Social Security Disability Insurance Program Worker Experience
ACTUARIAL STUDY NO. 118 by Tim Zayatz, A.S.A.

I. ENTITLEMENT TO DISABILITY INSURANCE BENEFITS

After onset of a severe physical or mental impairment, a worker may become entitled to monthly disability insurance benefits under the Social Security Old-Age, Survivors, and Disability Insurance (OASDI) program, provided he or she:

A worker's cash benefit is classified as an award at the time of initial payment. Additional auxiliary benefits may also be payable to other family members based on the earnings record of the entitled worker. This study analyzes the activity of disabled workers of the Social Security Disability Insurance (DI) program as described under title II of the Social Security Act.

A. Definition of Disability

For purposes of entitlement to DI benefits, disability is defined as the inability to engage in any substantial gainful activity (SGA) by reason of any medically determinable physical or mental impairment. The impairment must be expected to result in death or to last for a continuous period of at least 12 months. In addition, the disability must prevent the claimant from performing previous work, or engaging in any other kind of work in which a significant number of jobs exist. It is immaterial whether such work exists in the claimant's immediate area, or whether a specific job vacancy exists, or whether the claimant would be hired if he or she applied for work.

Several additional points are worth mentioning:

B. Disability Insured Status

To qualify for Social Security benefits for themselves and their dependents, individuals must work in employment covered by Social Security or be self-employed for a certain period of time, or have a specific amount of covered earnings in a year. Credit for this work is measured in quarters of coverage (QCs), or credits.²

An insured status test of the number of credits earned by a worker is required to establish a period of entitlement to any benefit. To be considered for disability benefits, a worker must satisfy disability insured status requirements, which consist of having obtained a requisite number of total credits as well as a specified number of credits earned in recent quarters. The worker must accrue a sufficient number of credits to be deemed fully insured—at least six credits and no more than 40 credits are required for this insured status.³ The recency-of-work test is satisfied if the worker has earned at least 20 credits during a 40-quarter period that ends with the quarter in which the waiting period begins.

A special recency-of-work test for younger workers provides an alternative to the 20/40 requirement. A worker who is under a disability which began before the quarter of attainment of age 31 satisfies the recency-of-work requirement if credits were earned for at least one-half of the quarters during the period beginning with the quarter after the quarter the worker attained age 21, and ending with the quarter in which the disability began. If this period contains 12 or fewer quarters—that is, if the disability begins in the quarter the worker attains age 24 or earlier—then a minimum of six credits must be earned in the 12-quarter period ending with the quarter in which the disability began.

Note that an individual disabled because of statutory blindness is not required to satisfy the recency-of-work test. Disability insured status is granted if the worker is fully insured only.

C. Waiting Period

The waiting period for DI benefits consists of 5 consecutive full calendar months beginning with the earliest full calendar month throughout which the worker satisfied both the definition of disability and the disability insured requirements. Benefits are not payable during the waiting period. However, the waiting period is waived for individuals who had a prior period of disability, which ended within 5 years of the current period of disability. In contrast, there is no waiting period for disability benefits under the SSI program.

II. EXPERIENCE OF DISABILITY DETERMINATIONS

A. Substantial Gainful Activity (SGA)

Substantial work activity involves the performance of significant physical or mental duties that are productive in nature. The degree to which an impairment limits an individual's ability to perform basic work activities is essential in determining the severity of the disability. Basic work activities include: sitting, standing, walking, lifting, carrying, handling, reaching, pushing, pulling, climbing, stooping, seeing, hearing, speaking, understanding, carrying out simple instructions, using judgment, responding appropriately in a work setting, and dealing with changes in work routine.

Gainful work activity is work performed for pay or profit. Certain earnings criteria have been established as reasonable indications of whether an individual is engaged in SGA. The dollar amount associated with defining SGA is specified in Federal regulations, and was originally set at $100 at the inception of the DI program. This amount had been updated on an ad hoc basis until January 2001, at which time the regulations were revised to provide for annual indexing of the SGA level based on the increase in average wages. From 1990 through 2000, the SGA level actually consisted of a primary amount and a lower secondary amount. Earnings above the primary amount ordinarily demonstrated SGA, whereas earnings less than the secondary amount demonstrated lack of SGA; earnings between the two amounts required consideration of all circumstances related to work activity. This tier structure was discontinued beginning in 2001. Since 1977, blind persons have been subject to a separate SGA amount.⁴ Figure 1 shows a history of the level of SGA.

Figure 1.—Substantial Gainful Activity Amounts
for the Disability Insurance Program 1957-2005

[D]

B. Impairments

To establish the presence of an impairment, an individual must provide supporting medical evidence along with the disability claim. The Social Security Administration's Listing of Impairments is used to determine the severity of the disability. The listings contain examples of common impairments for each of the major body systems that are deemed to be of such severity as to prevent a person from performing SGA. However, a diagnosis of a listed impairment alone may not be sufficient to establish disability; associated symptoms, clinical signs, and laboratory findings must accompany it. In addition, claimants are asked to provide the names of employers and job duties for the last 15 years.

Many individuals are found to be disabled even though impairments fail to meet the level of severity required in the impairment listings. In these cases, an individual's medical condition is evaluated in conjunction with age, education, and job skills. These vocational factors are given increasing weight with the advancing age of the worker, and are particularly significant in the determination of disability among workers age 50 or older.

