Measuring Functional Capacity and Work Requirements: Summary of a Workshop (1999)

This panel session was designed to assess the following questions:

• What are the specific components of the functional capacity domains?

• How are the specific components linked to demands of work?

• Is it possible to develop a baseline of work requirements? Can the Department of Labor's Occupational Informational Network (O*NET) be used or adapted to meet the Social Security Administration's (SSA) need for an occupational classification system?

The topic of this panel is central to the SSA disability decision process. Inherent in the SSA statutory definition of disability is a need to link impairment with inability to perform substantial gainful activity (SGA). Impairment alone, however, is not sufficient to meet the test of disability. Functional capacity is the concept linking impairment to the ability to perform SGA. Work requirements are a way to specify the components of work and the abilities, skills, and other activities needed to perform competitive work.

A systematic and valid method for linking impairment-related functional capacities with work requirements and the ability to perform SGA would be highly desirable in a redesigned SSA disability decision process. It is important

to be clear about constructs before measuring them and to be concerned about the reliability and validity of measures of these constructs.

There may be a theory that connects impairments, functional capacities, and work requirements, but there are many links in the theoretical chain, complicating development of an effective disability determination process. Hence, it is important to be able to distinguish among the constructs—impairment, functional capacity, and work requirement. However, our current ability to do that is limited.

The boundaries between impairments and functional capacity limitations are not absolutely clear despite the various conceptual models developed. There also is the need to identify the specific characteristics that are said to make different components of these concepts operational. One needs to know whether an individual is demonstrating a manifestation of one or the other of these impairments and then must be able to rate them in terms of severity or their degree of limitation. Three levels of limitation have been identified up to this point in the process—conceptualization, identification of the conceptual state in an individual, and then the rating of it. The next step is to develop a cost-effective functional capacity process to do the determination, and then, finally, it has to be implemented in the real world.

The multiplicative effect of error in determination of disability demonstrates that it is hard to make this process totally accurate. Research is needed to perfect abilities to measure in order to move the field closer to the ideal of accuracy in disability determinations. Three real-world examples relevant to linking assessment of functional capacity to work in the disability determination process, together with related research, can be described:

1. Review of the reliability of the Social Security Administration's psychiatric and mental impairment standards in the early 1980s: Many people's benefits were terminated as a result of the redetermination. A disproportionate number of people taken off the disability rolls had mental impairments. This led to a reexamination of the mental impairment standards, or Listings of Impairments, used in SSA's disability determination process. The review found that the Listings already made an implicit link between impairment, in this case a mental condition, and functional limitation. The listings had embedded in them both assessment of the severity of one's impairments in mental functioning and measures of functional capacity, such as activities of daily living, the ability to perform them, the ability to concentrate, and the ability to interact socially. The mental impairment standards were revised; not surprisingly the functional capacity measurement remained part of the determination process at the listing level. Since mental impairment influences the whole person and the ability to do certain things, these functional capacity measures clearly need to be embedded in the listing.

The problem of lack of clarity in the conceptual framework is illustrated by the example of concentration measure. Some view this measure as an impairment and not a functional limitation. Revised standards were tested to see if clinicians could do a better job ascertaining disability and agree among themselves on a determination of disability. A very high degree of agreement was found except in very difficult cases with limited documentation or confusing case histories. This experience showed that listings structured in this manner could be useful, but that there were limitations in making these assessments—lack of information, lack of clarity about these constructs, whether they were related to each other, and how long a person had or was likely to have these particular functional limitations.

2. One of the measures in the multi-axial approach to psychiatric diagnosis focuses on social functioning. A measure of global assessment of functioning that has been around for many years involves rating on a single ordinal scale the severity of a person's psychological functioning together with the severity of social and occupational functioning globally. At times it can be very difficult to rate particular cases, especially when there is not a high degree of correlation between the severity of psychopathology and social functioning. The conclusion was to separate the assessment of social and occupational functioning from the symptom severity measure. Physicians, unlike nonphysician raters, found it difficult, even with training, to think about social and occupational functioning without thinking about symptom severity. This finding speaks to the limits in the ability to go from concept to implementation. Even after the constructs are made clear and anchors provided in the instruments, the conceptual framework does not always work well in practice.

