and it succeeds rarely, which is exactly the point: the new-construction tier is the one place where the framework refuses to negotiate.

Q: Which elements get no safe harbor under the 2010 Standards?

The supplemental elements listed in 28 CFR 36.304(d)(2)(iii), the categories the 2010 Standards addressed for the first time and the 1991 Standards never specified. They include residential facilities and dwelling units, amusement rides, recreational boating facilities, exercise machines and equipment, fishing piers and platforms, golf courses and miniature golf, play areas, saunas and steam rooms, swimming pools and spas, and shooting facilities with firing positions. Because there was no earlier specification for these elements to satisfy, there is nothing for the safe harbor to protect, and they must be modified to the extent readily achievable to meet the 2010 Standards. An amusement park in perfect compliance with the 1991 rules still owes the 2010 rules for its rides. This is the sharpest illustration of the harbor’s limits: it preserves past compliance, and only past compliance, element by element.

Q: How does Title II treat existing facilities differently from Title III?

Title II does not use “readily achievable” at all. Instead it applies the program accessibility standard: each service, program, or activity of a state or local government, “when viewed in its entirety,” must be accessible to individuals with disabilities, 28 CFR 35.150(a). A public entity need not make every existing facility accessible; it can comply by relocating a program to an accessible building, providing home visits, or using other methods that make the program as a whole usable. Title III gives private places of public accommodation no such aggregate option: each existing facility must remove its own barriers where readily achievable. The two titles also diverge on new construction and alterations, which Title II governs separately at 28 CFR 35.151, including its own path-of-travel disproportionality rule. Telling a city it must remove barriers where readily achievable, or telling a restaurant it need only make its program accessible in the aggregate, cites the wrong title.

Q: How is undue hardship measured against an employer’s resources?

Undue hardship is the Title I employment defense, and it is measured employer by employer. The statute defines it as “an action requiring significant difficulty or expense, when considered in light of the factors set forth in subparagraph (B),” 42 U.S.C. 12111(10), and the factors examine the nature and cost of the accommodation, the financial resources of the facility and its headcount, the effect of the accommodation on the facility’s operation, and the resources, size, and operations of the covered entity as a whole. The same accommodation can therefore be reasonable for a large employer and an undue hardship for a small one, by statutory design. The defense never leaves the employment title: it does not excuse Title III barrier removal or Title II program access decisions, and invoking it there cites the wrong law. Employers should document the factors rather than asserting hardship as a conclusion, since courts examine the analysis, not the label.

Q: What spending order applies when the path-of-travel cap binds?

When the 20 percent cap limits path-of-travel spending, the regulation prescribes the order in which the available funds must be spent, 28 CFR 36.403(f)(2). The entrance comes first, then the accessible route to the altered area, then at least one restroom for each sex or a single unisex restroom, then telephones, then drinking fountains, and then other elements. The owner spends down the list until the capped amount is exhausted, which means a modest renovation may produce an accessible entrance and route but no upgraded restroom, while a larger project works further down the priority order. The sequence reflects a judgment about what matters most to a person arriving at the building: getting in, getting to the altered area, and then the supporting elements. Advisory guidance accompanying the regulation elaborates the same ordering for designers working through a constrained budget.

Q: Can a business owe the same barrier removal in a later year after declining it?

Yes, and this is the feature of the readily achievable duty that most frustrates owners seeking a final answer. The duty is continuing: it is reassessed as circumstances change rather than decided once and filed away. A modification properly declined as too costly in one year may become readily achievable in another because revenues rose, because the cost of the work fell, because a planned renovation made the work cheaper to perform alongside it, or because the business grew and its resources grew with it. The compliance question is never finally closed, which is why the article’s central claim holds that readily achievable is not a budget test. Prudent facility managers build the reassessment into an annual review, documenting the analysis each time, so that the record shows a living judgment rather than a single decision aging into noncompliance.

Q: When did the 2010 Standards take effect, and which dates matter?

The final rule was published September 15, 2010, became effective March 15, 2011, and required compliance beginning March 15, 2012. Between publication and the compliance date, builders could use either the 1991 or the 2010 Standards in full, but not a mixture. Compliance for newly constructed and altered swimming pools, wading pools, and spas was extended to January 31, 2013 by a rule published May 21, 2012. March 15, 2012 is also the safe harbor’s trigger date: only elements unaltered on or after that date keep the protection of 28 CFR 36.304(d)(2)(i), and Title II’s path-of-travel harbor at 28 CFR 35.151(b)(4)(ii)(C) similarly protects work built or altered to the earlier rules before that date. A compliance calendar organized around these dates, rather than around the statute’s 1990 enactment, is what keeps a facility on the right side of the transition.

Q: Why do federal courts disagree about websites and physical locations?

The disagreement is about what the public accommodations title covers, described here by reasoning rather than by case names. One line of reasoning treats a website as itself a place of public accommodation, or as the means by which such a place offers its goods and services, and concludes that the statute’s duties apply to it directly; on this view the absence of a physical threshold does not remove an online transaction from the law. The other line of reasoning requires a nexus between the website and a physical place of business, reading the title’s list of covered establishments as describing physical places that a purely virtual operation does not resemble. The two readings produce different answers for the same website in different courtrooms, which is why the question remained the field’s highest volume dispute at the reference date. The Department’s longstanding position favored coverage, expressed through rulemaking notices and litigation filings, but no adopted Title III technical rule settled the measurements.

Q: Who must comply with Section 508, and what does it cover?

Section 508 of the Rehabilitation Act of 1973 binds federal agencies, requiring that electronic and information technology they develop, procure, maintain, or use be accessible to federal employees and members of the public with disabilities. Through procurement, the duty reaches the vendors who sell technology to the government, making accessibility a condition of federal contracts. It covers software, systems, and electronic content rather than buildings, and it is enforced through agency procurement and administrative complaint processes rather than through the private lawsuits that enforce the ADA’s design rules. Private places of public accommodation are not directly bound by Section 508, and federal agencies are not directly bound by the ADA’s Title III facility rules. The two regimes are cousins with separate standards, separate enforcers, and separate case law.

Q: What distinguishes an alteration from ordinary maintenance?

The regulations attach the alteration tier to work that affects usability, which is the practical dividing line. A renovation that changes how a space is used or reached, such as reconfiguring a lobby, replacing a staircase, or remodeling restrooms, triggers the maximum-extent-feasible duty for the altered portions and can trigger the path-of-travel obligation where a primary function area is affected, 28 CFR 36.402 and 36.403. Routine upkeep that changes nothing about usability, such as repainting, replacing carpet with equivalent material, or servicing mechanical systems, does not trigger those duties. The distinction matters for project planning: labeling a project as maintenance does not control the legal question, and a phased “maintenance” program that cumulatively rebuilds a space will be judged by what it did to usability rather than by what it was called.