The leading diagnostic categories for disability vary by gender and year of award. Table 1 shows the leading causes among DI disabled workers. Ranked by overall percentage, musculoskeletal disorders represent the largest single category of impairment, accounting for 24.4 percent of all awards over the period 1998-2002. As the baby-boom generation (birth cohorts 1946-64) continues to age, arthritic, back, and bone disorders are expected to become increasingly more prevalent. It is anticipated that the DI program will continue to experience a growing proportion of awards in this impairment category. Also note that the jump in the percentage of musculoskeletal impairments in 1995 can be explained by a change in reporting method. Prior to 1995, the Office of Disability (OD) reported the diagnosis of all awards based solely on the distribution of allowances at the initial level. In essence, allowances at the appeals level were assumed to have the same distribution among impairment categories as those at the inital level. In 1995, OD began reporting the actual diagnosis of all awards at all levels of review. This change revealed a significantly higher percentage of musculoskeletal impairments at the appeals level. The accompanying drop in the percentage of mental impairments in 1995 suggests that at least some of the awards in this category were among those misclassified under the old reporting method.

Mental impairments rank second in overall percentage, accounting for 23.8 percent of all awards, and continue to have the highest percentage among females for the 5-year period ending in 2002. Mental disorders had been the leading cause of disability for both sexes until being displaced by musculoskeletal disorders in 1996. However, mental impairments became the leading cause again in 2001. This may be due, in part, to a special administrative initiative. Effective in that year, SSA began conducting a review of SSI recipients who are potentially eligible for DI benefits due to previously unrecognized disability-insured status. Many of the resulting awards were found to be individuals under age 35 who were diagnosed with a mental disorder. Although review of this special disability workload is expected to continue for a number of years, the impact on specific impairment categories is uncertain. Also note that revised listings for mental impairments (published in 1985) led to the re-adjudication of a large number of cases, resulting in a jump in new awards in 1986.

Among males, circulatory disorders have always been a leading cause of disability, accounting for roughly 14-16 percent of total awards over the last 10 years. Neoplastic disorders have also been steady in recent years, accounting for roughly 9-11 percent of total awards. An age comparison within these two categories reveals nearly five times as many circulatory impairments among those age 50 or older compared with those age 35-49; and three times as many neoplastic disorders for the same age distribution. Although medical advancements continue in these categories, the large disparities show that these types of impairments will continue to significantly impact the DI rolls as the baby-boom ages. Finally, the percentage of awards based on infectious disease has decreased significantly over the last 5 years, mainly due to the marginalized impact of HIV infection.

Awards based on neoplastic and metabolic impairments rank higher among females than males, whereas awards based on circulatory impairments rank lower. Higher prevalence of cancer, and hormonal and eating disorders, as well as increasing prevalence of diabetes among females may account for this. Note that the impairment listings were changed in 1999 to eliminate awards based solely on obesity, accounting for the large decline in the percentage of nutritional and metabolic impairments beginning in 2000.

C. Determination Process

At the initial stage of a claimant's request for disability benefits, the State Disability Determination Services (DDS) will make a decision to allow or deny the claim. A claimant who is dissatisfied with the initial decision may request further review. This review process consists of several steps, which must be requested within specified time intervals, and in the following order:

Table 2 presents data on the disposition of claims for DI disability benefits across the various review stages, for calendar years 1993-2004. The data are tabulated by year of filing, with an additional break out by program involvement—showing claims for DI benefits only, or concurrent entitlement to DI and SSI benefits. As mentioned earlier, both programs use the same definition of disability for adults. However, eligibility for SSI benefits is further dependent upon the claimant's countable assets and income, which may include DI benefits.

Many factors exist that affect the number of disability claims filed as well as the frequency of subsequent decisions to either allow or deny benefits. However, the impact of any one factor is difficult to gauge; in general, they may be administrative, economic, or demographic in nature. Below is a list of some of the leading determinants which may have a significant impact on both the number of claims filed and the rate of favorable determinations ⁵:

D. Applications

The number of DDS claims for DI disabled worker benefits declined steadily from 1.25 million in 1994 to 1.04 million in 1998. Some factors contributing to the decline in this period include a robust economic expansion and lower levels of unemployment; leveling off of female labor force participation; a decline in HIV-related impairments; and the elimination of drug addiction and alcoholism as material causes for disability.

Over the following 6 years, the DI program experienced nearly a 43 percent increase in applications to an estimated 1.49 million in 2004, with roughly a 10 percent increase alone in each of 2001 and 2002. This increase in DDS claims is not fully understood, except in general terms. Having an obvious impact is the aging of the baby-boom, which continues to progress through the peak ages of disability; and higher levels of unemployment resulting from the economic contraction which began in the second quarter of 2001. A small portion of additional claims come from the review of the special disability workload of SSI recipients who may be eligible for DI benefits, as previously mentioned.

E. Initial Decisions

From 1992 through 1995, the allowance rate declined even as application growth continued. This may be indicative of claimants—caught in the economic downturn of a post-recession period—exhibiting less severe impairments. The result was fewer allowances from a growing number of claims. It is worth noting that allowance rates tend to be much lower among those concurrently filing for DI and SSI benefits than those filing for DI benefits only. This may be due to differences in the composition and economic status of the filers. Concurrent filers tend to be of lesser means (reflective of the nature of the SSI program) and are thus more likely affected by changes in the economy. Many times the only alternative is to seek aid from Federal, State, or local programs. Consequently, concurrent filers may exhibit less severe disability, or provide less evidence of impairment, resulting in fewer allowances.

