United States General Accounting Office GAO Report to the Chairman, Subcommittee on Social Security, Committee on Ways and Means, House of Representatives July 2003 SOCIAL SECURITY DISABILITY Reviews of Beneficiaries’ Disability Status Require Continued Attention to Achieve Timeliness and Cost- Effectiveness GAO-03-662 July 2003 SOCIAL SECURITY DISABILITY Reviews of Beneficiaries' Disability Highlights of GAO-03-662, a report to the Status Require Continued Attention to Chairman, Subcommittee on Social Security, Committee on Ways and Means, Achieve Timeliness and Cost- House of Representatives Effectiveness The Social Security Administration SSA will likely face a backlog of about 200,000 continuing disability review (SSA) has had difficulty in (CDR) cases by the end of fiscal year 2003. SSA officials attribute the conducting timely reviews of pending backlog to its decision to reduce the number of cases reviewed as a beneficiaries’ cases to ensure they result of the delay in obtaining fiscal year 2003 funding. In addition, the are still eligible for disability pending backlog resulted from putting more emphasis on initial applications benefits. SSA has been taking steps to improve the cost- over CDRs. To ensure CDRs receive adequate attention, SSA has requested effectiveness of its review process. some fiscal year 2004 funds be “earmarked” for these reviews. Given SSA’s SSA has linked the review process ability to eliminate its previous CDR backlog using targeted funds, this to eligibility for a new benefit that maneuver could help SSA. Over the next 5 years, SSA has estimated that 8.5 provides return-to-work services. million CDRs, costing about $4 billion, are needed to stay current. If SSA generates another backlog, cost savings and program integrity may be This report looks at SSA’s ability to compromised by paying benefits to disability beneficiaries who are no longer stay current with future reviews, eligible to receive them. identifies potential improvements to the review process, and assesses SSA is not making the best use of available information when conducting its the review process–return-to-work CDRs, leaving opportunities for improvement. First, SSA’s decisions on the link. timing of CDRs are not based on systematic analysis of available information. Second, SSA’s process for determining which CDR method to use is not always based on the best available information. For example, SSA GAO recommends that the requires an in-depth review for all beneficiaries who, upon entering the Commissioner of SSA program, are expected to medically improve even if current information on • pursue a more comprehensive, certain of those beneficiaries indicates that improvement is unlikely and that data-driven approach to the the review would be better handled through a shorter, less expensive method it uses to decide when method. Third, SSA has not fully pursued medical treatment data available to assess individuals for on- from the Medicare and Medicaid programs despite their potential to improve going eligibility; • rely more readily on current SSA’s decisions regarding which review method to use. Fourth, SSA’s CDRs assessments of beneficiary continue to be hampered by missing or incomplete information on information rather than on beneficiaries’ case history. assessments made at time of program entry when deciding SSA delays the provision of new return-to-work benefits to beneficiaries which review method to use; expected to medically improve based on the assumption that such and beneficiaries are least likely to need them. However, according to SSA • study, and incorporate if cost- data, about 94 percent of such beneficiaries are not found to have effective, the more comprehensive use of medically improved upon completion of a disability review. As a result, Medicare/Medicaid data into some individuals who might benefit from return-to-work services are SSA’s decisions about the initially denied access to them. SSA is reviewing this policy and while review method to use. doing so, will need to consider how to best balance its financial stewardship and return-to-work goals. In its comments on a draft of this report, SSA generally agreed with GAO’s recommendations. www.gao.gov/cgi-bin/getrpt?GAO-03-662. To view the full product, including the scope and methodology, click on the link above. For more information, contact Robert E. Robertson at (202) 512-7215 or RobertsonR@gao.gov. Contents Letter 1 Results in Brief 3 Background 5 End of Targeted Funding and Other Issues Could Contribute to Another Backlog, Threatening Cost Savings 10 Further Opportunities Exist for SSA to Improve CDR Cost- Effectiveness 17 SSA’s Rationale for Postponing Return-to-Work Services to Some Beneficiaries Is Not Well-Supported by Program Experience 24 Conclusions 27 Recommendations to the Commissioner of SSA 28 Agency Comments and Our Evaluation 29 Appendix I Scope and Methodology 32 Appendix II Comments from the Social Security Administration 34 Appendix III GAO Contacts and Staff Acknowledgments 38 GAO Contacts 38 Staff Acknowledgments 38 Tables Table 1: Estimated CDR Activities, Fiscal Years 2004-08 13 Table 2: DDS Directors’ Reported Likelihood, If Any, of Experiencing an Event That Jeopardizes Meeting CDR Workload During Fiscal Years 2004 and 2005 14 Table 3: Extent That DDSs Have a Strategy to Manage Anticipated Events and Likelihood That Events Will Have Negative Impact on Workload Processing During Fiscal Year 2004 and 2005 15 Table 4: Summary of SSA’s CDR Activities During Special Funding Period, Fiscal Years 1996-2002 16 Page i GAO-03-662 Social Security Disability Abbreviations CDR continuing disability reviews CMS Center for Medicare and Medicaid Services DDS Disability Determination Services DI Disability Insurance EF Disability Electronic Folder MIE medical improvement expected MINE medical improvement not expected MIP medical improvement possible SGA substantial gainful activity SSA Social Security Administration SSI Supplemental Security Income This is a work of the U.S. Government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. It may contain copyrighted graphics, images or other materials. Permission from the copyright holder may be necessary should you wish to reproduce copyrighted materials separately from GAO’s product. Page ii GAO-03-662 Social Security Disability United States General Accounting Office Washington, DC 20548 July 24, 2003 The Honorable E. Clay Shaw, Jr. Chairman, Subcommittee on Social Security Committee on Ways and Means House of Representatives Dear Mr. Chairman: The Disability Insurance (DI) and Supplemental Security Income (SSI) programs are the largest federal income programs for disabled individuals, paying about $86 billion to about 10 million disabled beneficiaries in 2002. These programs have been growing in recent years and are poised to grow further as the baby boom generation ages. To help ensure that only eligible beneficiaries remain on the rolls, the Social Security Administration (SSA) is required by law to conduct continuing disability reviews (CDR) for all DI beneficiaries and some SSI disability recipients to determine whether they continue to meet the disability requirements of the law. In addition, to assist beneficiaries who want to return to work and leave the disability rolls, SSA began implementing the Ticket to Work and Self-Sufficiency Program in 2002. Under this program, beneficiaries are issued a “ticket,” or voucher, which they can use to obtain vocational rehabilitation, employment, or other return-to-work services from an approved provider of their choice. Through much of the 1980s and 1990s, the Congress and GAO, among others, emphasized the importance of CDRs for maintaining DI and SSI program integrity and, consequently, the critical need for SSA to conduct CDRs when they are due and in a cost-effective manner. However, SSA had difficulty completing all required CDRs when they were due, which resulted in the development of an enormous backlog of 4.3 million cases. In 1996, the Congress, in response to these difficulties, authorized funding Page 1 GAO-03-662 Social Security Disability targeted exclusively for CDRs from fiscal year 1996 through 2002 to eliminate the CDR backlog and conduct new CDRs as they became due.1 At the time beneficiaries enter the DI or SSI programs or continue their benefits following a CDR, state-based Disability Determination Services (DDS) determine beneficiaries’ due date for a CDR based on their potential for medical improvement. Beneficiaries classified as “medical improvement expected” are generally scheduled for a CDR within 6 to 18 months, beneficiaries classified as “medical improvement possible” are scheduled once every 3 years, and beneficiaries classified as “medical improvement not expected” are scheduled once every 5 to 7 years. Once the date for a review arrives, SSA compiles information such as age, length of time on the rolls, and qualifying medical condition to determine if it would be cost-effective to complete the CDR based on information reported by the beneficiary on a mailed-out questionnaire (“mailer”). In instances where SSA determines it is best to examine the beneficiary in person, SSA sends the beneficiaries’ case file to the DDS for a full medical review. To reduce work disincentives and address some beneficiaries’ fear that any work activity could result in the termination of their benefits through a CDR, the Ticket to Work and Self-Sufficiency Program prohibits SSA from conducting CDRs for beneficiaries who are using a ticket. However, SSA has decided that it will not issue a ticket to beneficiaries who are expected to medically improve until their first CDR is completed. SSA believes these beneficiaries do not require assistance to return to work. But some disability advocates and policy experts believe that beneficiaries expected to medically improve could benefit from early ticket services and, therefore, should not be subject to restrictions on ticket issuance. 