3. As part of its contract with SSA, The American Institutes of Research (AIR) is attempting to link functional capacity measures with work requirements, using a large matrix for rating purposes. They asked a group of raters who are knowledgeable in the subject to rate the extent to which a specific component of functional capacity (called "functional assessment constructs") is related to various demands of work from the O*NET on a 9-point ordinal scale. The scale ranged from "0" (no relationship) to ''8" (strongly related). In some cases no relationship could be imagined between the constructs (e.g., between arm strength and attention to detail); in other cases the constructs are identical (e.g., between mathematical reasoning and using mathematics in an occupation).

As part of this effort, the specific functional capacity domains, the components of work requirements from the O*NET,⁶ and the experience of trying to assess the relationship between them were reviewed.

Specific functional assessment constructs included sensory, physical, cognitive-intellectual, emotional stability, general work behavior, activities of daily living, and medical conditions affecting work. Each domain contains a subset of conceptual factors (36 in all). For example, the sensory domain includes factors for hearing, vision, smell, and cutaneous sensation. Hearing is subdivided into two "functional categories": auditory receptive safety, and the other for auditory receptive communication. There are 89 functional categories in all, each of which is the smallest subunit of functional assessment constructs.

This review considered five major categories of specific work requirements and demands: abilities, skills, work styles, general work activities (GWAs), and work context. Other O*NET domains include knowledge, education, training, experience, and licensing. Abilities include oral comprehension, memorization, finger dexterity, and depth perception, for example. Skills include reading comprehension, troubleshooting, and time management. Work styles include initiative, integrity, and innovation. GWAs include processing information, thinking creatively, developing and building teams, and staffing organizational units. Work context includes formality of communication, conflict situations, body position (e.g., sitting), and level of competition.

For example, oral comprehension is considered a work requirement in O*NET, that is, there are jobs that demand that you be able to demonstrate oral comprehension. Some other examples are a functional assessment measure for verbal communication, a math reasoning requirement for some jobs, and a functional assessment measure also called math reasoning. Then there is a stamina measure and an endurance measure. These measures have a high degree of relationship.

However, when it comes to psychiatric listings and some functional assessment measures that deal with the whole person in the context of certain roles, there is a blurring of these functional assessment measures and work requirements. This is more clearly seen in the measures chosen as the "B criteria" for functional limitation in SSA's listings of mental impairment. Activities of daily living (ADL), for example, are listed as functional assessment measures. ADLs speak to whether one

can accomplish the task of getting to the workplace, but O*NET, in the specification and requirements of work, does not deal with whether one can get to the workplace at all.

Then there are the issues of adaptation such as can the person deal with changes that might occur. These are constructs that are related to the whole person that should be thought of as functional assessment measures, and are clearly related to the ability to work in any work situation. Concentration at the extreme is one of the most basic mental functions. If you cannot attend, many other mental functions, and subsequently whole body functions, cannot be performed.

There are some fundamental functional assessments that, if measured in the extreme and rated at the extreme in an individual, preclude virtually all work. Impairments characterized by such extremes in function could be considered as universally disabling. The question—Are there certain functional capacity measures that, when below a particular level, preclude all work?—should be considered in developing, measures of functional capacity to work. Social interaction is another example. Virtually all jobs require interaction with peers, coworkers or the public, and with supervisors.

In conclusion, it should be emphasized that linking occupational demands with functional assessment measures is central to disability determination. It is, at best, tricky, but it is a necessary challenge.

This presentation summarizes a program of research, extending over 40 years, to develop methods that link job requirements to human capabilities for performing job tasks. Specifically, the presentation reviewed the research which identified abilities in the cognitive, psychomotor, physical, and sensory-perceptual domains of human performance. Second, the development of a job analysis method, designed to determine the levels of these different abilities required for jobs was described, as was a resource developed for linking specific tests to the abilities required in jobs. Finally, some examples of studies linking the ability and medical requirements of jobs were provided. This program eventually provided the conceptual and empirical foundation for the ability requirements section of the O*NET occupational classification system.

Identification of Ability Requirements. An initial program of research was concerned with the identification of psychomotor abilities accounting for perform

ance in a wide range of human tasks. A series of factor analysis studies examined the intercorrelations among proficiencies of hundreds of subjects, performing a wide range of tasks requiring different kinds of motor performance skills. It was found that these tasks could be clustered into nine categories requiring common abilities. The underlying abilities accounting for these performances were identified as: control precision, multilimb coordination, reaction time, response orientation, timing, arm-hand steadiness, finger and manual dexterity, and speed of limb movement (see, for example, Fleishman, 1972). From the task requirements, it was possible to provide detailed specifications for each ability requirement. It was also possible to identify the tests that were most diagnostic of each ability. It was shown that the kinds of motor abilities involved in psychomotor tasks were independent of the types of motor performance required in physically demanding tasks.