Also note that for years with pending decisions, the ultimate allowance rate will be lower than that shown in table 2. This is due to the greater processing time involved in unfavorable decisions.

F. Reconsideration

Allowance rates at the reconsideration level have been very consistent. Although the reconsideration stage is de novo ⁶ in concept, it is similar to the initial stage in that disability determination is mostly a "paper review" process where claimants are rarely observed by the decision-maker. Assuming some uniformity among the initial decision-makers, it follows that initial denials are seldom overturned at reconsideration.

Currently 10 States are participating in a test of the Prototype Model to redesign the disability determination process. Features of this model include elimination of the reconsideration step, along with use of a single decision-maker and process enhancements at the DDS and OHA levels. The results shown in table 2, therefore, reflect a mixture of experience for the standard decision model and the modified test process. The Prototype Model test is set to expire on December 31, 2005.

G. Appeals Beyond Reconsideration

The subjectivity inherent in assessing disability leaves considerable room for interpretation of evidence. As a result, overturned decisions at the OHA level and beyond remain relatively high. Factors that contribute to the high reversal rate include:

Class action suits can also have an impact on the determination process. Public pressure has surfaced in controversial areas such as mental impairment issues; the amount of leverage given to allegations of pain; statements by treating physicians in the absence of clinical evidence; how HIV-related impairments and cardiovascular diseases are evaluated; use of vocational factors in the absence of a single debilitating impairment; and the consistency of DDS decisions with SSA policy. Although the number of claimants directly involved in any one case may not be large, the outcome may have a broader and subtler influence on subsequent rulings and determinations.

Finally, Federal efforts aim to improve the disability determination process by striving to reach the proper determination at the earliest possible stage. This effort will help to maintain managable backlogs, as well as reduce the rate of overturned decisions at the OHA level.

III. EXPERIENCE OF DISABILITY INCIDENCE

A. History

Since the commencement of disability cash benefits in July 1957, dynamics of the DI program have been subject to many internal and external factors. Congressional action, public opinion, and court rulings have shaped program characteristics including: how disability is defined; the determination of entitlement; the level of benefits; the review process of current beneficiaries; and ultimate program cost. Prior to 1960, the DI program applied only to workers age 50 or older. Prior to 1965, a claimant needed to be permanently disabled to qualify for benefits. The Social Security Amendments of 1967 (Public Law 90-248) eased the insured status requirements for persons under age 31, allowing a substantial number of young beneficiaries to enter the rolls. From 1968 through 1970, disability incidence remained fairly stable; however, through the early 1970s program growth far exceeded any reasonable expectations.

The introduction of the Black Lung program (1970) and the SSI program (1974), and a severe economic recession (1974-75) led to hundreds of thousands of new disability claims. In addition, administrative policy also tended to change as the DI program became bigger and more complex. Notably, the SSI program generally requires applicants under the age of 65 to apply for benefits from all other programs including DI, which may partially or fully offset SSI benefits. As expediency in processing applications was naturally given high priority, central office review of DDS initial decisions fell to roughly 5 percent in 1972 from 100 percent prior to 1972. The increased public awareness and pressures of administering two new programs probably contributed significantly to the sharp increase in new awards from 1972 to 1976.

The Social Security Amendments of 1977 (Public Law 95-216) and the Social Security Disability Amendments of 1980 (Public Law 96-265) also had a significant impact on the DI program. The 1977 amendments changed the benefit formula used to calculate benefits awarded in 1979 and later. The 1980 amendments introduced a more restrictive limit on the total monthly amount of Social Security benefits payable on a disabled worker's account; and mandated a 65 percent review rate of DDS allowances to assure uniformity of decisions. The return to high levels of review during this period led the DDSs to give increasingly careful consideration to new allowances, and increased the chances of reversing an initially favorable decision. These circumstances contributed to steadily declining awards from 1977 through 1982.

By 1984, DI program policy had undergone another reversal. Congressional and public concern over the removal of a large number of beneficiaries (particularly the mentally impaired) resulted in an administrative moratorium on the review of the disability rolls while Congress considered new DI legislation. Many beneficiaries whose benefits had been terminated were returned to the rolls through the appeals process. This initiated a period of increased court appeals and class action suits. In response, Congress passed the Social Security Disability Benefits Reform Act of 1984 (Public Law 98-460). Provisions of the Act include: revised mental impairment standards; increased emphasis on treating physician opinion; emphasis given to the combined effects of multiple impairments in the absence of a single severe impairment; required proof of medical improvement prior to termination of benefits; and standards to evaluate pain.

B. Recent Experience

Table 3 shows the total number of DI disabled workers awarded benefits grouped by calendar age ⁷ at time of award, for calendar years 1980-2004. These awards are also illustrated in figure 2.

Figure 2.—Disabled Worker Awards
by Age Group, Calendar Years 1980-2004

(In thousands)

[D]

Table 4 shows the associated disabled worker incidence rates, which are expressed as annual awards per thousand disability insured not already receiving benefits. Adjusted figures are expressed as age-adjusted (male and female) or age-sex-adjusted (total) relative to the exposed population as of 2000. Tables 3 and 4 are tabulated as of the year the beneficiary is added to the rolls. Note, however, that the year of award may actually be different from the year of disability onset or entitlement. This is due to factors such as the waiting period, the nature of the determination process, and claims processing times in general. Consequently, the incidence rates shown are not necessarily representative of true morbidity rates for the stated calendar years.

