Changing Circumstances:  Experiences of Child SSI Recipients Before and After Their Age-18 Redetermination for Adult Benefits

 

Jeffrey Hemmeter, Social Security Administration

Jacqueline Kauff, Mathematica Policy Research, Inc.

David Wittenburg, Mathematica Policy Research, Inc.

 

February 26, 2008

 

 

 

 

 

 

We are grateful for the editorial and production assistance of James Cameron, Molly Cameron, Sharon Clark, and Cindy McClure and for feedback from Tom Fraker, Thomas Golden, Arun Karpur, Connie Marion, Kalman Rupp, David Stapleton, and Robert Weathers. 

 

Work on the article was supported by a grant to Cornell University from the Department of Education’s National Institute on Disability and Rehabilitation Research under its Rehabilitation Research and Training Center on Employment Policy for Persons with Disabilities (Cooperative Agreement No. H133B040013). The opinions, conclusions, and errors in this paper are the sole responsibility of the authors, and do not represent the official views of the U.S. Department of Education, Social Security Administration, Mathematica Policy Research, Inc., or Cornell University and do not assume endorsement by the Federal Government (Edgar, 75.620 (b)

 

Abstract

 

This paper provides an analysis of the dynamics of the transition of child Supplemental Security Income (SSI) recipients into adulthood using linked 2001-2002 National Survey of Children and Families (NSCF) survey and Social Security Administration (SSA) administrative data. We examine the interaction of impairment status, reported health needs, and other self-reported indicators of human capital on SSI program and employment outcomes after age 18.  Our primary objective is to examine the differences in pre-age-18 individual characteristics across subgroups of recipients by impairment status and determine whether these differences influence post-age-18 SSI participation and employment outcomes.  We find that after controlling for measures of disability severity, duration, and human capital, youth with other mental and behavioral disorders are much less likely to receive SSI at age 19.  The findings also suggest that non-health factors, particularly education, employment, and social indicators, play an important role in the probability of a child SSI recipient being on adult SSI after age 18.  Our findings indicate that, while some youth appear to be making a successful transition from child SSI benefits to adult benefits or other activities (off of SSI), others appear to have limited prospects for long-term self-sufficiency. A major concern is that some youth no longer on SSI after age 18, particularly those with other mental and behavioral disorders, may not have been sufficiently prepared for life without SSI and might not have enough human capital or other supports to become self-sufficient during their adult years.

 

Introduction

The child Supplemental Security Income (SSI) program, a means-tested cash benefit for children with severe disabilities, provides an important source of income for youth under the age of 18.  As mandated in the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996, child SSI recipients must have their eligibility for SSI redetermined under the adult eligibility criteria once they reach age 18.  Approximately one-third lose eligibility during this redetermination because they do not meet the adult SSI disability criteria or leave the program for other reasons (Social Security Advisory Board [SSAB] 2006). 

The age-18 redetermination is a significant event in the lives of child SSI recipients and their families, and is a source of controversy among advocates, members of the research community, practitioners, and public officials. Central to the debate is whether it is appropriate for youth to be redetermined under the adult criteria at age 18.  Some argue that Congress should extend the redetermination past age 18 to allow youth to obtain enough skills to find jobs that would enable them to live independently, as well as to be consistent with other federal policies for youth with disabilities.  This argument is based on the premise that under the Individuals with Disabilities Education Act children are entitled to receive special education services through age 21—recognizing the need for important educational and other supports leading to post-school success.  However, extending the age of the redetermination process would reverse the intent of the PRWORA legislation that emphasized the importance of distinguishing between child and adult disability status to address concerns regarding rapid program growth.  Additionally, the effect that extending the age of redetermination may have on child recipients’ accumulation of human capital (e.g., education, training, and early work experience) is unclear.  Some might be more likely and others less likely to accumulate such capital. 