1 The Balanced Budget and Emergency Deficit Control Act of 1985, as amended, established statutory limits on federal government spending for fiscal year 1991 through 2002. The act created, among other provisions, annual adjustable dollar limits (spending caps) on discretionary spending funded through the regular appropriations process. The act, as amended, also required that SSA’s discretionary spending caps that existed through fiscal year 2002 be adjusted upward to account for appropriations targeted for CDRs. The Contract with America Advancement Act of 1996 authorized about $4.1 billion to be paid from the Old-Age and Survivors Insurance Trust Fund and the Disability Insurance Trust Fund to process CDRs in fiscal year 1996 through 2002. Hereafter, in this report, we refer to this funding mechanism authorized by the Balanced Budget and Emergency Deficit Control Act of 1985 as “CDR-targeted funding.” Page 2 GAO-03-662 Social Security Disability Given the importance of CDRs in ensuring program integrity and determining beneficiary eligibility for assistance under the ticket program, the Chairman of the Subcommittee on Social Security, House Committee on Ways and Means requested that we review the CDR process. In response, this report discusses: (1) what impact the expiration of targeted funding for CDR processing would have on SSA’s ability to remain current with the CDR caseload, and what level of funding would be needed over the next 5 years to keep the workload current; (2) what opportunities exist for SSA to improve the cost-effectiveness of the CDR process; and (3) whether SSA’s rationale for delaying return-to-work and vocational services under the ticket program for beneficiaries who are expected to medically improve is adequately supported by program experience. To answer these questions, we reviewed SSA documents, including the agency’s budget request and estimates of the cost and savings from conducting CDRs. Also, we surveyed 52 DDS directors to assess the potential effect of the expiration of CDR-targeted funding on DDS operations. Moreover, we analyzed SSA data on CDR outcomes, reviewed SSA-contracted studies of the CDR process, examined legislation, regulations, and SSA policy guidance related to CDRs and the ticket program, and interviewed SSA officials. We performed our work in accordance with generally accepted government auditing standards between August 2002 and May 2003. See appendix I for a more detailed description of our scope and methodology. With the expiration of CDR-targeted funds at the end of fiscal year Results in Brief 2002, SSA is at risk of generating another CDR backlog; based on SSA’s cost and workload projections, it would cost a total of about $4 billion or more over the next 5 years to complete its CDR workload. However, most of the backlog expected to appear by the end of fiscal year 2003 will likely consist of SSI CDRs and, according to SSA officials, this makes the backlog less problematic than if it consisted mostly of DI cases. The expected shortfall is attributable to several factors. One factor was SSA’s decision to reduce the number of CDRs it processed pending fiscal year 2003 funding decisions. Other factors relate to workload capacity and the lower priority given to CDRs relative to initial claims. In the years ahead, a CDR backlog could grow due to an expected increase in the number of initial claims as well as DDS’ potential difficulty with replacing the disability examiners who leave through retirement or attrition. If another large CDR backlog is generated, SSA is at risk of foregoing cost savings and compromising the integrity of its disability programs by paying benefits to disability beneficiaries who are no longer eligible to receive them. Page 3 GAO-03-662 Social Security Disability While SSA has taken a number of actions over the past decade to significantly improve the cost-effectiveness of the CDR process, opportunities remain for SSA to better use information in CDR decision making. In particular, SSA’s process for deciding when beneficiaries should undergo a CDR is not based on systematic analysis of available information, and likely results in some CDRs not being conducted at the optimal time. Also, SSA’s process for determining which method to use in conducting a CDR—mailer or full medical review—is not always based on the best available information. For example, SSA requires a full medical review for all beneficiaries who, upon entering the program, are expected to medically improve even if current information on certain of those beneficiaries indicates that improvement is unlikely and that the CDR would be better handled through a much less expensive mailer. In addition, SSA has not fully studied and pursued the use of medical treatment data on beneficiaries available from the Medicare and Medicaid programs despite the potential of these data to improve SSA’s decisions regarding whether to use a mailer or full medical review to complete a CDR. Finally, SSA continues to be hampered in its CDR decisions by missing or incomplete information on beneficiaries’ case history. While the exact magnitude of this problem is unknown, 72 percent of DDSs reported that missing or incomplete information hinders their ability to determine whether medical improvement has occurred, thereby making it difficult for SSA to cease benefits for some individuals who no longer meet eligibility standards. SSA’s rationale for delaying issuance of a ticket to beneficiaries expected to medically improve, based on the premise that they will regain their capacity to return to work without SSA assistance, is not well-supported by program experience. The majority of these beneficiaries—about 94 percent—are not found to have medically improved upon completion of a CDR. As a result, some beneficiaries who might otherwise benefit from potentially valuable return-to-work assistance have to wait up to 3 years to access services through the ticket program. SSA has acknowledged the need to reexamine this policy, and agency officials have informed us that they are in the process of doing so. As SSA reexamines this policy, it will need to consider alternatives that better balance the agency’s program stewardship and return-to-work goals. This report contains recommendations for further improving the cost- effectiveness of SSA’s CDR process. In its comments on a draft of this report, SSA agreed with our recommendations and said that our review represents a comprehensive and accurate assessment of SSA’s accomplishments in improving the CDR process as well as opportunities Page 4 GAO-03-662 Social Security Disability to improve the process. SSA also provided a number of technical comments, which we incorporated where appropriate. The DI and SSI programs are the two largest federal programs providing Background cash assistance to people with disabilities. Established in 1956, DI is an insurance program that provides monthly cash benefits to workers who are unable to work because of severe long-term disability. Workers who have worked long enough and recently enough are insured for coverage under the DI program. In addition to cash assistance, DI beneficiaries receive Medicare coverage after they have received cash benefits for 24 months. In 2002, SSA paid about $60 billion to 5.5 million disabled workers, with average monthly cash benefits amounting to $834 per person.2 DI cash benefits are paid from the Federal Disability Insurance Trust Fund.3 SSI, created in 1972, is a means-tested income assistance program that provides a financial safety net for disabled, blind, or aged individuals who have low income and limited resources. Unlike the DI program, SSI has no prior work requirement and no waiting period for cash or medical benefits. Eligible SSI applicants generally begin receiving cash benefits immediately upon entitlement and, in most cases, receipt of cash benefits makes them eligible for Medicaid benefits. In 2002, about 5.5 million people with disabilities received SSI benefits.4 In the same year, federal SSI cash benefits paid to SSI beneficiaries with disabilities equaled $26 billion, 2 Included among these 5.5 million beneficiaries are about 1.2 million beneficiaries who were dually eligible for SSI benefits because of the low level of their income and resources. In 2002, the DI program also paid about $6 billion in cash benefits to about 1.7 million spouses and children of disabled workers. 3 Most disabled Social Security beneficiaries are disabled insured workers who receive benefits through the DI program based on their own earnings record. However, as of 2002, about 952,000 Social Security disability beneficiaries were disabled surviving spouses and disabled adult children who qualified for disability benefits based on the earnings record of an insured spouse or parent. Many of these disabled surviving spouses and adult children receive their disability benefits through the Social Security Old-Age and Survivors Insurance (OASI) program, not the DI program. OASI disability benefits for disabled surviving spouses and disabled adult children are paid from the Old-Age and Survivors Trust Fund. To receive OASI disability benefits, surviving spouses and adult children must meet the DI program’s disability criteria, and they are also subject to the requirement for CDRs. For simplicity, our report refers to all disabled Social Security beneficiaries (in both DI and OASI) as “DI beneficiaries.” 4 About 3.9 million of these individuals were working age adults aged 18 to 64. Page 5 GAO-03-662 Social Security Disability and average monthly federal SSI cash benefits amounted to about $398 per person. SSI cash benefits are paid from general tax revenues. The DI and SSI programs use the same statutory definition of disability. To meet the definition of disability under these programs, an individual must have a medically determinable physical or mental impairment that (1) has lasted or is expected to last at least 1 year or to result in death and (2) prevents the individual from engaging in substantial gainful activity (SGA). Individuals are considered to be engaged in SGA if they have countable earnings above a certain dollar level.5 Moreover, for a person to be determined to be disabled, the impairment must be of such severity that the person not only is unable to do his or her previous work but, considering his or her age, education, and work experience, is unable to do any other kind of substantial work that exists in the national economy. SSA contracts with state DDS agencies to determine whether applicants are disabled. To help ensure that only eligible beneficiaries remain on the rolls, SSA is required by law to conduct CDRs for all DI beneficiaries and some SSI disability recipients to determine whether they continue to meet the disability requirements of the law. In 1980, because of concerns about the effectiveness of the CDR process and growing disability rolls, the Congress enacted a law requiring that CDRs be conducted at least once every 3 years for all DI beneficiaries whose disabilities are not considered permanent and at intervals determined appropriate by SSA for DI beneficiaries whose impairments are considered permanent. SSA issued regulations in 1986 stating its policy of conducting CDRs for SSI disability beneficiaries with the same frequency as it conducts CDRs for DI beneficiaries. In 1994, the Congress established the first statutory requirement for SSI CDRs, requiring that CDRs be conducted for a relatively small proportion of SSI beneficiaries. Welfare reform legislation enacted in August 1996 focused on CDRs for SSI children.6 This legislation required that SSA (1) conduct CDRs at least once every 3 years for SSI children under age 18 if their impairments are not considered permanent and for infants during their first year of life if they are receiving SSI benefits due to low birth weight and (2) review the cases of all SSI 5 For 2003, SSA considers countable earnings above $800 a month to be substantial gainful activity for persons who are not blind and above $1,330 a month for persons who are blind. 