The incidence of disability increases considerably among workers beginning at age 50. The chance of DI entitlement increases naturally with advancing age, but also as a result of greater consideration given to vocational factors. This program characteristic continues to have a significant impact on costs as the baby-boom generation progresses through ages of higher incidence. Note that the rates shown in table 4 for ages 60 and older are likely to understate the true incidence of disability since, beginning at age 62, a disabled worker may elect to forgo disability benefits, opting instead to receive reduced old-age benefits. A number of factors influence this decision, including: the disability waiting period (old-age benefits are payable immediately); the potential for worker's compensation offset; differences between disability and old-age maximum family benefits payable; and the possibility for denial of disability benefits.

Age-specific disability incidence among female workers is typically lower than males. However in 1997, female incidence began to exceed that for males at ages 35-54. Although both genders experienced a general decline in incidence in this age range over the following five years, the decline is more pronounced among males. This is likely due to the elimination of drug and alcohol addiction from the DI rolls and a sharp decline in HIV impairments, both of which are predominantly male incidence categories. It is also interesting to note that the decline in incidence rates among females did not necessarily translate into a decline in the number of benefits awarded in this specific age range. This is due to the relatively large increases in female labor force participation. Over the 10-year period ending in 1996, labor force participation rates among females age 35-54 increased between 3.5-10.0 percentage points resulting in an additional 10.7 million disability insured. Over the same period and age range, male labor force participation decreased by roughly 2.0-2.5 percentage points.

Some of the same factors that affect the number of DI applicants can also affect incidence rates. In general, the decline in incidence between 1975-82 is attributable, in part, to a stricter program. Following a very low-growth period in incidence from 1983-89, the program experienced a surge in claims beginning in 1990, and incidence rates rose significantly and remained relatively high through 1995. In the late 1990s, the prevailing economic and political environment was characterized by robust economic expansion, low unemployment, and legislative restrictions on the qualifications of certain impairments. In addition, advancements in medical treatment; public need for employer-sponsored healthcare protection; and pursuit of financial goals for retirement may have provided incentive to remain in the labor force, if possible. These factors contributed to the decline in applications and awards during this period.

The sharp increase in incidence rates over 2001-04 represents a notable departure from the experience of the late 1990s, which generally showed modest annual declines in the age-sex-adjusted rate. While the aging of the baby-boom is always recognized as a factor in program growth, the increase in incidence in 2001 is likely due in large part to the severe economic contraction experienced in that year. However, special administrative activity by SSA beginning in 2001 has also contributed slightly to the surge in awards. As previously mentioned, the special disability workload was the result of discovering a substantial number of SSI recipients whose disability-insured status was not previously recognized. As this caseload is processed over the next several years, the resulting disability awards will contribute to temporarily higher incidence rates than would be expected from current underlying trends.

Table 5 presents historical termination data for disabled workers. Termination experience is discussed in detail in the next section.

Table 6 shows the number of disabled workers in current-payment status, at the end of calendar years 1980-2004. The current-payment population is derived from the in-force beneficiary population (not shown), which is reduced by the number of entitled individuals whose benefits are suspended. Common reasons for the suspension of DI benefits include: engagement in SGA following completion of a trial work period; worker's compensation offset; or imprisonment. Over the period 1983-2004, the current-payment population has grown at an average annual rate of roughly 4.3 percent, as growth in awards has steadily outpaced growth in terminations. Note that as a consequence of the scheduled increase in normal retirement age, there appear DI beneficiaries age 65 or older for the first time in 2003. The following table summarizes the progression of the DI rolls.

IV. EXPERIENCE OF DISABILITY BENEFIT TERMINATION

A. Background

The reasons for termination of DI disabled worker benefits can be grouped into four main categories:

Generally, the final month of entitlement to disability benefits for a worker is the earliest of the following:

The law contains several provisions for individuals who wish to return to work, but continue to have a disabling impairment:

The trial work period (TWP) is a 9-month period—not necessarily consecutive—during which an entitled beneficiary may work without affecting the right to benefits. Earnings during the 9 months are not counted toward SGA, and benefits will continue as long as the beneficiary has not medically recovered.

Individuals who continue to have a disabling impairment following the 9-month TWP, receive an extended period of eligibility (EPE). Earnings during the EPE are counted toward SGA, and monthly benefits will not be paid when such earnings exceed the SGA limit. If earnings fall below the SGA limit anytime during the EPE, benefits are automatically reinstated. Effective January 1, 1988, the law was amended to lengthen the EPE from 15 months to 36 months for individuals entitled to benefits in January 1988 or later.

After 24 months of disability entitlement, a beneficiary becomes eligible for Medicare coverage—regardless of age—and receives services as long as DI entitlement continues. In the case of an individual engaging in SGA, coverage is provided throughout the TWP and EPE. Medicare benefits were further extended through a provision of the Ticket to Work and Work Incentives Improvement Act of 1999 (Public Law 106-170)—effective October 1, 2000, coverage for beneficiaries who return to work will continue for an additional 4 1/2 years beyond the EPE.