The debate over the age-18 redetermination has been hampered because of limited data on the characteristics and transition outcomes of child SSI recipients after redetermination. Some studies have shown substantial variation in the demographic, family, and social characteristics within the child SSI population, particularly across those who are on versus off of adult SSI benefits after age 18 (Rogowski et al. 2002; Wittenburg and Loprest forthcoming; Loprest and Wittenburg 2007).  However, the information is more limited regarding how individual characteristics of child SSI recipients prior to age 18—especially impairment and pre-redetermination activities (e.g., schooling or employment)—influence program and employment outcomes in adulthood. 

This paper provides a more comprehensive analysis of the dynamics of the transition process of child SSI recipients by examining the interaction of impairment status, reported health needs, and other self-reported indicators of human capital on SSI program and employment outcomes after age 18.  Our primary objective is to examine the differences in pre-age-18 individual characteristics across impairment subgroups of recipients and determine whether these differences influence post-age-18 SSI participation and employment outcomes.  We examine these dynamics using linked 2001-2002 National Survey of SSI Children and Families (NSCF) survey and Social Security Administration (SSA) administrative data.  While the paper does not address whether a redetermination at age 18 is appropriate, it does provide insights about which recipients are more likely to remain on or no longer receive SSI after the redetermination, how child recipients have fared after redetermination, and differences in experiences across subgroups. 

KEY FINDINGS IN BRIEF

Results show that the SSI program and employment outcomes following the age-18 redetermination vary across impairment categories, although we find that the probability of no longer receiving SSI after age 18 correlates with several other pre-age-18 individual and human capital characteristics.  Consistent with published statistics, child SSI recipients with mental and behavioral impairments other than mental retardation are least likely to remain on SSI at age 19, relative to those with other impairments.  Additionally, the probability of remaining on SSI inversely correlates with functioning and pre-age-18 education, social, and employment characteristics.  We find that after controlling for measures of disability severity, duration, and human capital, youth with other mental and behavioral disorders are much less likely to receive SSI at age 19.  The findings also suggest that non-health factors, particularly education, employment, and social indicators, play an important role in the probability of a child SSI recipient being on adult SSI after age 18.  Of particular concern is that many youth who are off SSI at age 19 report a high arrest history prior to age 18, which could indicate that some youth are “falling through the cracks’ of the system and left with no income support and poor employment prospects.  While some youth appear to be successfully transitioning from child SSI benefits to adult benefits or other activities, some consideration might be given to extending supports for those who appear to have limited prospects for long-term self-sufficiency, particularly those with mental impairments. 

Background

The child SSI program, administered by SSA, provides an important source of cash benefits to low-income families.  In 2008, the maximum federal SSI payment for a child was $637 per month (approximately 75 percent of the poverty line for an individual), and many states provided a separate supplement to the federal payment, ranging from a few dollars to approximately $150 per month.  That year, there were 1.1 million SSI recipients under the age of 18 (SSA Office of Policy 2007).  There is no guarantee that a child’s benefits will continue after age 18, however, because eligibility rules for children and adults differ.  For youth with disabilities, the transition from childhood to adulthood can be complicated further by disparate support systems and confusing incentives and disincentives for developing human capital during the transition period. 

SSI Eligibility Criteria Differ for Children and Adults and, at Age 18, All Child Recipients Must be Redetermined for SSI Under the Adult Criteria

To qualify for SSI, children and their families must meet strict income, asset, and disability eligibility criteria. The income and asset criteria are based on a complex set of rules that require the parent’s (or guardian’s) and child’s income and assets to fall below a certain threshold.  To meet the disability criteria, a child must have “a medically determinable physical or mental impairment which results in marked and severe functional limitations, and which can be expected to result in death or which has lasted or can be expected to last for a continuous period of not less than 12 months” (U.S. Code 42 2007).  Once on the rolls, recipients must continue to satisfy the strict income, asset, and disability eligibility criteria to continue receiving benefits.  SSA is mandated to conduct Continuing Disability Reviews (CDRs) at least once every three years to assess the child’s eligibility, unless there is a determination that the child’s health condition is not expected to improve.