6 Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (Pub. L. No. 104-193). Page 6 GAO-03-662 Social Security Disability children beginning on their 18th birthdays to determine whether they are eligible for disability benefits under adult disability criteria. The redeterminations for 18-year-olds are considered part of the CDR workload. CDR Process At the time beneficiaries enter the DI or SSI programs, DDSs determine when beneficiaries will be due for CDRs on the basis of their potential for medical improvement. Based on SSA regulations, DDSs classify individuals into one of three medical improvement categories, called “diary categories”: “medical improvement expected” (MIE), “medical improvement possible” (MIP), or “medical improvement not expected” (MINE). Based on the diary categories, DDSs select a “diary date” for each beneficiary, which is the date that the beneficiary is scheduled to have a CDR. The diary date is generally within 6 to 18 months if the beneficiary is classified as MIE;7 once every 3 years if classified as MIP; and once every 5 to 7 years if classified as MINE. Upon completion of a CDR, DDSs reassess the medical improvement potential of beneficiaries who remain eligible for benefits to determine the most appropriate medical improvement category and time frame for conducting the next CDR. Beneficiaries classified as MIE are not eligible to receive Ticket to Work services until either the completion of their first CDR, or until they have received benefits for 3 years. While SSA uses diary categories to determine the timing of CDRs, it has developed another method, called profiling, to determine the most cost- effective method of conducting a CDR. Profiling involves the application of statistical formulas that use data on beneficiary characteristics contained in SSA’s computerized records—such as age, impairment type, length of time on disability rolls, previous CDR activity, and reported earnings—to predict the likelihood of medical improvement and, therefore, of benefit cessation. For example, SSA found that the longer an individual is on the disability rolls, the less likely he or she is to have benefits terminated. In addition, once an individual undergoes a CDR, the chance that a new CDR will result in benefit termination is reduced substantially. Reported earnings, on the other hand, greatly increase the likelihood of termination. 7 Although SSA’s policy guidance indicates that CDRs for MIE beneficiaries should generally be scheduled at intervals of 6 to 18 months, the guidance provides DDS personnel with flexibility to establish a diary date for any time period between 6 and 36 months. Page 7 GAO-03-662 Social Security Disability Through its profiling formulas, SSA assigns a “score” to beneficiaries indicating whether there is a high, medium, or low likelihood of medical improvement. In general, beneficiaries with a high score are referred for full medical reviews—an in-depth assessment of a beneficiaries’ medical and vocational status—while beneficiaries with lower scores are, at least initially, sent a questionnaire, known as a “mailer.”8 The mailer consists of a short list of questions asking beneficiaries to report information on their medical conditions, treatments, and work activities. If beneficiaries’ responses to a mailer indicate possible improvement in medical condition or vocational status, SSA may refer these individuals for a full medical review. However, in most cases, SSA decides that a full medical review is not warranted and that benefits should be continued. In contrast to mailers, full medical reviews are labor intensive and expensive. These reviews generally involve the following steps: (1) SSA headquarters personnel determine that a CDR is due and notify the SSA processing center; (2) personnel at the processing center locate the beneficiary’s file and send it to the appropriate SSA field office; (3) field office personnel contact the beneficiary, conduct a lengthy interview, and send the file to the appropriate DDS; (4) the DDS requests medical records from the beneficiary’s physicians and other medical sources and, if these sources cannot provide sufficient evidence, schedules medical or psychological examinations with consulting physicians outside the DDS; and (5) a DDS team, consisting of a disability examiner and a physician or psychologist, determines whether the beneficiary continues to meet SSA disability criteria. CDR Backlog As of fiscal year 1996, about 4.3 million CDRs were due or overdue. In response, SSA and the Congress focused on providing funding to conduct overdue CDRs and new CDRs as they became due. SSA developed a plan for a 7-year initiative to conduct about 8.2 million CDRs during fiscal years 1996 through 2002. In the Contract with America Advancement Act of 1996 (Pub. L. No. 104-121), the Congress authorized a total of about $4.1 billion to fund the 7-year CDR plan.9 In addition, The Personal 8 While SSA uses mailers primarily for beneficiaries with low profile scores, the agency has recently expanded its use of mailers to some beneficiaries with medium and high profile scores. 9 The act also required that SSA report to the Congress annually for fiscal years 1996 - 2002 on its CDR activities, including the estimated savings resulting from CDRs. Page 8 GAO-03-662 Social Security Disability Responsibility and Work Opportunity Reconciliation Act of 1996 (Pub. L. No. 104-193) required SSA to conduct CDRs on several beneficiary groups, such as low birth weight babies and authorized an additional $250 million for CDRs in fiscal years 1997 and 1998. The actual amount appropriated during the 7-year period, about $3.68 billion, was less than the amount authorized in 1996. SSA reported to the Congress in its fiscal year 2000 CDR report that in that year, the agency became current with the backlog of CDRs for all DI beneficiaries. SSA officials indicated to us that although they are in the midst of preparing the final statistics for its fiscal year 2002 CDR report, it became current with the backlog of CDRs for all SSI beneficiaries by the end of fiscal year 2002. CDR Cost-Effectiveness Since first implementing the profiling and mailer processes in the early 1990s, SSA has continued its efforts to improve the cost-effectiveness of the CDR process. Most notably, SSA has refined the statistical formulas used in profiling to identify which method—mailer or full medical review—should be used to conduct the CDR. According to SSA officials and studies of the profiling process, these improvements have led to some beneficiaries receiving a mailer who otherwise would have received a full medical review, thereby allowing SSA to reduce the overall cost of the CDR process. Conversely, by improving SSA’s ability to identify beneficiaries who are likely to medically improve, these refinements have also helped the agency better ensure that it is conducting full medical reviews—and ceasing benefits—when appropriate.10 In addition to improvements in its profiling process, SSA has also implemented other CDR process improvements such as introducing an automated review of mailers. 10 According to SSA’s study of its profiling model, the agency’s recent improvements in statistical profiling have resulted in hundreds of millions of dollars in annual savings from being better able to identify and cease the benefits of individuals who have a relatively high likelihood of medical improvement. Page 9 GAO-03-662 Social Security Disability In the midst of its first year following the cessation of CDR-targeted funds, End of Targeted SSA appears to be developing another CDR backlog; the agency estimates Funding and Other it will cost several billion dollars in total over the next 5 years to keep its workload current. By the end of fiscal year 2003, on the basis of SSA’s Issues Could current projections, the agency will likely face a backlog of 200,000 CDRs, Contribute to Another though the characteristics of the backlog may mitigate its negative effects. SSA attributes the mounting backlog to the management decisions it made Backlog, Threatening at the beginning of the fiscal year during budget deliberations, as well as Cost Savings the need to process a larger than expected workload of initial disability applications. SSA has estimated that it will need a total of about $4 billion to process its projected CDR workload over the next 5 years. However, SSA’s updated estimate, expected to be available later this year, will likely show a higher cost as the disability rolls continue to expand. Aside from funding issues, DDSs reported that challenges associated with processing initial disability applications and maintaining enough disability examiners could jeopardize their ability to stay current with the CDR workload over the next few years. If another large CDR backlog is generated, SSA is at risk of foregoing cost-savings, thereby compromising the integrity of its disability programs as a result of paying benefits to disability beneficiaries who are no longer eligible to receive them. CDR Backlog Likely to At the end of March 2003—six months after the expiration of separate Reemerge, Though Its authorized CDR funding—SSA was on a pace to generate a CDR backlog Characteristics May by the end of the current fiscal year. However, most of the backlogged claims will consist of SSI CDRs, which may make the backlog less Minimize Negative Effects problematic than it otherwise would have been because, among other reasons, SSI CDRs have lower long-term savings than DI CDRs. In its fiscal year 2003 budget justification, SSA indicated that it needed to process about 1.38 million CDRs during fiscal year 2003 to stay current with its CDR workload. Yet, SSA expects to process a total of 1.18 million CDRs, if not more, by the end of the fiscal year.11 By the end of March 2003—the midpoint of the fiscal year—SSA had processed about 539,000 CDRs. To reach the 1.18 million end-year revised total, SSA will need to process CDRs during the second half of the fiscal year at a pace similar to that 11 On May 14, 2003, SSA released its revised final performance plan for fiscal year 2003. The plan projects that SSA will process 1,129,000 CDRs during fiscal year 2003. SSA also expects to process an additional 20,000 CDRs initiated for reasons other than maturation of the scheduled diary date (e.g., a third party reports that the individual may no longer be disabled). Page 10 GAO-03-662 Social Security Disability achieved during the first 6 months of the fiscal year.12 Nevertheless, while it appears that SSA should be able to achieve this outcome, by the end of fiscal year 2003, it will have accumulated a backlog of 200,000 CDRs. SSA officials attributed the delay in obtaining a fiscal year 2003 budget as the main factor in hampering their ability to conduct all of the planned CDRs for the fiscal year.13 Because of the uncertainty surrounding the agency’s funding level, SSA reduced the number of CDRs it sent to DDS officials for processing as well as froze DDS hiring and overtime pay. SSA officials told us that they took these actions because they were concerned that the fiscal year 2003 appropriations would not support CDR activity at the fiscal year 2002 level. SSA officials recognize that a hiring freeze can have a longer-term impact because it disrupts the normal replacement of disability examiners lost through attrition. SSA officials explained that disability examiners generally do not increase overall productivity when first hired. In fact, new disability examiners could initially decrease productivity because experienced examiners may devote some of their time to training these new examiners. SSA officials noted that it generally takes 1 to 2 years before disability examiners become proficient. SSA’s management strategy to cut back on the number of CDRs it processed during the delays to the extended fiscal year 2003 budget process reflects the agency’s higher priority for processing of initial applications for disability benefits. Specifically, while SSA cut back on the number of CDRs, no similar action was reported with DI and SSI initial eligibility decision making. SSA officials indicated that the application rate for disability benefits increased during the beginning months of fiscal year 2003, further affecting its ability to stay current with CDRs. SSA officials told us that although SSA sets a goal to process all CDRs and initial applications, initial eligibility decisions are given the highest priority. Officials said that, due to political pressure, getting disability benefits to people in a timely manner is emphasized over reviewing whether current beneficiaries remain eligible for benefits. DDSs, likewise, place a greater priority on processing initial applications. Three-fourths (75 percent) of 12 SSA