Employment Rates and Wage Disparities

The results of studies both on what happened to the wage and employment rates of people with disabilities in the decade following the passage of the ADA and whether the ADA was the cause of the various changes that they identified differ (Hotchkiss, 2003). There is consensus that the employment rate declined in the 1990s for people reporting that they had conditions that limit their ability to work, but whether this trend extended to wages or to people more likely to be protected by Title I is still debated. Disagreement starts with whether the most commonly used data sets properly define disability and also extends to more arcane points of study methodology. Analysis is complicated by the changing judicial interpretation of “disability” over time and the fact that in many states people with disabilities had substantial protection from discrimination even before the ADA was passed. These issues are fully discussed elsewhere (Blanck et al., 2003; Kruse and Schur, 2003; Stapleton and Burkhauser, 2003). In this section, we briefly summarize the findings and most significant methodological issues.

Most studies of employment rates use one of three nationally representative data sets: the Current Population Survey (CPS), the National Health Interview Survey, and the Survey of Income and Program Participation (SIPP), all of which define “disability” primarily in terms of a self-reported health condition that limits or entirely prevents an individual from working. The researchers analyzing these data generally agree that the rate of employment for working-aged adults with disabilities declined during the 1990s, not only absolutely but also relative to the rate for those without disabilities (Acemoglu and Angrist, 2001; Burkhauser et al., 2003; DeLeire, 2000; Houtenville and Burkhauser, 2004; Kruse and Schur, 2003; Moon and Shin, 2006; Stapleton and Burkhauser, 2003).

Fewer studies have looked at changes in wages over time. DeLeire analyzed SIPP data for men between 1986 and 1995 and found no significant declines (DeLeire, 2000). Acemogolu and Angrist, using similar methods, analyzed CPS data for men and women between the ages of 21 and 58 years over the same time period and reached the same conclusion (Acemoglu and Angrist, 2001). Moon and Shin, using SIPP data for men between 1990 and 1992, found that the logarithm of the real wages of people with disabilities had declined relative to those of men without disabilities (Moon and Shin, 2006).

Title I’s definition of disability excludes people who cannot work (or who require more than a “reasonable” accommodation to do so), and it covers people who have no serious impairment or work limitations but are mistakenly treated as if they had. Thus the employment rate of people with serious work limitations is a flawed indicator of the statute’s effect on those whom it was designed to help. A set of questions on the SIPP about functional and activity limitations that do not prevent work allowed Kruse and Schur to investigate changes in a segment of the disabled population that more closely approximates that covered by Title I (Kruse and Schur, 2003). They found that employment rates did decrease in the 1990s for individuals reporting a work disability but that employment rates actually increased for those with functional and activity limitations that do not prevent work (Kruse and Schur, 2003). Disagreement about the validity of studies based on a work-limitation definition of disability persists, even as the Bureau of Labor Statistics works to craft a new definition more compatible with the ADA for use in the CPS (Burkhauser et al., 2002; McMenamin et al., 2006; National Council on Disability, 2004a).

Researchers have tested many explanations for the employment outcomes that they found. They have studied the importance of demographic factors and education (Houtenville and Daly, 2003), changes in the nature of work or in the job market (Stapleton et al., 2003), the changing size and composition of the disabled population (Kaye, 2003), changes in the costs of health care and modes of health care financing (Goodman and Waidmann, 2003; Hill et al., 2003; Yelowitz, 1998), and expansions of the Social Security Disability Income (SSDI) and Supplemental Security Income (SSI) programs, including both lowering of the eligibility requirements and increases in benefits for some recipients (Burkhauser et al., 2001; Houtenville and Burkhauser, 2004). Reviewing these data, two leading researchers concluded that SSDI and SSI expansion had played a more significant role than any other factor, including the ADA (Stapleton and Burkhauser, 2003).

Table E-1 lists the studies that have focused on the effect of the ADA on employment among people with disabilities. Most used the work-limitation definition of disability and depended primarily on the temporal associa tion between the date of passage (1990) or the effective date (1992) of the ADA’s Title I and employment changes to prove causation. Using both simple pre-post and year-by-year analyses, Moon and Shin (2006), DeLeire (2000), and Acemogolu and Angrist (2001) all found robust declines in employment rates: Moon and Shin on the order of 6 percent and DeLeire of 7 percent. Acemoglu and Angrist reported sharp drops in annual weeks of employment for men with disabilities aged 21 to 58 years and women with disabilities under age 40 years. Moon and Shin also found a decline in the logarithm of the real wages of men with disabilities of 5.3 percent relative to that of men without disabilities, although it was significant only at the 0.1 level. Econometric modeling led both DeLeire and Acemoglu and Angrist to attribute the declines specifically to the reasonable accommodation requirement of Title I; Acemoglu and Angrist thought that the reduction in the rates of employment among disabled individuals also reflected employers’ expectation of increased lawsuit costs.