       Upon reaching age 18, child SSI recipients must undergo a redetermination of eligibility, using the adult definition of disability, to continue receiving SSI benefits.  Since 1996, SSA has been required to conduct redeterminations within one year of a child’s 18th birthday, although this regulation was relaxed in 2006 by Public Law 105-33 to allow for longer periods if necessary. Unlike the child criteria, the adult disability criteria is based on a person’s ability to perform work at a substantial gainful activity level (SGA), defined in 2008 as $940 per month.  If the decision is unfavorable, SSI eligibility ends the last day of the second month following the month of the decision, and individuals are notified and made aware of their right to appeal.  Between 1996 and 2004, more than 400,000 recipients had age-18 redeterminations (SSAB 2006) and after appeals, approximately one-third of the redeterminations resulted in a cessation of benefits.[1]  Rogowski et al. (2002) found eligibility differences across subgroups; children who had a diagnosis of mental retardation were much more likely to be found eligible as adults, relative to those with other diagnoses.

The Age-18 Redetermination Process has Important Implications for a Youth’s Human Capital Decisions

The uncertainty of the age-18 redetermination might influence a youth’s decision to seek education, training, and work skills prior to age 18, as well as family decisions regarding work and investment in human capital (Wittenburg and Loprest 2003; Auxter, Halloran, Berry, and O’Mara 1999).  Recipients who believe they are likely to lose benefits at redetermination might be more likely to accumulate human capital (e.g., work experience or schooling), while those who think that they will remain on SSI after age 18 might be less likely to do so for fear of jeopardizing future benefits.  Similarly, parents might be wary of working because their earnings could jeopardize their children’s benefit levels and potential Medicaid coverage. 

       In part because of these concerns, the SSAB (2006) recommended that SSA take an active role in informing educators, parents, and caregivers about the implications of the age-18 redetermination, and in broadening the opportunities for further human capital development.  Specifically, they recommended that SSA begin a process of notifying SSI youth and their parents or caregivers about the age-18 redetermination as early as age 14.  They also recommended expanding the use of work incentives (e.g., the student earned income exclusion and Section 301 protections that expand eligibility for youth in training and vocational rehabilitation activities) to promote early development of human capital.

The Transition to Adulthood, Particularly for Those Receiving SSI, Can Be Complicated Further By Disparate Support Systems

While they are in school, students with disabilities, including many SSI recipients, receive services from one provider—the school system—and those services typically are provided in a relatively organized manner.  However, as students with disabilities exit the school system, they face an array of service providers that operate with different eligibility requirements and usually are not located under one roof.  While publicly supported education services are an entitlement for youth with disabilities until they receive a diploma or turn age 22 (or older, at state option), most other state services cease upon attainment of one of those milestones, and these youth must then enter the adult support system.

The transition to adulthood can be problematic because access to many adult services regarded as essential for successful transition are not entitlements, including vocational rehabilitation (VR), developmental disabilities, and mental health services.  Long waiting lists for these services are common (Stancliffe and Lakin 2005).  In addition, eligibility determination processes vary by state so the initiation of adult services for youth in transition may be timely in one state and significantly delayed in another.

Data Description

There is limited information on how the interaction of individual impairment, demographic, and family characteristics influence longitudinal program and employment outcomes.  SSA data include detailed program information that can be used to track recipient characteristics and program outcomes, as was done in Rogowski et al. (2002), but these data do not include information on the youth’s family characteristics (e.g., income or family structure), school activities, and participation in other activities (e.g., VR) that might influence these experiences.  Conversely, the NSCF includes more detailed information on these missing characteristics, but does not include some of the longitudinal information contained in SSA administrative records.   In part because of the limited information on outcomes of child SSI recipients by impairment subgroups, the SSAB (2006) suggested that SSA use restricted files of the NSCF that could be linked to SSA program and earnings data to examine further the relationship between pre-age-18 recipient characteristics and activities and post-age-18 SSI participation and employment outcomes.