At the time, many terms were in use to describe physical performance, such as speed, agility, muscular endurance, and strength, but no one was sure about the most appropriate categories, their definitions, generality, overlap, and so forth. Also, the tests most diagnostic of various physical abilities had not been specified. Subsequently, in a series of factor analysis studies, involving the administration of comprehensive batteries of physical tests to several hundred subjects, nine physical ability factors were identified from the correlations among their performances. These included four strength factors (static, dynamic, explosive, and trunk strength), two flexibility factors, an equilibrium factor, a gross-body coordination factor, and a stamina (cardiovascular endurance) factor. It was possible to define these different abilities quite precisely in terms of the tasks to which they extended and to delimit their generality. Detailed definitions of these physical abilities have been provided (Fleishman, 1964; Fleishman and Reilly, 1992). Furthermore, it was possible to specify the tests most diagnostic of these abilities, where each test had high reliability and high factor loadings on particular abilities (see Fleishman, 1964, 1969; Myers et al., 1993). Extensive factor analytic and experimental research by many investigators has been carried out in the cognitive and sensory-perceptual domains as well (Carroll, 1993; Fleishman and Reilly, 1992).

Development of an Abilities Taxonomy and a Job Analysis System. In a program originally sponsored by the Advanced Research Projects Agency of the Department of Defense, an extensive project on taxonomic issues relevant to human performance research and measurement was carried out (summarized in Fleishman and Quaintance, 1984). As a facet of this work, alternative ways to estimate the ability requirements of jobs and job tasks were examined. One extensive effort concerned the development of a job analysis methodology for rating job tasks in terms of their ability requirements (Fleishman, 1975; 1979; 1982). The methodology developed involved presenting very carefully defined abilities, based on the best factor analysis research information about the ability, and a series of rating scales containing empirically derived task anchors repre-

sentative of that ability at different points on each scale. The positions of these anchors on each scale were obtained empirically.

Fifty-two such scales were developed covering the abilities in the cognitive, sensory-perceptual, psychomotor, and physical domains of human performance. These rating scales have been combined into the Fleishman Job Analysis Survey (F-JAS) (Fleishman, 1992). Respondents, including either job incumbents, supervisors, or job analysts, examine jobs or job tasks and assign them scale values on each of the ability rating scales. The mean scores of these groups of raters provides the job's ability requirements profile. Thousands of jobs have now been studied and interrater reliabilities are high. There is also very high agreement between profiles obtained from incumbents, supervisors, and job analysts (see, for example, Fleishman and Mumford, 1991).

A recent line of research has examined the domain of interpersonal abilities, most relevant to jobs involving interactions with others. The resulting taxonomy, ability definitions, and behaviorally task-anchored rating scales have been developed for 21 such abilities (e.g., social confidence, dependability, social sensitivity) (Abod et al., 1996). These scales have now been incorporated as a part of the F-JAS job analysis methodology.

Thus, we have five domains of human abilities (cognitive, psychomotor, physical, sensory-perceptual, and interpersonal), each composed of distinct abilities with differential relations to job performance. Extensive use of these methods has produced tests and assessment methods with empirical validity in predicting on the job performance (see, e.g., Fleishman, 1988; Fleishman and Mumford, 1991). The Handbook of Human Abilities (Fleishman and Reilly, 1992) has been developed; it provides specifications for tests that measure each ability as well as examples of available published tests that measure each ability. In this publication, commercially available tests are classified according to the ability measured within each functional assessment domain.

Ability Requirements and O*NET. The research described provided a foundation for the ability requirements subsequently included in O*NET. Research in the O*NET project assisted in the refinement of these scales and their tryout on a national sample of jobs. The reliability and utility of the scales were confirmed in this study (Fleishman et al., 1996). A database describing the levels of each of the 52 abilities required in more than 1,200 occupations has been developed (Peterson et al., 1996). Thus, the ability requirements of occupations can be described in the same terms as the functional assessment measures. A challenge for O*NET is to see if the system can identify jobs with minimal requirements suitable for individuals with different medically disabling conditions.