TABLE E-1

Studies of the Effect of the ADA on Employment Rates and Wages.

The use of a different definition of disability and different time periods leads to different results. Kruse and Schur used 14 different measures of disability from the SIPP that represented permutations along three dimensions (activity limitations, receipt of disability income, and reported ability to work) to facilitate comparisons between people more and less likely to be covered by the ADA (Kruse and Schur, 2003). They found that the subgroup of people most likely to be covered by Title I saw an improvement in employment rates, while for others employment declined.

Houtenville and Burkhauser (2004) replicated the study of Acemoglu and Angrist (2001), testing the sensitivities of different definitions of disability, as well as the effects of different periods of time. They confirmed that the employment rate for people with a work limitation did indeed decline when the rate was measured by the annual number of weeks that an individual worked during one calendar year. The use of a 2-year time period, however, suggested that employment began to decline in the mid-1980s and actually sharply improved in 1992 for some age-sex categories. They concluded that there was “little evidence of a negative effect of the ADA on the population with longer term disabilities and some evidence of a positive effect of the ADA” (Houtenville and Burkhauser, 2004, p. 7).

All the studies that have attempted to test the impact of the ADA by temporal association with national trends in employment and wages suffer from the same flaw: many states had laws against discrimination on the basis of disability before the ADA came into effect, so we cannot assume that the implementation of Title I represented a change in the rules for all employers and workers. Two studies specified state-law variables that made it possible to compare the rates of employment and wages for people with disabilities in states with and without various ADA-like disability protections at the same points in time.

Beegle and Stock compared state-level employment conditions for people with disabilities at three points, in 1970, 1980, and 1990. Because all three times were before the time of the implementation of the ADA, it should be understood that the study used similar state laws as a proxy for the federal statute (Beegle and Stock, 2003). They found that disability discrimination laws were associated with lower relative earnings among disabled individuals and with slightly lower rates of relative labor force participation rates among disabled individuals. However, once they controlled for differential time trends in employment among disabled and nondisabled individuals, there was no systematic negative relationship between the laws and the relative employment rates of disabled individuals. Of course, this also suggests that there was no substantial positive impact of state disability discrimination laws.

Jolls and Prescott studied states with three different legal conditions between 1990 and 1993: states with no protections comparable to the ADA, in which the ADA would represent an entirely new influence; “protection without accommodation” states, whose law prohibited employment discrimination against people with disabilities but did not require reasonable accommodation; and “ADA-like” states, whose protections for the employment rights of people with disabilities both prohibited discrimination and required reasonable accommodation (Jolls and Prescott, 2004). They found that employment rates for disabled individuals began to decline in 1993 relative to pre-ADA levels in all three categories of states. The extent to which the ADA was new law made no significant difference, making it difficult as a general matter to attribute the declines to employers’ reaction to new legal requirements.

The design of the study of Jolls and Prescott (2004) allowed them to test the effects of specific ADA mandates. They found that the employment rates of individuals with disabilities fell 10 percent in the early days of the ADA in “protection without accommodation” states compared with that in states that already had full “ADA-like” statutes, indicating that the reasonable accommodation requirement had an independent negative effect on employment. Yet they also found little or no differences between pre- and post-ADA employment rates among people with disabilities in states that had had no protection before, indicating that the ADA did not have a significant impact in states where it brought entirely new mandates. Considering other important confounding variables—including the size of the employer covered, differences between states in eligibility for disability benefits and the amounts of disability benefits, states’ economic environments, and preexisting state group-specific trends in employment rates among disabled individuals—Jolls and Prescott concluded that apart from a short-term effect of the new reasonable accommodation requirement, there was no link between the ADA and the employment declines experienced by people with disabilities starting in 1993 and continuing forward (Jolls and Prescott, 2004).