indicated that 710,000 CDRs had been processed nearing the end of April 2003. This year-to-date completion rate positions SSA to complete all 1.18 million CDRs. 13 The federal government had operated under a series of continuing resolutions from the beginning of the fiscal year through February 20, 2003. A continuing resolution is legislation that may be enacted to provide budget authority for agencies to continue in operation when the Congress and the President have not completed action on appropriations by the beginning of the fiscal year. Page 11 GAO-03-662 Social Security Disability directors said processing initial disability claims were a top priority relative to CDRs, whereas far fewer directors (23 percent) said that processing initial claims and CDRs were equal priorities. SSA has recently proposed an approach to avoid this competition between CDRs and initial claims. Specifically, in SSA’s fiscal year 2004 budget request, the Commissioner requested that almost $1.5 billion be earmarked for three activities that could provide a return on investment—CDRs, SSI nondisability redeterminations,14 and overpayment workloads. While we did not review the sufficiency of the level of this request, the earmarking of funds for activities such as CDRs could help SSA keep current with these activities. For example, if the number of initial applications for disability benefits continues to increase over the next several years, holding apart the necessary funds for CDRs could be a prudent measure. SSA has indicated in its annual CDR reports, as well as in its performance and accountability report, that its ability to complete all CDRs as they become due in the future is dependent upon adequate funding. In 2000, SSA estimated that a total of about $4 billion was needed to process the CDR workload during the 5-year period between fiscal year 2004 and 2008 (see table 1). SSA based these “rough estimates” on cost and workload projections available at that time. SSA expects to release updated workload and cost projections in the summer of 2003. While the estimates made in 2000 are not inconsistent with recent years’ authorized CDR funding levels, they rely upon assumptions that may change in the years ahead. For instance, the updated numbers for the fiscal year 2004 to 2008 period will likely be higher than the past estimate for this time period because of the recent growth in the disability rolls. 14 To determine whether beneficiaries remain financially eligible for SSI benefits after the initial assessment, SSA conducts nondisability redeterminations to verify eligibility factors such as income, resources, and living arrangements. Beneficiaries are reviewed at least once every 6 years, but reviews may be more frequent if SSA determines that changes in eligibility are likely. Page 12 GAO-03-662 Social Security Disability Table 1: Estimated CDR Activities, Fiscal Years 2004-08 CDRs to be processed during CDR expenses Cessationsa Fiscal year year (in thousands) (dollars in millions) (in thousands) 2004 1,637 $716 61 2005 1,682 $729 59 2006 1,632 $787 61 2007 1,769 $896 65 2008 1,793 $857 62 Source: SSA’s Office of the Chief Actuary, May 2000 estimates. a Estimated ultimate cessations after all appeals. Despite the likely reemergence of a CDR backlog, the characteristics of the backlog may mitigate its negative consequences. During fiscal year 2003, SSA has focused on DI CDRs. SSA officials cite four reasons for this: (1) cessations of beneficiaries receiving DI benefits lead to higher savings than cessations of recipients receiving SSI benefits, (2) SSA desires to protect the DI trust fund, (3) legislation sets out a clearer mandate to complete CDRs on beneficiaries receiving DI benefits than for adult beneficiaries receiving SSI benefits, and (4) external auditors cite SSA for noncompliance with the law when SSA does not complete the required CDRs for DI beneficiaries. As a result, most of the backlog that is expected to reemerge by the end of fiscal year 2003 will likely consist of SSI CDRs and, according to SSA officials, this makes the backlog less problematic than if the backlog consisted of mostly DI cases. SSA maintains that not only do SSI adult CDRs result in lower long-term savings, but also the legislative mandate for conducting SSI CDRs is less prescriptive. Therefore, the negative effects of falling behind on SSI CDRs are less severe. DDS Directors Expressed Several of the issues that have contributed to the pending fiscal year Concerns about Their 2003 CDR backlog will also appear, in the views of DDS directors, in the Ability to Meet Future future. First, nearly all directors expect to process a higher number of initial disability claims than in the past. Most DDS directors have a CDR Workload strategy in place to deal with this rising initial claims workload, but still expect increased initial claims to negatively affect their ability to process their CDR workload. Second, most directors expect to experience difficulties in maintaining an adequate level of staffing, caused by many examiners leaving and difficulties finding replacements. Most DDSs who anticipate facing these staffing challenges reported that they have Page 13 GAO-03-662 Social Security Disability strategies in place to manage them. Nevertheless, nearly all believe that these staffing issues will negatively impact their ability to stay current with their expected CDR workloads. Tables 2 and 3 provide more specific results. Table 2: DDS Directors’ Reported Likelihood, If Any, of Experiencing an Event That Jeopardizes Meeting CDR Workload During Fiscal Years 2004 and 2005 Numbers in percent Not at all Somewhat Very Event likely likely likely Higher number of initial disability claims than in past (n=51) 2 35 63 State budget shortfalls causing constraints (e.g., personnel restrictions) (n=49) 25 29 47 Difficulties hiring disability examiners (n=51) 28 31 41 High turnover of disability examiners due to reasons other than retirement (n=51) 35 51 14 Large number of disability examiner retirements (n=51) 39 39 22 Source: GAO survey of DDS directors, February 2003. Page 14 GAO-03-662 Social Security Disability Table 3: Extent That DDSs Have a Strategy to Manage Anticipated Events and Likelihood That Events Will Have Negative Impact on Workload Processing During Fiscal Year 2004 and 2005 Likelihood event will reportedly have a negative impact on staying current with projected CDR workloads in fiscal year 2004 and 2005, even with a strategy (in percentage) Percentage of DDSs anticipating event that have a strategy Not at all likely Somewhat likely Event currently in place or not sure or very likely Higher number of initial disability claims than in past (n=50) 78 8 92 State budget shortfalls causing constraints a (e.g., personnel restrictions) (n=37) 57 17 83a Difficulties hiring disability Examiners (n=37) 70 8 92 High turnover of disability examiners due to reasons other than retirement (n=33) 79 3 97 Large number of disability examiner retirements (n=31) 81 3 97 Source: GAO survey of DDS directors, February 2003. a Percentage based on 35 responses (2 of the 37 DDS directors did not indicate a response about the likelihood of the event having a negative impact on CDR workload). Cost Savings and Program To the extent that funding, staffing, and other issues limit SSA’s ability to Integrity Could Be process its CDR workload, the full realization of CDR cost savings could Jeopardized If CDR be in jeopardy. SSA maintains that the return on investment from CDR activities is high. In fact, SSA’s most recent annual CDR report to the Backlog Grows Again Congress summarizes its average CDR cost-effectiveness during fiscal year 1996 to 2000 at about $11 returned for every $1 spent on CDRs.15 SSA has noted, however, that such rates of return are unlikely to be maintained because as SSA works down the backlog and beneficiaries come up for their second and third CDRs, the agency does not expect as many cessations and, therefore, the cost-benefit ratio could decline. 15 SSA calculated its annual cost-effectiveness ratios by dividing the estimated present value of total lifetime benefits saved with respect to CDR cessations (including Old-Age, Survivors, and Disability Insurance, SSI, Medicare, and Medicaid savings) by the dollar amount spent on periodic CDRs in a given year. SSA points out that the ratios should be considered an approximation because, for example, costs do not include the costs of appeals processed after the end of a given year. However, SSA officials also noted that the administrative costs for CDRs in a given year include the costs of appeals of CDR cessations in prior years which are processed in that year. Page 15 GAO-03-662 Social Security Disability Since the Congress’ provision of dedicated CDR funding starting in fiscal year 1996, SSA has reported completing millions of CDRs that resulted in substantial long-term savings. Table 4 shows the number of CDRs processed annually between fiscal year 1996 and 2001, which ranged from about 500,000 to over 1.8 million. SSA has reported that these annual CDRs will lead to long-term savings ranging from about $2 billion to $5.2 billion.16 Table 4: Summary of SSA’s CDR Activities During Special Funding Period, Fiscal Years 1996-2002 10-year estimated Number of CDR Costs savingsb CDRs Estimated (dollars in (dollars in a Fiscal year processed cessations millions) millions) 1996 498,400 26,500 $208 $2,040 1997 690,478 49,700 $330 $3,555 1998 1,391,889 70,300 $462 $4,435 1999 1,703,414 87,300 $547 $5,185 2000 1,836,510 76,000 $609 $5,060 2001 1,730,572 63,600 $603 $4,245 2002 SSA expects to report these data to the Congress in fall, 2003 Source: SSA’s annual CDR reports submitted to the Congress, fiscal year 1996-2001. The law that authorized dedicated CDR funding for SSA between fiscal year 1996 and 2002 (Pub. L. No. 104-121) required SSA to report to Congress for each of those fiscal years specific information such as the amount spent on CDRs and the estimated savings that would result from the cessation of benefits. a Estimated ultimate cessations after all appeals. b Estimates of the reductions in benefit payments for the following programs: the Old-Age, Survivors, and Disability Insurance program; the SSI program; the two Medicare programs, Hospital Insurance and Supplemental Medical Insurance; and the Medicaid program. In addition to a favorable return on investment, SSA’s CDR activities help protect DI and SSI program integrity. Keeping current with the CDR workload can help build and retain public confidence that only qualified 16 Although we did not independently verify these savings estimates, we discussed how SSA made its calculations and believe its approach is reasonable. To estimate long-term savings, SSA calculated the value of the reduction in both cash and medical insurance coverage that otherwise would have been provided to individuals whose benefits were ceased following the completion of a CDR. SSA factored in the effect of appealed cases: SSA did not count savings from those beneficiaries who were initially found ineligible for continued benefits but whose cessations were later successfully appealed. Moreover, SSA officials told us that to estimate savings over 10 years, they took into account the likelihood that some individuals whose benefits were ceased through a CDR would likely have left the disability rolls through death, retirement, and other reasons pertaining to eligibility. Page 16 GAO-03-662 Social Security Disability individuals are receiving disability benefits. In addition, it helps protect the programs’ fiscal integrity and allows SSA to meet