The Ticket to Work Act also allows certain previously entitled individuals to request expedited reinstatement (EXR) of disability benefits when a disabling impairment no longer permits the performance of SGA. The provision provides a safety net for those who successfully return to work and, consequently, lose entitlement to disability benefits. EXR allows an individual to request reinstatement without filing a new application, and provides up to 6 months of provisional (temporary) cash benefits while SSA conducts a medical review to determine whether the individual can be reinstated. The request for reinstatement must be filed within 5 years of the termination of disability benefits. EXR became effective January 1, 2001.

Special provisions are granted to persons experiencing multiple periods of disability. Individuals who become re-entitled to benefits within 5 years of the end of a previous period of disability are not required to satisfy the 5-month waiting period. In addition, the 24-month waiting period for Medicare coverage need not be consecutive months and time may accrue over multiple periods of disability.

Beginning at age 62, a disabled worker may elect to receive old-age benefits in lieu of disability benefits. Although most disabled beneficiaries elect to receive DI benefits until normal retirement age—at which time conversion to benefits payable under the Old-Age and Survivors Insurance (OASI) program occurs automatically—some choose to switch earlier. This decision is influenced by economic as well as personal factors. One of the most common reasons is the existence of benefits payable outside the DI program. For example, worker's compensation benefits may partially or totally offset a DI benefit, but would not affect an OASI benefit. Another common economic factor is the difference in maximum family benefits payable under the DI program, which may be lower than the maximum payable under the OASI program. Personal factors may include the beneficiary's own assessment of health and life expectancy.

Recovery from disability may occur when the beneficiary either notifies SSA of an improved disabling condition, demonstrates the ability to engage in SGA, or is judged to no longer meet the definition of disability. The DDS or the central office will conduct a continuing disability review (CDR) from time to time based on warranting situations such as:

Certain "outside" variables such as disability caseloads, backlogs, budget restrictions, and legislation can have an impact on CDR activity and disability recovery rates in general, without actually affecting the underlying rate of medical improvement. These factors shape the prevailing administrative policy and can exert considerable influence on the nature of allowances, and the degree of overall impairment-severity of the DI rolls. To a lesser extent, mortality rates are also affected by exogenous variables. For example, the elimination of drug and alcohol related impairments, and an increase in allowance based on vocational factors or mental impairments may lead to an improvement in the overall mortality profile of the DI rolls.

B. History

Many variables can affect the rate at which beneficiaries are terminated from the disability rolls, including:

The Social Security Amendments of 1965 (Public Law 89-97) modified the definition of disability by replacing the requirement of permanent disability with the expectation that the disability last at least 12 months. This led to the entitlement of less seriously impaired claimants and lower mortality rates among the disabled. The 1967 amendments eased the insured status requirements for claimants under age 31. A growing portion of younger and relatively healthier beneficiaries further contributed to the decline in the mortality rates of the DI rolls.

As mortality rates fell in the early years of the program, the gross recovery rate generally increased. With the introduction of government-funded rehabilitation programs, elimination of the "permanently disabled" condition, and the extension of benefits to younger claimants, the recovery rate among beneficiaries rose from 19 per thousand in 1965 to 32 per thousand by 1967. Thereafter, the gross recovery rate decreased rapidly through 1975. This was due in large part to changes in the administration of the program. With the introduction of the Black Lung and SSI programs in the early 1970s, workload pressures resulted in the suspension or curtailment of some administrative review procedures. For example, by 1972 the central office reviewed only 10 percent of DDS continuances in which medical recovery was expected. Previously, 100 percent of such continuances were reviewed. By 1976, the gross recovery rate began to increase again as central office review of continuances returned to 100 percent.

Throughout the 1970s, the DI program experienced substantial increases in cost, mainly the result of significant growth in incidence. Under then-current policy, reviews were performed only in those cases where the beneficiary's condition was expected to improve, or voluntary reports or posted earnings indicated work activity. However, by the late 1970s measures to curtail inaccurate award determinations and improve the review process were intensified. One significant provision of the 1980 amendments required that beneficiaries with non-permanent impairments be reviewed every 3 years, and permanently disabled beneficiaries be reviewed at intervals determined by the Commissioner. Using that legislative mandate, the Reagan Administration initiated a major review of the disability rolls that resulted in a large number of cases in which it was determined that recovery had occurred.

Ensuing public disapproval of the newly implemented review process led to a moratorium on reviews of all cases of mental impairment disability.¹¹ Revision of mental impairment criteria and the review process followed and more than half of those removed from the rolls were reinstated upon appeal. The result was a sharp drop in recoveries as well as a sharp increase in new awards throughout the remainder of the 1980s.

C. Recent Experience

In the latter part of the 1980s, the agency experienced reductions in both work force and administrative funding. By the early 1990s, there existed a shortage of personnel needed to handle a significant increase in claims, as well as to meet review schedules. In an effort to free up resources to process initial claims, the agency sharply curbed the review of existing beneficiaries. Beginning in 1994, growth in initial claims began to level-off and once again attention shifted to performing mandated reviews. Congress enacted the Contract With America Advancement Act of 1996 (Public Law 104-121), which included a provision authorizing the appropriation of funds to be used exclusively to conduct additional CDRs. As intended, the special funding helped to eliminate DI backlogs by the end of 2000, and to become essentially current in the processing of SSI-only CDRs by the end of 2002. Since then, administrative budget delays have caused the agency to fall behind in processing certain SSI CDRs, emphasizing the need for renewed funding and revised scheduling plans.