Employer Practices and Attitudes

The ADA requires employers to eliminate discriminatory practices and promote equal opportunity by making reasonable accommodations for workers with disabilities. We can assess the impact of the ADA on employ ment by asking what employers have done to comply with the law and whether there is evidence that the attitudes of employers to people with disabilities have improved. Existing data show a high level of awareness of the law and substantial compliance activity. They do not shed much light on the sincerity or the effectiveness of employer efforts and suggest that negative attitudes toward disabled people persist.

At the time of enactment, many employers were uninformed about the ADA and had significant concerns about the costs that it was going to impose. This seems to have changed rather quickly. More durable has been the negative attitudes about people with disabilities, especially those with certain types of disabilities (Bruyere et al., 2006; Greenwood and Johnson, 1987). Employers feel more positively about people with physical or sensory disabilities than they do about people with psychiatric or cognitive disabilities (Baldwin, 1992; Scheid, 2005). There is less acceptance of people whose disabilities are perceived as having been caused by factors under their control than for “innocent” victims (for example, there is less acceptance of a person paralyzed in an accident caused by her own drinking than of a person paralyzed by the actions of a drunken stranger) (Bruyere et al., 2006; Hazer and Bedell, 2000; Mitchell and Kovera, 2006).

Given the poor empirical correlation between attitudes about disabled people and actual employer behavior (Bruyere et al., 2006), attitude studies may be more suggestive of the complexity of workplace decision making and the need for further research than conclusive about the employer response to the ADA. Organizations have historically made changes in policies, procedures, and organizational structures in response to a new national civil rights law (Edelman and Suchman, 1997). Organizational compliance actions create the environment in which managers make decisions about hiring, promotion, and accommodation. Perceptions of organizational adherence to the ADA have been found to be a better predictor of how managers put the ADA into practice than their personal attitudes about people with disabilities (Thakker and Solomon, 1999). Larger employers typically respond to antidiscrimination laws by creating policies and internal equal employment opportunity offices to help the organization draw the line between legal and illegal behavior and to minimize and resolve discrimination disputes. In a 1998 survey of human resource managers, 72 percent reported that their companies had formal procedures for requesting reasonable accommodations and presenting grievances (Bruyere, 2000). Employers too small to support a separate human resources or equal employment opportunity staff may respond to antidiscrimination law in less formal but still important ways (Barnes and Burke, 2006).

Aside from a number of small surveys, qualitative studies, and dissertations (Blackburn, 2003; Harlan and Robert, 1998; Unger, 1999), the best evidence on the organizational response to the ADA comes from a series of large probability surveys of human resources managers conducted by Bruyere and colleagues at Cornell University (Bruyere et al., 2006). General awareness of the law was high, with a large majority (80 to 90 percent) of both federal and private respondents reporting that they had received training in specific elements of the law, particularly reasonable accommodation and nondiscrimination in hiring (Bruyere, 2000).

Title I puts considerable emphasis on reasonable accommodation as an instrument for workplace inclusion, making the extent to which employers provide accommodations a good measure of compliance. Bruyere’s surveys tracked 10 types of accommodations, including changes in accessibility, transportation, supervisory methods, and job requirements. The research found overall that most employers were making accommodations of all types, but there were important variations. As an employer, the federal government was substantially more likely to report the provision of accommodations of virtually every kind than private employers (Bruyere, 2000), and large employers reported a greater use of accommodations than smaller ones (Bruyere et al., 2006). The differences between federal and private employers and between small and large employers were largely accounted for by those that reported that they had never been asked for an accommodation but may also reflect a greater emphasis on compliance within the federal government. A lack of requests for accommodation could indicate an unsupportive environment, employee unfamiliarity with their rights, or the use of informal, undocumented accommodations in smaller workplaces. For all employers, physical accessibility, workplace or tool reconfiguration, and policy change accommodations were more common than changes in training, supervisory methods, or job structure (Bruyere, 2000).

Most federal government (95 percent) and private (82 percent) employers reported that they had made facilities more accessible. There were sharper differences in other areas. Government employers were twice as likely to report providing communication access to the hearing impaired (91 percent versus 43 percent) or visually impaired (77 percent versus 37 percent). Less than half of employers, federal or private, had provided flexible test-taking procedures or a scent-free environment, but most employers that did not provide these accommodations reported that no one had ever asked for them (Bruyere, 2000). A similar pattern and explanation emerged in studies comparing large and small private employers, with smaller employers consistently less likely to report various compliance activities (Bruyere et al., 2006).