its financial stewardship responsibilities. To the extent the agency falls behind in conducting CDRs, a CDR backlog undermines these positive outcomes. While SSA has taken a number of actions over the past decade to Further Opportunities significantly improve the cost-effectiveness of the CDR process, Exist for SSA to opportunities remain for SSA to better use program information in CDR decision making. While DDS personnel study available information on Improve CDR Cost- beneficiaries to decide when they should undergo a CDR, they do not Effectiveness conduct a systematic analysis of this information. As a result, CDRs may not be conducted at the optimal time. Also, SSA’s process for determining what method to use for a CDR—mailer or full medical review—is not always based on the best information available. In addition, SSA has not fully studied and pursued the use of medical treatment data on beneficiaries available from the Medicare and Medicaid programs despite the potential of these data to improve SSA’s selection of the most appropriate CDR method. Finally, SSA continues to be hampered in its CDR decisions by missing or incomplete information on beneficiaries’ case history, which may prevent SSA from ceasing benefits for some individuals who no longer meet eligibility standards. Decisions on Timing of While DDS personnel review available information on beneficiaries to CDRs Are Not Based on establish a diary date indicating when beneficiaries should undergo a CDR, Systematic Analysis of they do not conduct a systematic analysis of this information. Diary decisions are inherently complex because DDS personnel must assess a Available Information beneficiary’s likelihood of medical improvement and how such medical improvement will affect that person’s ability to work. Based on these judgments, beneficiaries are placed in a diary category indicating either that medical improvement is “expected,” “possible,” or “not expected.” DDS personnel then assign a diary date that corresponds with the diary category; the more likely a beneficiary is to medically improve, the earlier the diary date. Although SSA has established guidance for DDS personnel on diary date decisions, SSA officials told us that, ultimately, such decisions are difficult to make and are based on the judgment of the DDS staff. An SSA contracted study of the diary process found that this process is often subjective and that the setting of diary categories and dates is “almost an afterthought” once the case file is developed and a disability determination has been made. SSA’s study identified shortcomings in the diary date process. For example, most beneficiaries assigned to the diary category Page 17 GAO-03-662 Social Security Disability indicating they are expected to medically improve are not found to have improved when a CDR is conducted. Our analysis of SSA data indicates that between 1998 and 2002, only about 5 percent of beneficiaries in the MIE category17 were found to have medically improved to the point of being able to work again. SSA’s diary process study indicated that diary predictions of medical improvement could be substantially improved through the use of statistical modeling techniques similar to those used in the CDR profiling process that SSA uses to determine whether a mailer or a full medical review is needed. The study noted that this systematic, quantitative approach to assigning diary categories and dates would likely enhance disability program efficiency by reducing the number of CDRs that do not result in benefit cessation.18 Another benefit derived from a more systematic approach to diary categorization, according to SSA’s study, is improved integrity of the diary process. Such integrity improvements will result from more timely CDRs and from actual medical improvement rates that more closely correlate with the diary categories that SSA assigns to beneficiaries. For example, SSA’s study indicates that the actual medical improvement rate for beneficiaries assigned to the MIE diary category would increase to about 29 percent under this improved process. SSA officials told us that, in response to the diary study recommendations, the agency has begun to revise its diary process to introduce a more systematic approach to selecting a CDR date. In particular, SSA is developing a process that will use beneficiary data collected at the time of benefit application, such as impairment type and age, in a statistical formula to help determine when a CDR should be conducted. While this change is likely to result in some improvements in the timing of CDRs, the fundamental diary categorization process used by DDSs will remain the same. Despite the study’s findings and recommendations, SSA officials told us that they will not replace SSA’s current process for assigning diary categories with a statistical process because of what they believe would be significant costs involved in changing this system across DDSs. However, SSA’s study acknowledged the potential cost of implementing a new 17 This figure includes all MIE beneficiaries—those who have already undergone a CDR as well as those who have not yet had a CDR. 18 The study recommended that DDSs continue to assign diary categories because this process is useful for indicating the severity of an impairment. The statistical formula would then factor in this DDS diary category in developing an ultimate diary determination. Page 18 GAO-03-662 Social Security Disability process in DDSs, and instead recommended that a revised diary process be centrally administered in order to avoid such high costs. The officials also said that such fundamental changes in the diary process would require a change in regulations. SSA’s Process for SSA’s process for determining what method to use for a CDR is not always Determining CDR Method based on the best information available. In the 1990s, SSA introduced a Not Always Based on Best system that develops a “profile score” for each beneficiary. The profile score indicates the beneficiary’s likelihood for medical improvement Information Available based on a statistical analysis of beneficiary data. The purpose of the profile score is to allow SSA to determine whether it is more cost-effective to send a mailer or to conduct a full medical review. SSA’s own contracted studies indicate that profiling results provide the best available indication of whether a beneficiary is likely to medically improve. Nevertheless, for some beneficiaries, SSA continues to use the diary category that was judgmentally assigned by DDS personnel as the basis for their decision about whether to send a mailer or conduct a full medical review. SSA requires a full medical review for all beneficiaries whose diary category indicates that medical improvement is expected (MIE) and who have not yet undergone a CDR.19 This is the case even when the profile score indicates that improvement is unlikely. In fiscal year 2002, about 14 percent of beneficiaries in the MIE diary category were assigned to the “low” profile category, which indicates that medical improvement is not likely. SSA officials acknowledged that their policy requiring full medical reviews for all beneficiaries in this diary category departs from their usual practice of using mailers for beneficiaries in the low profile category, but they believe that this policy is reasonable given that these beneficiaries are more likely to medically improve than those assigned to other diary categories. However, SSA’s data from 1998 to 2002 shows that most beneficiaries in this category—about 94 percent—do not medically improve to the point of being able to work. 19 SSA applies a different process for MIE beneficiaries who have undergone one or more CDRs. These beneficiaries may receive a mailer if their CDR profile score indicates that they have a low likelihood of medical improvement. However, most beneficiaries assigned to the MIE category have not yet undergone a CDR; in fiscal year 2002, about 88 percent of all beneficiaries in this diary category had not had a CDR. When referring to MIE beneficiaries in the remainder of our discussion in this section, we are describing only those beneficiaries who have not yet had a CDR. Page 19 GAO-03-662 Social Security Disability For other CDR cases, SSA may require that a mailer be sent even when the profile score indicates that conducting a full medical review would be most cost-effective. Specifically, SSA’s policy is to send a mailer to all beneficiaries who were assigned a diary category that indicates medical improvement is not expected (MINE),20 even if the profile score indicates a relatively high likelihood of medical improvement.21 Whether or not these beneficiaries subsequently receive a full medical review will be based on the results of their mailer. SSA officials said that MINE beneficiaries with a high profile score are more likely to receive a full medical review based on their mailer responses because SSA conducts a more stringent review of their mailer responses.22 However, it is not clear that sending mailers to beneficiaries in the high profile category is the most cost-effective approach. SSA studies of the mailer process have indicated that, while this process is effective, it does not provide the same assurance as full medical reviews that medical improvement will be identified. As a result, the use of mailers for beneficiaries whose profile scores indicate a high likelihood of improvement could result in SSA identifying fewer benefit cessations.23 SSA Has Not Fully Studied SSA has not fully studied and pursued the use of medical treatment data and Pursued the Use of on beneficiaries available from the Medicare and Medicaid programs Medical Treatment Data despite the potential of these data to improve SSA’s decisions regarding whether to use a mailer or full medical review to complete a CDR. In from Medicare and 2000, an SSA contracted study found that the use of Medicare data from Medicaid the Center for Medicare and Medicaid Services (CMS)—such as data on hospital admissions and medical treatments—resulted in a significant improvement in SSA’s ability to assess potential medical improvement through CDR profiling. Based on these results, SSA, in fiscal year 2003, 20 SSA officials told us that while it is their intention to do mailers for all MINE beneficiaries, they may be unable in some years to send mailers to all of these beneficiaries if their overall funding for mailers is insufficient. 21 In addition to sending mailers to high profile beneficiaries in the MINE diary category, SSA has recently begun to send mailers to some high profile beneficiaries in the MIP diary category. 22 SSA also sends mailers to medium profile beneficiaries in the MINE diary category. However, SSA has some evidence from its profiling studies indicating that issuing mailers to medium profile beneficiaries is likely to be cost-effective. No similar evidence exists regarding high profile beneficiaries. 