Relating These Methods to Medical Symptomatologies and Impairments in Jobs. The importance of providing relevant information about job requirements that can be linked to information about disabilities, medical symptomatologies and diagnosis, and rehabilitation was discussed. Recent attempts in reengineering

major disability assessment programs require better information about the job tasks that individuals with different disabilities can and cannot perform safely and effectively. This part of the presentation briefly mentioned some of these issues and described some prototype demonstrations of how the types of information provided by the F-JAS ability requirement scales have been utilized previously by those concerned with these problems in the workplace.

One line of work involved classifying jobs in terms of common levels of requirements in each of the F-JAS physical ability scales (Fleishman, 1988). Occupational medical specialists were able to link disqualifying symptomatologies in relation to the different levels of job ability requirements (e.g., Gebhardt et al., 1981; Fleishman, 1988; Hogan et al., 1978). In other studies (e.g. Fleishman et al., 1996), the linkage of information obtained from the rating, methodology to physiological and ergonomic indices of work capacity was demonstrated. Other applications of these ability scales involved development of a computerized support system integrating information about the physical requirements of jobs with diagnostic procedures practiced by physicians for use in occupational health and personnel services (Halpern, 1996). The O*NET system, which includes ability requirement scales, has considerable potential for providing information relevant in dealing with the issues discussed in this presentation.

At the heart of the Social Security Administration's two disability programs, Social Security Disability Insurance (SSDI) and Supplemental Security Income (SSI), is the statutory definition of disability. It is defined by the Social Security Act as the ''inability to engage in any substantial gainful activity by reason of any medically determinable physical or mental impairment which can be expected to result in death or which has lasted or can be expected to last for a continuous period of not less than 12 months" (Social Security Act, sec. 23(d)(1)(A)). The definition of disability can only be changed by an act of Congress. The regulations that put it into operation can be modified more readily. The following presentation will be based on this definition of disability and the notion that the way that it is implemented has the potential for change.

Three elements relate to this workshop on functional capacity and work requirements for the Institute of Medicine's Social Security disability study. The first is work. What is it? What are general and specific work requirements that SSA needs to consider? The second is functional capacity. What can the person do? What functional abilities and limitations does the person have that relate to work? The third element is how to fit them all together. What can the person do—or what can be predicted that the person could do—on a routine basis for

the customary workweek, under usual work conditions, that are activities which constitute work in our national economy?

The examples used in this presentation are taken from mental health. They are, however, not exclusive to mental disorders; they are equally applicable to all disorders. People with mental disorders who are disabled for work are the largest proportion of people currently receiving Social Security disability benefits in both the SSI and SSDI programs. Over the last decade, both the actual number of people receiving benefits on the basis of mental disorders as well as the proportion of beneficiaries receiving SSA disability benefits on the basis of mental disorders has increased. In addition, people with mental disorders who are receiving SSDI and SSI benefits are younger than people who are receiving benefits on the basis of other health conditions. They are in their prime earning years and, because people do not tend to die from chronic mental disorders, they are likely to remain on the disability roles.

In linking functional capacity with work requirements, this presentation focused on mental health, disability for work, and the World Health Organization's (WHO) International Classification for Impairments, Activities, and Participation (ICIDH-2). The ICIDH-2 offers a new conceptual model and a refined and expanded taxonomy that is intended for application in social security programs. Each item in the ICIDH-2 has an operational definition and at least one rating scale. In addition, research disablement assessment instruments, being developed in an independent project, are based on the ICIDH-2 and are intended for use both in surveys and in clinical settings. These research questionnaires could be adapted for the assessment of the ability to perform work and employment-related activities.

An International Task Force on Mental Health and Addictive, Behavioral, Cognitive, and Developmental Aspects of ICIDH has been established. It is charged with looking at: (1) the consequences of mental and addictive, behavioral, cognitive, and developmental disorders, and (2) the mental health aspects of any other disease or disablement, that is, any issue arising because of any impairment, disability, or handicap that deals with the mental well-being of a person. The task force is responsible for these areas in the ICIDH-2. To fulfill this responsibility, it is depending not only on expert opinion, but also is conducting international field trials of the concepts, clarity, organization, cultural sensitivity, and innovations to build an empirical base either to substantiate the draft document or to indicate areas in need of additional revision. Representation on the task force from the United States includes professional organizations such as the American Psychiatric Association; people with disabilities and their advocates, such as the National Alliance for the Mentally Ill; private disability insurance companies; and U.S. federal agencies, including SSA. The task force would like to see increased participation by SSA to make the work and the associated functional capacities classified in ICIDH-2 relevant and practical for application in SSA's disability determination.