The ADA requires a large number of changes in routine procedures that can also be used to mark employer compliance. Brueyere tracked 10 indicators in the hiring process, ranging from changing where recruiting was conducted to overcoming communication barriers with technology or interpreters and complying with restrictions on obtaining information about medical examinations and medical histories. As with the provision of reasonable accommodations after a disabled individual is hired, employers were much more likely to make what they reported to be “easy” changes, like ensuring physical accessibility in the interview site, than to take on communication barriers. Again, the federal government as an employer was much more likely to report providing these accommodations than private employers, apparently because private employers were less aware of how to overcome communication barriers.

A survey focusing on the private-sector use of adaptive information technologies found that less than half of the respondents reported that their organizations had experience in modifying a computer to make it accessible to an employee with a disability (Bruyere et al., 2005). Respondents employed by large organizations were more likely than those employed by small organizations to report familiarity with assistive technology and accessible web designs and to report that their organizations had made modifications to computers and adaptations to workstations. In general, however, there was a low level of familiarity with assistive technology. Almost half (46 percent) were familiar with screen magnifiers, approximately one-third were familiar with speech recognition software, about one-fourth were familiar with video captioning, about 20 percent were familiar with braille readers/displays, about 16 percent were familiar with screen readers, and only 13 percent were familiar with accessible web design.

These studies leave little doubt that employers are generally aware of the ADA and have taken steps to meet its requirements. They show that employers are making accommodations in hiring processes and in the workplace. The studies do not allow confident conclusions to be made about how well these measures are being implemented and the extent to which people with disabilities have benefited or perceived any benefits. Bruyere’s respondents report that negative attitudes and a lack of information still stand as barriers to the hiring and retention of workers with disabilities and that smaller employers may need more help to understand and implement the law’s requirements. More research is also recommended (Bruyere et al., 2006).

Empowerment Effects

The ADA could have a positive impact by empowering workers with more tools for achieving their goals or vindicating their rights. Filing an ADA complaint is one way in which people can use the law to defend their rights. More than 200,000 workers filed discrimination claims under ADA in the decade after the statute was passed (Moss et al., 2005), but research on how people “use” the law would predict that only a small minority of people with disabilities will invoke the ADA to deal with discrimination. People often use a language of rights to define their social goals and position, and disadvantaged groups and their advocates often see the defining of rights as a policy solution to social problems. At the same time, research has consistently shown that Americans are loathe to sue (Felstiner et al., 1980-1981; Galanter, 1983) and that civil rights laws are among the least often invoked (Curran and Spalding, 1974; Engel and Munger, 2003). Research on the ADA suggests that people with disabilities conform to this pattern of law avoidance not simply because they are unaware of the law or are unable to get a lawyer but because filing a complaint implicates a variety of deep social and psychological issues (Engel and Munger, 2001; Moss et al., 2005; Studdert, 2002).

Just as organizations may comply with the law without active enforcement, individuals may “rely” on the law in their own strategies for daily living without formally invoking it (Burris and Moss, 2000). An important study by Engel and Munger explored this in the context of the ADA. Detailed life-history interviews with 60 people with a variety of disabilities found at least three ways in which the ADA was changing their lives, apart from litigation:

First, rights can change the self-perceptions of individuals with disabilities, enabling them to envision more ambitious career paths by incorporating in their plans the reasonable accommodations and the nondiscriminatory treatment guaranteed by the ADA…. Second, ADA rights become active through cultural and discursive shifts even when rights do not directly transform an individual’s self-perceptions. By becoming part of everyday speech, thought, and action, ADA rights affect the way others perceive individuals with disabilities as employees…. Third, ADA rights may become active through institutional transformations that are not directed at any particular individual…. [R]ights are sometimes implemented unilaterally by … employers, rather than through advocacy by the rights-bearers themselves (Engel and Munger, 2003, pp. 243–244).

While these “empowerment effects” are difficult to quantify, they must be seen as among the most important forms of impact that a civil rights law can have. Although the Engel and Munger study is qualitative, it offers intriguing suggestions that the ADA may be working through these mechanisms. Bruyere’s work and some unpublished studies have found signs that people with disabilities are reluctant to ask for accommodations (Baldridge, 2002; Bruyere, 2000; Frank, 2004). Further research along these lines would be valuable.