23 Although a relatively small proportion of beneficiaries have their benefits ceased based on a CDR, the savings from these benefit cessations are substantial, as noted earlier in this report. Page 20 GAO-03-662 Social Security Disability implemented a process that uses CMS Medicare data in CDR profiling to determine if DI beneficiaries who are initially identified as candidates to receive a full medical review should instead receive mailers.24 SSA expects that this will result in administrative savings due to the reduced number of full medical reviews the agency must conduct. SSA has also initiated a study to assess whether CMS Medicaid data can be used in the same way to decide if SSI beneficiaries, scheduled to receive full medical reviews, could instead be sent mailers. But SSA’s efforts to obtain and use CMS Medicare or Medicaid data are incomplete because the data will only be used to reclassify full medical reviews to mailers but not to reclassify mailers to full medical reviews. SSA officials told us that they have no plans to pursue this additional use of the data because they believe their current profiling system is sufficient for identifying beneficiaries who have a low likelihood of medical improvement. While they agreed that the CMS data could potentially be useful for reclassifying mailers to full medical reviews, they noted that they would need to first study this particular use of the data and would need to develop another interagency agreement with CMS to authorize and obtain data for this purpose. Also, they said that any action to reclassify mailers to full medical reviews would require SSA to publish a Federal Register notice describing this action. SSA could potentially achieve substantial program savings from conducting additional full medical reviews in cases where CMS data indicate that beneficiaries originally identified as mailer candidates have a relatively high likelihood of medical improvement. Using CMS Medicare data for this purpose would be consistent with the results of an SSA study that recommended that these data be used whenever it improves the agency’s ability to accurately predict medical improvement. For example, the study noted that the CMS data would be useful for enhancing SSA’s profiling of beneficiaries with mental impairments, including those with a low likelihood of medical improvement for whom SSA would usually send a mailer. To the extent that CMS data improves SSA’s ability to identify beneficiaries for full medical review, the program savings from reduced 24 SSA is using CMS Medicare data to reassess the prospects of medical improvement for beneficiaries who, based on their initial CDR profiling results, are considered to have a high or medium likelihood of medical improvement. Typically, SSA would conduct full medical reviews for these beneficiaries. However, SSA’s reassessment may indicate that some of these beneficiaries instead have a low likelihood of medical improvement and therefore should receive mailers. Page 21 GAO-03-662 Social Security Disability lifetime benefit payments to those beneficiaries whose benefits are ceased could easily exceed any increased administrative costs resulting from additional full medical reviews. Missing or Incomplete SSA continues to be hampered in its CDR decisions by missing or Case Folders May Result in incomplete information on beneficiaries’ case history, which may prevent Fewer Benefit Cessations SSA from ceasing benefits for some individuals who no longer qualify for benefits. To cease benefits based on a CDR, SSA must determine if the beneficiary has improved by comparing information about the beneficiary’s current condition to information from the agency’s previous decision regarding the beneficiary’s medical condition. This previous decision and the evidence supporting it are recorded by SSA and maintained in case folders that are usually stored in SSA records storage facilities. However, in conducting CDRs, DDSs sometimes have difficulty retrieving the case folders or the key medical evidence that is maintained in these folders. Without the information contained in case folders, DDSs cannot establish a comparison and, therefore, cannot determine if medical improvement has occurred. As a result, SSA is legally required to keep the beneficiary on the disability rolls even though the beneficiary may have been judged to no longer qualify for benefits had the DDS been able to establish a comparison. SSA’s inability to cease benefits in cases where folders are missing or incomplete could result in a substantial cost to the federal government arising from continued payments of benefits—cash and medical—to people who no longer meet eligibility standards.25 Our discussions with SSA officials, survey of DDSs, and review of SSA studies indicate that missing or incomplete folders present an obstacle to effective processing of CDRs. However, evidence on the extent of this problem is mixed. In responding to our survey on CDRs, about 72 percent of DDSs informed us that missing or incomplete information from case folders negatively impacted the quality or timing of CDR decisions to a moderate or great extent. An August 2002 study of missing or incomplete folders conducted by SSA’s Office of the Inspector General reported that DDSs, as well as other SSA components such as field offices, complained 25 Missing or incomplete case folders may also result in additional administrative costs to the extent that SSA and DDS personnel spend time attempting to locate or reconstruct missing information. Page 22 GAO-03-662 Social Security Disability that a large proportion of cases were missing information.26 This study found that case folder retrieval is a significant problem for SSA. Among the problems identified were untimely receipt of case folders, nonreceipt of requested folders, and folders provided without necessary medical evidence. The report questioned SSA’s oversight of folder inventory and retrieval processes and recommended that SSA take various actions, such as independent quality assurance reviews, to improve management of case folders. A study contracted by SSA also identified problems with disability case folder management, such as misrouted or missing folders. The study noted that “inefficient folder management increases administrative and program costs and risks data integrity” and recommended that SSA “analyze the reasons for missing folders and provide recommendations for process and systems improvements.” SSA headquarters officials we spoke with said that SSA has examined the incidence of missing or incomplete case folders and found that the problem is not as significant as claimed by DDSs. For example, in fiscal year 2000, SSA investigated allegations of substantial numbers of missing case folders in two DDSs. SSA officials told us that they were able to locate many of the folders that had been reported as missing. The officials attribute the discrepancy between their findings and the allegations of DDSs, in part, to staff shortages and workload pressures at field offices, which result in a failure of these offices to take further steps to look for folders. However, our survey of DDSs indicates that regardless of SSA’s ability to locate many case folders upon further investigation, DDSs are still having difficulty obtaining the information they need to make CDR decisions. In a 2002 memorandum to SSA’s Inspector General, the SSA Commissioner acknowledged that missing or incomplete case folders are a problem in the CDR process, but noted that the problem had been overstated. The memorandum cited data indicating a lost folder rate of about 0.5 percent for DI CDRs and about 3 percent for SSI CDRs.27 The Commissioner also said that SSA had taken a number of actions in recent years to reduce the incidence of lost folders, such as issuance of additional guidance and training on this issue. In addition, the Commissioner noted that the agency 26 Office of the Inspector General, Social Security Administration, Case Folder Storage and Retrieval at the Social Security Administration’s Megasite Records Center, A-04-99-62006 (Washington, D.C., 2002). 27 Data are based on CDRs conducted from 1997 to 2001. Page 23 GAO-03-662 Social Security Disability was committed to building a system of electronic folders28 that will “virtually eliminate the incidences of lost folders.” While electronic folders may be a key initiative in resolving SSA’s problems with missing or incomplete case folders, SSA does not plan to fully implement this system until mid-2005.29 In addition, these electronic folders will be established only for new disability cases; cases established prior to implementation of electronic folders will remain in a paper format. Therefore, problems in handling these older case folders will likely continue. SSA’s rationale for postponing issuance of a ticket to beneficiaries SSA’s Rationale for expected to medically improve—those who are assigned an MIE diary Postponing Return-to- category—is not well-supported by program experience. In issuing regulations implementing the ticket act, SSA decided to postpone issuance Work Services to of tickets to MIE beneficiaries who have not yet had a CDR based on the Some Beneficiaries Is premise that these beneficiaries could be expected to regain their capacity to work without SSA assistance.30 However, our analysis of SSA data Not Well-Supported indicates that the vast majority of MIE beneficiaries in the DI and SSI by Program programs—about 94 percent—are not found to have medically improved Experience upon completion of a CDR. As a result, some beneficiaries who might otherwise benefit from potentially valuable return-to-work assistance must wait up to 3 years to access services through the ticket program.31 Some disability advocacy groups and SSA’s own Ticket to Work and Work Incentives Advisory Panel have questioned SSA’s policy of delaying the 28 SSA is currently developing a Disability Electronic Folder (EF) which, when completed, will be the repository of all information used in the disability process and should eventually replace the paper folders. As a result, processing components should not have to rely on a paper folder to take adjudicative actions. The EF is planned to be linked to all existing and future systems that support the disability case process. Information will be captured electronically during the case intake process and transmitted to the EF. Documentation and forms received from external sources (e.g., claimants, medical providers, third parties, etc.) will be converted to an electronic format (e.g., scanning and imaging) and added to the EF. Electronic documents received from medical providers will be indexed and added to the EF. 29 SSA plans to begin rollout of electronic disability folders in January 2004 and plans to achieve national implementation over an 18-month period. 30 The Ticket to Work Act gave the SSA Commissioner authority to determine which disabled beneficiaries would be eligible to participate in the ticket program. 31 SSA’s policy on ticket eligibility states that any MIE beneficiary who has been on the disability rolls for at least 3 years will be eligible for a ticket, even if they have not yet had a CDR. Page 24 GAO-03-662 Social Security Disability issuance of tickets to MIE beneficiaries. In particular, they have commented that delaying tickets to all MIE beneficiaries when only a small proportion of these beneficiaries return to work underscores the inherent weakness of relying upon the MIE category as a basis for granting access to ticket services. Furthermore, the ticket panel cited research indicating that the sooner a person with recent work history receives employment services, the more likely the person will be to return to work. In our prior work examining DI and SSI return-to-work policies, we also noted that delays in the provision of vocational rehabilitation services can diminish the effectiveness of such return-to-work efforts.32 Delaying services to some disability beneficiaries, therefore, undermines SSA’s recent efforts to increase its emphasis on helping these beneficiaries return to work. In publishing its final regulations implementing the ticket program,33 SSA wrote that many commenters on the draft regulations had indicated that the agency should provide tickets to all beneficiaries, regardless of their diary category. The commenters also referred to the MIE diary category as an “administrative convenience” that is “not a sufficiently precise tool to deny beneficiaries immediate access to a ticket.” In responding to these comments, SSA wrote that use of the MIE category to identify which beneficiaries should receive tickets “is the most administratively feasible approach currently available to us.” SSA acknowledged that it might be possible to improve the system for identifying such beneficiaries and wrote that it planned to conduct an evaluation to identify possible improvements. SSA officials told us that they are examining the current policy of issuing tickets to MIE beneficiaries to identify possible alternatives but they are not sure when this assessment will be completed.34 However, they noted that their policy of limiting ticket issuance reflects congressional interests in striking an appropriate balance between program stewardship and 32 U.S. General Accounting Office, SSA Disability: Program Redesign Necessary to Encourage Return to Work, GAO/HEHS-96-62 (Washington, D.C.: Apr. 24, 1996). 