Table 5 shows the historical number of terminations and gross termination rates for disabled workers, by reason for decrement. As evident from the data, most terminations occur as a result of death or conversion. With the exception of a sharp decrease in conversions in 2003¹², both categories exhibit steady long-term trends with little variation from year-to-year. In contrast, the number of recoveries can deviate considerably from a somewhat normal level. As previously mentioned, exogenous variables can greatly impact this category—and consequently the recovery rate—without actually affecting the underlying rate of medical improvement. As an example, we see that a spike in recoveries occurred in 1997 as the result of a provision of Public Law 104-121, which eliminated drug and alcohol addiction from the impairment listings and explicitly denied benefits in cases where drug or alcohol addiction were contributing material factors to the disability. Although categorized as "recoveries", this group of beneficiaries was simply eliminated from the rolls due to a change in law. "Other" is a relatively small category comprised mostly of individuals who switch to old-age benefits prior to normal retirement age. Figure 3 shows the distribution of DI disabled worker terminations by reason.

Figure 3.—Disabled Worker Terminations
by Reason, Calendar Years 1980-2004

(In thousands)

[D]

As mentioned, death and conversion account for most of the terminations that occur, and the general trend in termination rates has been downward. Figure 4 shows the distribution of DI disabled worker termination rates by reason. Several trends in the disability rolls have developed over the years which help explain the decline: mortality improvements and a reduction in the average age of beneficiaries.

Over the period 1980-95, the average age among male disabled workers in current-payment status steadily declined from 52.8 to 49.8 years. The proportion of the DI rolls made up of beneficiaries aged 35-49 grew from 20 percent in 1980 to 34 percent in 1995, while the proportion of 50-64 year olds fell from 71 percent to 55 percent. Similar trends were experienced among female disabled workers as the average age declined from 53.7 to 49.9 years. Higher incidence of mental disorders has led to an increase in younger and (physically) healthier beneficiaries. The result is fewer deaths, as well as a smaller percentage of beneficiaries converting to old-age benefits each year.

Since 1995, the average age among disabled workers in current-payment status has steadily increased. In 2004, the average age of disabled males was 51.6 years. The proportion of the DI rolls made up of beneficiaries aged 35-49 declined slightly to 29 percent, while the proportion of 50-64 year olds grew to 62 percent. Similar trends were experienced among female disabled workers as the average age climbed to 51.4 years. Note that the increase in average age is mostly an effect of the aging baby-boom generation and has not yet resulted in any significant increase in conversions. The trend in termination rates continues to be downward, in large part due to mortality improvements, as well as the scheduled increase in normal retirement age.

Figure 4.—Disabled Worker Termination Rates
by Reason, Calendar Years 1980-2004

(Per thousand exposed)

[D]

Medical progress has significantly contributed to longer life expectancy among the general population. Certain advancements have also had a considerable, albeit less quantifiable, impact on selected segments of the disability population. Over the period 1980-89, the DI rolls show gross rates ranging from roughly 50-55 deaths per thousand males, and 35-40 deaths per thousand females. Throughout the early 1990s, mortality gradually improved. A noticeable decline in the death rate after 1995, especially among males, is due in part to the rapidly diminishing impact of HIV-related impairments and the elimination of drug and alcohol addiction from the impairment listings. Other significant trends leading to a lower mortality profile of the disability rolls include: increasing prevelance of musculoskeletal and mental impairments, which tend to be less life-threatening than circulatory or neoplastic disorders; advancements in the treatment of malignant growths; and a greater proportion of awards to older workers, whose determinations are based on a set of vocational factors rather than a single severe disability. Over the period 2000-04, the DI rolls show stable gross rates of roughly 35 deaths per thousand males, and 26 deaths per thousand females.

D. Death Experience (1996-2000)

Tables 7A-7C show the probability of death for male and female DI disabled workers, by select age at entitlement to disability benefits; and duration—measured in years since selection. Data reflect the actual experience of the DI rolls from January 1, 1996 through December 31, 2000. The methods used in table construction and graduation are detailed in the appendix.

Mortality among disabled workers generally increases with select age. For any given select age, the probability of death is greatest during the first year of entitlement, then decreases dramatically during the second and third years of entitlement. Death probabilities tend to level off sometime around the fifth or sixth duration for males, and the third or fourth duration for females. For older select ages, mortality is lowest at these durations before trending upward in the later durations as general demographic factors such as age of the beneficiary begin to have an increasing effect. Greater consideration given to vocational factors, especially after age 50, may cause a slight decline in mortality. Disability mortality among males is higher than females for virtually all attained ages. Unique data considerations affecting estimates of disability mortality—such as death during the waiting period and tracking DI beneficiaries upon conversion to old-age benefits—are discussed in the appendix.

An analysis of broad age categories may explain some of the trends and characteristics of disability mortality. Disabled males under age 35 show the highest concentration of mental disorders, infectious and nervous impairments, and injuries. Disabled males aged 35-49 show a consistent mix of various impairments following the predominant mental and musculoskeletal categories; an increase in circulatory and neoplastic impairments is apparent. Disabled males aged 50 or older show the highest concentration of musculoskeletal awards, along with considerable increases in circulatory, neoplastic, and respiratory disorders; though still an important category, the prevalence of mental impairments is down considerably from the younger age groups.