ICIDH-2 conceptualizes three key dimensions of disablement. Each dimension is subclassified into domains and items that are each more detailed aspects of each dimension. The first dimension, impairments, classifies body parts or body systems (such as mental functions, including attention and memory) or organ systems (such as cardiovascular and respiratory functions). The second dimension, activities, classifies the activities in which people are typically engaged. These range from the very basic activities of movement of limbs through such fundamental activities of daily life as grooming and bathing (commonly known as activities of daily living) to more complex activities such as work. The third dimension classifies participation, the involvement of the person in life situations. (A fourth dimension, context, has been proposed by ICIDH-2 and is conceptualized as extrinsic factors that have positive or negative impact on functioning, performance, and involvement.) An example of participation in work would include an individual who is capable of working at the level of performing all the activities required of and related to a job, but is not hired because of a diagnosis of a mental disorder. This situation is one the Americans for Disabilities Act (ADA) was enacted to eliminate. The person might not be considered as disabled for the activity, but is systematically denied participation in work. Note that the contextual factors that impinge on or foster participation, for example, the lack or existence of such legislation as the ADA, are mutable. Other factors are changeable as well. For instance, the unions won the number of hours in the current workweek after many years of effort. This is currently described in SSA regulations as the "customary workweek." Actual employment has been traditionally operationalized using the Dictionary of Occupational Titles (DOT). SSA is currently considering the use of a grid called O*NET being developed to replace DOT.

For the purpose of SSA's determination of disability claims based on mental disorders, the ICIDH-2 Impairment chapter on mental functions and the Activities chapter that includes work are appropriate and applicable. They contain components that can be rated for purposes of adjudicating disability claims. Indeed, SSA's standards and guidelines were reviewed for input in developing the Activities section on work. It is not difficult to understand that mental impairment items such as attention, for instance, are necessary to work. Focusing attention, sustaining attention, and shifting attention are needed skills in both manual and nonphysical occupations. It is stating the obvious to note that recent memory and remote memory are also fundamental to all types of employment. The mental impairment of executive functioning may not apply to all types of work. As delineated in ICIDH-2, it includes concept formation, planning, flexibility, and judgment. Although the name of the mental function (i.e., executive function) is not intended to reflect the employment hierarchy, it does more suitably describe professional and supervisory work requirements than basic manual labor.

In the ICIDH-2 Activities dimension, work is currently classified along with school-related activities, since they tend to be differentiated by age rather than

the actual task the person is performing. For example, among the basic work activities are following directions, working independently, and working in groups. Both work and school require such other generic activities as attending regularly, being punctual, and responding to feedback. ICIDH-2 has a section on work acquisition and retention skills: it is not enough for a person to be able to get a job, the person has to be able to maintain it. Furthermore, there are items in other Activities chapters that classify work-related activities already acknowledged by SSA. For example, the ICIDH-2 chapter on Interpersonal Behaviors includes a section on interacting with persons in formal settings, that contains interaction with coworkers, superiors, and subordinates. The ICIDH-2 dimensions, domains, and items—along with their operational definitions—could be used by SSA to document relevant functions and activities, rate the person's performance on each item, and calculate the person's ability to work.

The above examples of ICIDH-2 Impairments and Activities highlight some of the basic functional capacities generic to work. The statutory definition of disability does not specify certain jobs; it states "substantial gainful activity" and can be understood to mean paid employment in the general economy. In the determination process, items such as those from ICIDH-2 would need to be assessed on the basis of being able to perform them on a continuous basis, over workweeks, over time, once the connection of ICIDH-2 Impairments and Activities to actual work is made.

Two examples of studies funded by the National Institute for Mental Health highlight the kinds of research that can be drawn upon in the process of revisiting SSA's disability determination process. They are predicated on and further examine the relationships between mental functions and work. Studies such as these begin to expand the boundaries of traditional research in rehabilitation associated with mental disorders. The studies illustrate the association of particular mental functions with specific activities, provide empirical support for the statutory definition of disability, and can be applied as the framework for developing both a process and guidelines for SSA's disability determination. Although not consciously designed with the ICIDH-2 model in mind, they show that the ICIDH-2 is in line with current conceptualization of disablements in the research community.