33 66 Fed. Reg. 67370, Dec. 28, 2001. 34 In May 2003, SSA announced in the Federal Register (Social Security Administration: Semiannual Regulatory Agenda, 68 Fed. Reg. 31240, May 27, 2003) that its long-term plans include a proposal to revise its rules to allow the immediate issuance of tickets to MIE beneficiaries. However, SSA’s Associate Commissioner responsible for reviewing the ticket policy for MIEs told us that SSA has not made a final decision regarding any changes to the current policy and that the agency’s review has not been completed. Page 25 GAO-03-662 Social Security Disability encouraging return to work. Moreover, they explained that reversing the current policy would be costly. SSA’s actuaries have estimated that issuing tickets to all MIE beneficiaries would cost an additional $822 million over 10 years because the ticket law prohibits SSA from conducting CDRs on beneficiaries who are using a ticket. Therefore, SSA would continue to pay DI and SSI benefits to some beneficiaries who might have otherwise had their benefits terminated. The drawbacks of SSA’s current policy of postponing issuance of tickets to MIE beneficiaries and the potential costs associated with an alternative policy that would allow immediate issuance of tickets to these beneficiaries highlights the need for SSA, as part of its policy reexamination, to consider other policy alternatives that might better balance the agency’s program stewardship and return-to-work objectives. While we did not conduct an in-depth assessment of potential alternatives to SSA’s current policy,35 our review of the CDR program and ticket provisions indicate that other options may exist that would achieve a better balance among SSA’s program objectives. For example, SSA could develop a better means of identifying beneficiaries who are expected to medically improve. Earlier in this report, we noted that an SSA-contracted study of the diary process recommended implementation of an improved system that, among other things, would better identify MIE beneficiaries through statistical modeling of diary decisions. One effect of such improved identification, according to the study, would be to substantially reduce the proportion of beneficiaries with an MIE diary category. For instance, the study found that although SSA, over the past decade, has assigned the MIE diary category to about 9 percent of DI beneficiaries, a statistically-based diary process would result in about 3 percent of DI beneficiaries being assigned to the MIE category. This would potentially minimize the number of beneficiaries initially denied tickets and may also provide more assurance, within and outside SSA, that such beneficiaries can truly be expected to improve. SSA might also consider an option that provides for the issuance of tickets to all MIE beneficiaries while allowing CDRs to be conducted as scheduled for these beneficiaries. This policy would require a legislative change because, as we noted earlier, the Ticket to Work Act currently prohibits 35 Given the recent implementation of the ticket program, insufficient data were available during the period of our review to conduct the analysis necessary to fully evaluate such alternatives. Page 26 GAO-03-662 Social Security Disability SSA from conducting a CDR while a person is using a ticket.36 While the ticket program’s prohibition on CDRs for ticket users was intended to remove a potential disincentive for beneficiaries to return to work, MIE beneficiaries currently get neither a ticket nor protection from a CDR. A policy allowing CDRs to be conducted on these beneficiaries while they use a ticket would at least give these beneficiaries immediate access to return-to-work services offered under the ticket program. In addition, SSA would still be able to achieve the cost savings that are derived from CDRs for beneficiaries that it considers most likely to medically improve. Failure to cost effectively process CDRs as they become due could Conclusions negatively affect DI and SSI program integrity. SSA and DDSs are to be commended for bringing the CDR workload current as of the end of 2002. SSA is also to be commended for the improvements it has made in the CDR process. However, a confluence of events, such as the expiration of targeted CDR funding and an increase in initial applications, is increasing the chances of a CDR backlog recurring, which could result in SSA paying out billions of dollars in the long term to beneficiaries who no longer qualify for benefits. In its fiscal year 2004 budget request, SSA has asked the Congress for targeted funding for several program activities, including CDRs, that provide a return on investment. If approved, the targeted funding could increase SSA’s chances of staying current with its CDR workload because this workload would not have to compete internally for funding with the initial determination workload. While SSA has taken a number of steps to improve the CDR process, it has not taken advantage of other opportunities that could further improve the cost-effectiveness of this process and its ability to stay current. In particular, although a more systematic and quantitative process for assigning diary categories and dates would likely improve the timing of CDRs, SSA does not intend to make comprehensive revisions to the diary process based on this more rigorous approach. In addition, despite SSA’s reliance on profiling formulas to improve the agency’s ability to predict medical improvement and benefit cessation, SSA is ignoring or not giving full consideration to information from these formulas in its decisions to conduct mailers or full medical reviews for some beneficiaries. Also, although SSA acknowledges that medical treatment data from Medicare and, possibly, Medicaid improve the agency’s ability to determine when a 36 However, the prohibition on CDRs for all other ticket users could remain in effect. Page 27 GAO-03-662 Social Security Disability mailer should be used, it does not see a need to consider the use of these data to help determine when a full medical review might be preferable. Furthermore, despite long-standing concerns, SSA has not fully addressed the problem of missing or incomplete case folders, which limits SSA’s ability to achieve cost savings through the CDR process. Finally, SSA’s initial assessments of which beneficiaries are most likely to improve are not very accurate and, therefore, may not be the most appropriate criteria to use for delaying beneficiary access to a ticket for return-to-work services. The ticket program is relatively new so little program information is available for SSA to draw upon in reexamining its current policy on ticket access for beneficiaries most likely to improve. SSA has the challenge of developing a policy that will make return-to-work assistance available to beneficiaries at the appropriate time while providing adequate mechanisms for ensuring program integrity. To further improve the cost-effectiveness of the CDR process, we Recommendations to recommend that the Commissioner of SSA take the following actions: the Commissioner of • Pursue more comprehensive enhancements of the CDR diary process— SSA beyond those already being considered—to ensure that the full benefits of a more systematic, quantitative approach to diary setting are attained. Among such key enhancements would be the use of a statistical approach to determine diary categories. Given the significant implications of such changes for the DI and SSI programs, SSA could consider pilot testing the revised diary process before fully implementing it. • Given the cost-effectiveness of conducting mailers in cases where there is a low likelihood for benefit cessation, revise SSA’s policy to allow mailers to be sent whenever appropriate—as indicated by the profiling scores—to beneficiaries with a diary category indicating that they are expected to medically improve. For beneficiaries assigned to a diary category indicating that they are not expected to medically improve, SSA should conduct a thorough analysis of its current policy, which allows mailers to be used for all of these beneficiaries regardless of their profile scores. SSA’s analysis should evaluate the overall cost-effectiveness of this policy, taking into account both the potential reduction in administrative costs from conducting fewer full medical reviews and the potential increase in benefit payments from reduced cessations. If this analysis indicates that the current policy results in higher overall costs for SSA’s disability programs, SSA should revise the policy to make it consistent with the agency’s general profiling approach—which prescribes the use of full Page 28 GAO-03-662 Social Security Disability medical reviews in cases where profiling indicates that a beneficiary has a relatively high likelihood of medical improvement. • Study the use of Medicare and Medicaid data for the purpose of deciding whether to use a full medical review in conducting a CDR for beneficiaries who would otherwise receive a mailer. If found to be cost-effective, SSA should incorporate Medicare and Medicaid data into its CDR process for this purpose. In commenting on a draft of this report, SSA agreed with our Agency Comments recommendations. SSA noted that our review represents a comprehensive and Our Evaluation and accurate assessment of SSA’s accomplishments in improving the CDR process as well as opportunities to improve the process. While agreeing with each of our recommendations, SSA supplied additional information describing its current or planned actions and the basis for such actions. With regard to our recommendation that SSA pursue more comprehensive enhancements of its diary process, SSA said that it is currently studying recommendations made by its contractor regarding the establishment of a statistically-based diary process and that SSA staff will be meeting in the near future to explore implementation options. However, SSA noted it has not yet made a decision regarding implementation. Regarding our recommendation that SSA revise its policies for determining what method to use for a CDR—mailer or full medical review—SSA said that while it generally agreed with our recommendation, it believes we were overly harsh in stating that it is not making the best use of available information. SSA noted that its policy for allowing mailers to be used for all MINE beneficiaries supplements information produced through profiling, thereby improving the