Females show comparable trends in broad age categories with a few categorical changes. Disabled females under age 35 show the highest concentration of mental disorders—over 43 percent of new awards occurred in this category in 2002. Disabled females aged 35-49 continue to show high incidence of mental and nervous impairments, along with large increases in musculoskeletal and neoplastic disorders. Disabled females aged 50 or older show the highest prevalence of new awards in the musculoskeletal, mental, and neoplasm categories.

Due to the many reasons already discussed, disability mortality over the 5-year period covered in this study (1996-2000) has improved for roughly 80 percent of all attained ages among male and female beneficiaries, when compared to the previous 5-year period (1991-95)¹³. Among males, mortality in the more recent period is generally 5-20 percent lower than mortality in the earlier period, with dramatic reductions of 35-45 percent at some younger select ages. The improvements at the younger ages are largely due to advancements in HIV treatment and legislation eliminating drug and alcohol addiction from the impairment listings. Mortality improvement among females is as widespread but less dramaitic as that seen among males, generally ranging from 5-15 percent lower than the earlier period.

Tables 8A-8C show the progression of a series of cohorts—each for a given select age—reflecting the probabilities of death shown in tables 7A-7C. These survival tables are a concise way of representing the probabilities of a particular population living to a particular age. See appendix for details on table construction and usage.

Tables 9A-9C show the expected future lifetime of DI disabled workers. Values are based on the survivorship experience shown in tables 8A-8C. As with the general population, disabled females display a higher future lifetime than males. Note that life expectancy is generally greater during the second or third year of entitlement than during the first year of entitlement. This is due to relatively higher mortality rates within the first several years of disability, after which workers seem to show a greater propensity for disability continuation or non-death termination.

Tables 10A-10C show the absolute rate of death per thousand entitled disabled workers. These rates are derived from the probabilities of death shown in tables 7A-7C, and represent a different concept in measuring the rate of termination from the DI rolls. Absolute rates consider only the effect of a single decrement in a multiple-decrement environment, reflecting the probability of survival and ultimately termination under that one decrement. These are not true probabilities and as such are expressed on a "per thousand" basis. See the appendix for further details.

Tables 11 and 12 show the aggregate probability of death and expected future lifetime, by select and attained age. Probabilities are based on aggregate counts of exposure and deaths across all durations. They represent the average probability of death within one year for beneficiaries originally entitled at a particular select age (table 11), or who have attained a particular age (table 12). Similarly, aggregate future lifetime represents the average life expectancy of beneficiaries for a particular select or attained age. Values are exposure-weighted averages of the select-and-ultimate future lifetimes shown in tables 9A-9C.

Table 13 shows the aggregate probability of death and expected future lifetime, by duration. Probabilities are based on aggregate counts of exposure and deaths across all select ages, and represent the average probability of death within the next year of entitlement to disability benefits. Aggregate future lifetime represents the average life expectancy of beneficiaries who have been entitled for the stated number of years.

E. Recovery Experience (1996-2000)

Tables 14A-14B show the probability of recovery for male and female DI disabled workers, by select age and duration. In general, chances of recovery decline with advancing select age. For any particular select age, the probability of recovery exhibits a bimodal distribution. Disability recovery tends to peak during the second year of entitlement before declining the following year, then peaks again at its maximum level in the fifth year of entitlement before declining thereafter. This effect is likely caused by CDR schedules, which are based on the likelihood of medical improvement. Beneficiaries for whom medical improvement is expected (MIE) account for roughly 5-10 percent of all reviews, which are scheduled anywhere from 6 to 24 months following the most recent disability decision. Roughly 10.5 percent of MIEs result in an initial cessation (i.e., those occurring before any appeal of decision) of benefits, causing the first peak in recoveries. Beneficiaries for whom medical improvement is possible but less likely to occur within the first 2 years (MIP) account for roughly 65-70 percent of reviews, which are scheduled every 36 months. Roughly 8.5 percent of MIPs result in an initial cessation of benefits, causing the second peak in recoveries. Beneficiaries for whom medical improvement is not expected (MINE) account for roughly 25-30 percent of reviews, which are scheduled every 5 to 7 years. Roughly 4.5 percent of MINEs result in an initial cessation of benefits.

As previously stated, the probability of death among males exceeds that for females for virtually all attained ages, as males experience nearly twice the incidence of high risk circulatory disorders and two-thirds more injuries. The probability of recovery among males also exceeds that for females in over 85 percent of all attained ages. Evidence suggests that this may be due to differences in the concentration of impairment categories among genders, especially at older ages. During the observation period, both sexes exhibit high prevalence of mental and musculoskeletal impairments. However the relative concentration of mental impairments among females is much greater—18 percent higher overall, and 25 percent higher among ages 50 or older. The nature of these types of impairments seems to contribute to both lower mortality and lower chance of recovery among female disabled workers.

Analysis of the 5-year period covered in this study shows a dramatic increase in the rate of disability recovery for virtually all male and female beneficiaries, when compared to the previous 5-year period (1991-95)¹⁴. For males, most ages exhibit a 50-100 percent increase in the probability of recovery, while females show increases ranging from 25-75 percent. A disproportionate amount of the improvement is attributable to the elimination of beneficiaries whose disability was based on drug or alcohol addiction, causing a one-time spike in recoveries in 1997. That year, the overall recovery rate among males more than doubled to 27 per thousand entitled versus 12 per thousand in the prior year. Females also experienced a surge to 16 per thousand versus 10 per thousand in the prior year.