Title I in the Courts: Narrowing the Protected Class

Title I did not purport to provide legal protection of employment opportunity to everyone with a disability but does provide legal protection of employment opportunity to those who meet its definition of a “qualified person with a disability.” Many proponents of the ADA have argued that Congress intended the definition to be liberally applied, as it had been under the earlier Rehabilitation Act, so that even substantial accommodations would be deemed “reasonable” to encourage the inclusion of people with disabilities who have low levels of functioning and so that “impairment” and “major life activity” would be liberally construed to ensure that qualified people did not suffer employment discrimination simply because of prejudice, fear, or outdated stereotypes (Burgdorf, 1991 Burgdorf, 1997; Feldblum, 2000). Over the last 15 years, however, the U.S. Supreme Court has led the federal courts in interpreting Title I narrowly (National Council on Disability, 2004b). We summarize the main rulings and then review the data on case outcomes.

The Title I Enforcement System: Implementation Problems

Title I is enforced by the same agencies and under the same procedures as the nation’s other employment discrimination laws. People who believe that they have been discriminated against in employment on the basis of a disability may file an administrative charge with the EEOC or a state or local FEPA that contracts with the EEOC, which initiates an administrative dispute resolution process. If the administrative process fails to produce a satisfactory result, the worker can file an ADA lawsuit in state or federal court. The effectiveness of this enforcement system has been weakened by chronic underfunding of the EEOC in the face of ever growing workloads (Moss et al., 2001).

Countervailing Policies

For people whose disabilities are severe enough to place them on the margins of employability, the effects of Title I may be blunted by other elements of national disability policy. Expansion of eligibility for federal Social Security programs² made it easier for marginally employable people to get income and health benefits, a phenomenon that some believe was the main driver of the declining employment rates among the disabled in the 1990s (Stapleton and Burkhauser, 2003). Eligibility for these programs required leaving the workforce and proving that one was unable to engage in any substantial gainful activity. Historically, people receiving benefits under these programs encountered substantial disincentives to going back to work. Workers returning to low-wage or part-time work faced the prospect of earning less than they received in benefits. Because Medicare/Medicaid benefits were tied to social security disability eligibility, going back to work could also mean a gap in or total loss of health insurance (Stapleton et al., 2005). Court decisions under the ADA added another wrinkle: Defendants in discrimination cases may use an individual’s application for or acceptance of disability benefits as evidence that the individual is not “otherwise qualified” to work and therefore not protected from discrimination (Cleveland v. Policy Management Systems Corp., 1999).

The government has acted in recent years to reshape Social Security to encourage beneficiaries to return to work and assist them with doing so. Legislation has expanded eligibility to programs that allow people to work and still receive benefits for a time, that allow increased access to vocational rehabilitation programs, that give state Medicaid administrators greater leeway to cover people returning to work from disability, that defer medical eligibility reviews for people trying to return to work, and that have funded advocacy organizations to improve communication between individuals with disabilities and state agencies (Ticket to Work and Work Incentives Improvement Act of 1999). Research is under way to assess the implementation and outcomes of these initiatives.

Cases

Albertson’s, Inc. v. Kirkingburg, 527 U.S. 555 (1999)

Cleveland v. Policy Management Systems Corp., 526 U.S. 795 (1999)

Murphy v. United Parcel Service, 527 U.S. 516 (1999)

School Board of Nassau County v. Arline, 480 U.S. 273 (1987)

Sutton v. United Airlines, Inc., 527 U.S. 471 (1999)

Toyota Motor Manufacturing, Kentucky, Inc. v. Williams, 534 U.S. 184 (2002)

Statutes, Regulations, and Legislative Materials

Age Discrimination in Employment Act, 29 U.S.C. §623 (2006)

Americans with Disabilities Act of 1990, 42 U.S.C. §§12101–12213 (2006)

Education for All Handicapped Children Act of 1975, P.L. 94-142, 89 Stat. 773 (Nov. 29, 1975) (now called the Individuals with Disabilities Education Act and codified at 20 U.S.C. §1400 et seq.)

Fair Housing Amendments Act of 1988, 42 U.S.C. §3613 et seq. (2006)

Rehabilitation Act of 1973, 29 U.S.C. §§791–794 (2006)

The Ticket to Work and Work Incentives Improvement Act of 1999, P.L. 106-170, 113 Stat 1860 (December 17, 1999)

Title VII of the Civil Rights Act of 1965, 42 U.S.C. §2000e-2(a) (2006)