process for selecting a CDR method. SSA said that this policy is based on evaluation and analysis of several thousand similar cases and noted that it will verify the cost- effectiveness of this policy through its ongoing integrity reviews. We continue to believe that any departure from SSA’s analytically-based process for using profiling scores to select a CDR method should be based on sound analysis indicating that an alternative process would result in improved cost-effectiveness. We, therefore, are encouraged by SSA’s plans to evaluate the cost-effectiveness of its current policy. However, it is not Page 29 GAO-03-662 Social Security Disability clear that SSA’s integrity reviews37 will be adequate for assessing the cost- effectiveness of the agency’s mailer policy for MINE beneficiaries due to the potential limitations of these reviews. For example, an SSA-contracted study identified several problems with the integrity reviews that SSA conducts for beneficiaries in the low profile category, such as the drawing of integrity samples that are not consistently representative of the mailer population. To the extent that such problems remain unresolved, SSA may need to develop an alternative means of evaluating its mailer policy for MINE beneficiaries. With regard to our recommendation on the use of Medicare and Medicaid data for deciding whether to use a full medical review for beneficiaries who would have otherwise received a mailer, SSA said that it intends to contract for such a study in fiscal year 2004 if funding is available. SSA noted that if the concept is found to be feasible, it will develop a pilot for this approach. SSA also provided additional comments intended to update or clarify some information we provide in this report. In particular, SSA noted that, due to its efforts to keep as current as possible, it believes its CDR backlog by the end of fiscal year 2003 will be significantly less than the potential backlog of 200,000 CDRs that we cited. While there is always a certain degree of imprecision associated with any projection, our backlog figure is based on the best information that was available during our review. We developed our potential backlog figure based on extensive discussions with SSA officials and reviews of SSA’s CDR workload and budget projections. SSA did not provide us with any revised official estimates or analyses that would have led us to revise the CDR backlog figure we report. In addition, SSA said that our report implies that it does not take seriously the shortfall in completing CDRs for the SSI program. SSA is apparently referring to our discussion of the CDR backlog where we note that most of the backlog that is expected to develop by the end of fiscal year 2003 will consist of SSI CDRs, which may make the backlog less problematic than it otherwise would have been because, among other reasons, SSI CDR cessations have lower long-term savings than DI CDR cessations. We did not intend to imply that SSA does not take the SSI backlog seriously. 37 Each year, SSA conducts integrity reviews in which it selects samples of beneficiaries to whom mailers were sent and sends these cases to DDSs for full medical reviews. The results of these full medical reviews are intended to provide a basis for assessing the reliability of using results from profiling formulas to identify appropriate mailer recipients. Page 30 GAO-03-662 Social Security Disability Rather, we included this information to more accurately characterize the nature of the potential backlog because that could provide important insights as to how to deal with it. Finally, SSA said that although our survey of DDS directors indicates that attrition among disability examiners is an issue for DDSs, SSA and DDSs are accustomed to dealing with such issues and DDSs are still able to complete their workloads. Although we are aware that DDSs regularly confront multiple challenges to completing their disability program workloads, we cannot ignore the clear implications of DDS directors’ answers to our survey questions. Given that a clear majority of DDS directors indicated that disability examiner attrition is somewhat or very likely to jeopardize their ability to complete their CDR workload, we believe that it is important for us to identify this issue as a potentially significant factor in the possible development of a CDR backlog in the years ahead. SSA’s comments appear in appendix II. SSA also provided additional technical comments that we have incorporated in the report, as appropriate. Copies of this report are being sent to the Commissioner of SSA, appropriate congressional committees, and other interested parties. The report is also available at no charge on GAO’s Web site at http://www.gao.gov. If you have any questions about this report, please contact me at (202) 512-7215. Other contacts and staff acknowledgments are listed in appendix III. Sincerely yours, Robert E. Robertson Director, Education, Workforce, and Income Security Issues Page 31 GAO-03-662 Social Security Disability Appendix I: Scope and Methodology Appendix I: Scope and Methodology To evaluate the impact of the expiration of separate funding for continuing disability review (CDR) processing and the level of funding needed to remain current with the CDR caseload, we interviewed Social Security Administration (SSA) officials from the Office of Budget, the Office of the Chief Actuary, the Office of Disability and Income Security Programs at SSA headquarters in Baltimore. We also reviewed SSA documents, including the agency’s budget request and estimates of the cost and savings from conducting CDRs. In addition, we surveyed Disability Determination Services (DDS) directors to assess the potential effect of the expiration of special CDR funding on DDS operations. To develop the survey, we identified information that would help us address the research questions. We generated specific survey items by reviewing SSA CDR reports submitted annually to the Congress and drawing upon interviews we conducted early in the assignment with SSA officials and the National Association of Disability Examiners. We validated our survey instrument by obtaining feedback from SSA officials and pretesting it with several current DDS directors. In consultation with SSA, we excluded 2 of the 54 DDSs from our survey as well as the federal DDS. The two DDSs excluded were Guam and South Carolina’s DDS serving persons who are blind–both relatively small DDSs that are run by one person. We excluded the federal DDS because responses from this site might skew results as the site (1) is a federal entity and as such it is “different” than other DDSs, (2) is used to process the overflow of CDRs, and (3) serves as SSA’s test unit. The remaining 52 DDSs (essentially 1 for each of the 50 states, plus the District of Columbia and Puerto Rico) comprised the study universe. All 52 of these DDSs responded to our survey. To assess the opportunities for SSA to improve CDR cost-effectiveness and to examine SSA’s rationale for delaying return-to-work services to some beneficiaries under the ticket to work program, we interviewed SSA officials from the Office of Disability and Income Security Programs and the Office of the Chief Actuary at SSA headquarters in Baltimore. We also reviewed legislation, regulations, and SSA policy guidance related to the CDR and the ticket programs. In addition, we examined various studies and reports on the CDR and ticket to work programs, including reports from SSA’s Office of the Inspector General and a wide range of contractor- produced reports analyzing the cost-effectiveness of the CDR process. Finally, we analyzed data from SSA on the number of adult DI and SSI beneficiaries, aged 19-64, assigned to various CDR diary and profiling categories and the CDR outcomes—cessation or continuance—for these beneficiaries. These data were derived from SSA administrative data sets Page 32 GAO-03-662 Social Security Disability Appendix I: Scope and Methodology used by the agency to select cases for review and to track the results of CDRs. We did not independently verify these data, but based on comparison to SSA’s previously published data, and discussion of minor discrepancies with SSA officials, we determined that they were sufficiently reliable for our purposes. We performed our work in accordance with generally accepted government auditing standards between August 2002 and May 2003. Page 33 GAO-03-662 Social Security Disability Appendix II: Comments from the Social Security Administration Appendix II: Comments from the Social Security Administration Page 34 GAO-03-662 Social Security Disability Appendix II: Comments from the Social Security Administration Page 35 GAO-03-662 Social Security Disability Appendix II: Comments from the Social Security Administration Page 36 GAO-03-662 Social Security Disability Appendix II: Comments from the Social Security Administration Page 37 GAO-03-662 Social Security Disability Appendix III: GAO Contacts and Staff Appendix III: GAO Contacts and Staff Acknowledgments Acknowledgments Shelia D. Drake, Assistant Director (202) 512-7172 GAO Contacts Brett S. Fallavollita, Analyst-in-Charge (202) 512-8507 The following individuals also made important contributions to this report: Staff Mark Trapani, Melinda L. Cordero, and Corinna A. Nicolaou. Acknowledgments (130194) Page 38 GAO-03-662 Social Security Disability The General Accounting Office, the audit, evaluation and investigative arm of GAO’s Mission Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO’s commitment to good government is reflected in its core values of accountability, integrity, and reliability. The fastest and easiest way to obtain copies of GAO documents at no cost is Obtaining Copies of through the Internet. GAO’s Web site (www.gao.gov) contains abstracts and full- GAO Reports and text files of current reports and testimony and an expanding archive of older products. The Web site features a search engine to help you locate documents Testimony using key words and phrases. You can print these documents in their entirety, including charts and other graphics. Each day, GAO issues a list of newly released reports, testimony, and correspondence. GAO posts this list, known as “Today’s Reports,” on its Web site daily. The list contains links to the full-text document files. To have GAO e-mail this list to you every afternoon, go to www.gao.gov and select “Subscribe to daily E-mail alert for newly released products” under the GAO Reports heading. Order by Mail or Phone The first copy of each printed report is free. Additional copies are $2 each. A check or money order should be made out to the Superintendent of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more copies mailed to a single address are discounted 25 percent. Orders should be sent to: U.S. General Accounting Office 441 G Street NW, Room LM Washington, D.C. 20548 To order by Phone: Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202) 512-6061 Contact: To Report Fraud, Web site: www.gao.gov/fraudnet/fraudnet.htm Waste, and Abuse in E-mail: firstname.lastname@example.org Federal Programs Automated answering system: (800) 424-5454 or (202) 512-7470 Jeff Nelligan, Managing Director, NelliganJ@gao.gov (202) 512-4800 Public Affairs U.S. General Accounting Office, 441 G Street NW, Room 7149 Washington, D.C. 20548
Social Security Disability: Reviews of Beneficiaries' Disability Status Require Continued Attention to Achieve Timeliness and Cost-Effectiveness
Published by the Government Accountability Office on 2003-07-24.
Below is a raw (and likely hideous) rendition of the original report. (PDF)