As previously mentioned, exogenous variables can also influence recovery rates. During 1991-95, the DI program experienced significant growth in claims, which limited the number of CDRs performed over that period. This trend was reversed through enactment of Public Law 104-121 in 1996, which provided special funding for CDR activity, and contributed significantly to the large increase in recoveries over 1996-2000.

Tables 15A-15B show the number of lives for a series of cohorts—each for a given select age—reflecting the probabilities of recovery shown in tables 14A-14B. Note that in this case, the concept of "survival" refers to beneficiaries who remain on the DI rolls by not recovering.

Tables 16A-16B show the expected number of future years of disability entitlement for those who have not recovered and remain on the DI rolls. Values are based on the survivorship experience shown in tables 15A-15B, which reflect termination due to recovery only, and automatic conversion to old-age benefits upon attaining normal retirement age. As previously discussed, males tend to exhibit a greater chance of recovery than females. Analysis shows that, on average, females remain on the disability rolls roughly 6 months longer than males.

Tables 17A-17B show the absolute rate of recovery per thousand entitled disabled workers, which are derived from the probabilities of recovery shown in tables 14A-14B. See the appendix for details on the derivation of these rates.

Tables 18 and 19 show the aggregate probability of recovery and expected time on the DI rolls, by select and attained age. Probabilities are based on aggregate counts of exposure and recoveries across all durations. They represent the average probability of recovery within one year for beneficiaries originally entitled at a particular select age (table 18), or who have attained a particular age (table 19). Similarly, expected time on the DI rolls represents the average number of future years of entitlement for those who remain on the rolls. These values are exposure-weighted averages of the values shown in tables 16A-16B, which reflect termination due to recovery only, and automatic conversion to old-age benefits upon attaining normal retirement age.

Table 20 shows the aggregate probability of recovery and expected time on the DI rolls, by duration. Probabilities are based on aggregate counts of exposure and recoveries across all select ages, and represent the average probability of recovery within the next year of entitlement to disability benefits. Expected time on the DI rolls represents the average number of future years of entitlement—reflecting termination due to recovery only—for those who have not recovered after the stated number of years.

F. Combined Experience (1996-2000)

Tables 21A-21B show the combined probability of death or recovery for male and female DI disabled workers, by select age and duration. Values are derived from death probabilities shown in tables 7A-7B and recovery probabilities shown in tables 14A-14B.

Tables 22A-22C show the number of lives for a series of cohorts, reflecting termination from DI entitlement due to death or recovery. The tables also reflect termination from OASI entitlement due to death, following conversion to old-age benefits.

Tables 23A-23B show the expected number of future years of combined DI and OASI entitlement for those originally entitled to disability benefits. Values are based on survivorship experience shown in tables 22A-22C, which reflect automatic conversion to old-age benefits upon attaining normal retirement age.

Tables 24A-24B show the expected number of future years of DI entitlement for those who have not died or recovered. Values are based on survivorship experience shown in tables 22A-22B. Entitlement is not considered after conversion to old-age benefits.

Tables 25 and 26 show the aggregate probability of death or recovery and expected time on the beneficiary rolls, by select and attained age. Values are derived from the average death and recovery probabilities discussed previously. They represent the average probability of total decrement within one year for beneficiaries originally entitled at a particular select age (table 25), or who have attained a particular age (table 26). Aggregate expected time on the OASDI rolls represents the average number of future years of combined entitlement for those originally entitled to disability benefits. These values are exposure-weighted averages of expected time on the combined rolls shown in tables 23A-23B, and reflect automatic conversion to old-age benefits upon attaining normal retirement age. Aggregate expected time on the DI rolls represents the average number of future years of disability entitlement. These values are exposure-weighted averages of expected time on the DI rolls shown in tables 24A-24B. Entitlement is not considered after conversion to old-age benefits.

Table 27 shows the aggregate probability of death or recovery and expected time on the beneficiary rolls, by duration. Values are derived from average death and recovery probabilities, and represent the average probability of total decrement within the next year of entitlement. Aggregate expected time on the OASDI rolls represents the average number of future years of combined entitlement for those originally entitled to disability benefits who have not died or recovered after the stated number of years. Aggregate expected time on the DI rolls represents the average number of future years of disability entitlement for those who have not died or recovered after the stated number of years.

G. Annuity Tables

Tables 28A-31D show the present value of a stream of payments to a disabled worker, by select age at entitlement. Annual or monthly payments are made at the beginning (annuity-due) or end (annuity-immediate) of each period. Receipt of payment is contingent upon survival to the next payment date. Note that "survival" denotes remaining entitled to benefits, which may imply not recovering as well as not dying. Payments are discounted using the stated annual effective interest rate and various survivorship assumptions. Tables 28A-29D show the actuarial present value of a life annuity of $1 payable to a disabled worker. Values are based on survivorship experience shown in tables 8A-8C, which reflect termination due to death only, and automatic conversion to old-age benefits upon attaining normal retirement age. Tables 30A-31D show the actuarial present value of a life annuity of $1 payable to a disabled worker until age 65. Values are based on survivorship experience shown in tables 22A-22B, which reflect termination due to death or recovery.

Source: Annual Statistical Supplement to the Social Security Bulletin (1984-2002); Annual Statistical Report on the Social Security Disability Insurance Program (2003).