Finally, the ICIDH-2 is linked to WHO research instruments that assess disablements, as noted above. At present, along with the two versions of the WHO Disability Assessment Schedule (DAS-II) (instruments intended for clinical and survey research), there is a checklist for use in clinical practice that can provide an overview of a person's disablement, and a 12-item screening questionnaire. All of these could be used sequentially as part of the medical evidence or as a way of documenting the review of disability claims. The checklist provides the overview of the person's functioning, the screener identifies what areas should be examined in more depth, and the WHO DAS-II offers a more detailed picture of the person's functioning, performance, and involvement. Clearly, these instruments are not designed currently for direct application by SSA, but they offer

an alternative that could be adapted for disability benefit determination purposes. The advantage they offer is that the assessment instruments are being developed based on a research protocol that will determine their scientific and psychometric properties.

In summary, aspects of functional capacity, components of work, and ways of fitting the two together have been depicted with mental health examples. The WHO ICIDH-2 offers a conceptual model and taxonomy that is substantiated by research. In addition, there are disablement assessment instruments based on ICIDH-2 that can be adapted to SSA's disability determination. As mentioned, mental health has been used illustratively; the ICIDH-2 and the research instruments are designed for use with all health conditions.

The Public Health Service establishes goals for health every 10 years. One of the recommendations for Healthy People 2010 is to have every medical encounter form include an assessment of functioning and to use it for purposes of reimbursement and accreditation. ICIDH-2 has been recommended as the system to use. This move will also begin to build a base from which the SSA can request functional status information as part of medical evidence. ICIDH-2 can be used as the basis for determining disability both in the standards and in the evidence supplied.

Some of the key issues that surfaced during the general discussion are:

• The focus of the human performance research has been on capacity to do specific tasks and their correspondence to what people do in work. However, an issue remains concerning the predictive validity of the capacity to do those kinds of tasks at a broader level, that is, to do actual work, actual employment, actual performance of the work, and not just specific tasks. Studies have been conducted in various companies using performance criteria on the job. These studies have shown the correlation between performance on these tasks of either applicants or people who have just been hired and their subsequent performance on the job. The question that still needs to be answered is: Can these assessments made with job applicants be applied to people who are applying for benefits and not jobs? The existing literature needs to be examined intensively for leads and for application in the disability determination work.

• Disability assessment and determining disability according to the statutory definition are different tasks; the application of the assessment is to a specific job, not to any or all jobs in the national economy.

• To phrase the issue another way: Are there certain areas of functional capacity that are associated with a high degree of correlation with a vast

number of work requirements, such that they are embedded in virtually all work and where incapacity in the area would preclude a vast array of work? However, with many new jobs emerging in the economy, answering this question would be enormously complex and involves forecasting what work is going to look like in a global economy, what will be the requirements of those jobs, and whether one can begin to develop some descriptive components of those jobs (or requirements), and then begin to identify the skill sets that go into them.

• Moreover, as different components of performance of a job are defined, inevitably from a research or conceptual perspective, it gets increasingly complicated, and the complexity makes any practical application difficult. Two separate questions emerge: (1) Is there a way to reduce complexity in a manner that has practical application for the SSA definition; and (2) With a less complex set of factors, how does one establish a threshold for the low level of incapacity? The question, then, is whether the human performance research generalizes to that setting and whether such assessment would matter.

• Skills other than physical capacity will be needed for jobs in the future, and both the workplace environment and the person should be assessed. The issue of accommodations becomes crucial. Therefore, continuing to emphasize only the person in determining capabilities and eligibility for benefits is doing a disservice to people with disabilities. The ICIDH-2 should be considered seriously as a conceptual framework and a potential classification system, because it now includes components of environment.

• SSA makes three to four million decisions every year as to which applicants are disabled and who is eligible for disability benefits. It therefore needs a decision process that is better than what it has now. The NAS committee's second interim report raised several important questions about what the problems are, the likelihood of other approaches being better, and the options available to SSA. SSA faces the daunting challenge to develop a process with more specificity than the determination process currently in place. It has to be able to implement nationally whatever process develops, and the redesigned process has to allow SSA to make better decisions. The disability program started with the medical listings. Over time they have come to be a proxy for a lot of people for the decision on work disability. There are real questions about how good a proxy these listings are for the decision about whether a person can work or not. Also, SSA has noted that several of the listings have incorporated functional definitions along with the purely medical impairments basis. That was the genesis of increased emphasis by SSA on function in deciding if the individual is engaged in substantial gainful activity, that is to say, is the person able to work.