Following this SSAB recommendation, our analysis uses restricted access data from SSA that include linked survey data from the NSCF and historical SSA administrative program and earnings records to examine program and other outcomes among child SSI recipients facing the redetermination process.  The combination of survey and program administrative records also allows us to examine dynamic differences in outcomes across important subgroups. 

Linked NSCF-SSA Administrative Files

The NSCF is the first nationally representative survey of current and former SSI recipients and applicants conducted since 1978.  It includes detailed information on the program, work, and education experiences of current and former child SSI recipients unavailable from any other source (Ireys et al. 2004).  SSA contracted with Mathematica Policy Research, Inc. (MPR) to conduct the NSCF, which was fielded between August 2001 and July 2002.  There were two goals of the NSCF: (1) to evaluate the effects of welfare reform on child SSI recipients and their families, and (2) to provide information on the characteristics, experiences, and needs of the current cross-section of SSI child recipients and their families.  Thus, the target population comprised two cohorts of children (ages 0 to 17) and young adults (ages 18 to 23) in the SSI applicant and recipient administrative program files who were eligible or had applied for SSI either as of December 1996 or December 2000.  Children who were not recipients at either of these times also were eligible for the survey if they had previously been recipients or had applied for SSI after January 1, 1992.  A complex clustered sampling design was used to oversample certain subgroups, including youth approaching the redetermination age either in 1996 or 2000. 

The final NSCF sample included eight sampling strata that could be weighted to be nationally representative of current and former child SSI recipients from the 1996 and 2000 cohorts.  After extensive locating efforts—which included identifying respondents from administrative records, community database searches, and other contacts (e.g., information from relatives or neighbors)—MPR interviewed 73 percent of the sample (8,726 eligible respondent households) by telephone or in person.  Rupp et al. (2005/2006) provides a profile of all children included in the NSCF.

SSA linked the NSCF data to three SSA administrative data sources that can be used to track longitudinal program earnings and death outcomes of recipients: the Supplemental Security Record (SSR), the Master Earnings File (MEF), and the Numident file.  The SSR contains the program characteristics, monthly SSI eligibility, and payment history of individuals since the SSI program began in 1974.  The MEF contains yearly Social Security-covered earnings from the Internal Revenue Service’s tax records.  Because not all earnings are covered by Social Security, our estimates of employment and earnings are for the covered sector only.  The most recent year for which earnings data are available is 2004, which provided up to three years of earnings information following the NSCF interview.  Finally, the Numident file stores the Death Master File of the U.S. government and provides the most accurate date of death available for individuals.

Sample Selection and Weighting

Our analysis sample includes two cohorts of NSCF respondents who were receiving child SSI payments one year prior to their age-18 redetermination.  We include members of the NSCF sample who were: (1) age 17 or 18 at the time of the survey (2001 or 2002) and (2) receiving SSI payments in the January during which they were age 17 (according to administrative records that range from 1999 to 2003).  Selecting the sample based on SSI benefits in January allows us to examine earnings, which are measured only on an annual basis, and the restriction that recipients were receiving benefits at age 17 ensures that we selected a population of eligible recipients just before their age-18 redetermination.  We selected this age group because these recipients presumably are at an age to make important decisions regarding their future participation in SSI as adults and, for many, decisions about transitions from secondary school. Additionally, recipients ages 17 to 18 were oversampled in the NSCF 2000 cohort, providing a sufficient sample for tracking pre- and post-redetermination outcomes. 

We use the sample weights included in the NSCF file to ensure that the results would be representative of child SSI recipients facing redetermination in their respective cohorts. These weights ensure that the sample is representative of the SSI populations in either December 1996 or December 2000.  Because we use multiple cohorts, and there was a lag between when the sample was selected and when the survey was administered, it is difficult to assess whether our selected sample represents any one cohort.  Nonetheless, we have no reason to believe that our selected sample excludes any particular subgroup; it should represent the activities of all SSI recipients in the year prior to their redetermination. 

We present standard errors for all estimates, using balanced repeated replicate weights created by MPR and estimated using Stata’s complex survey estimation procedures, which take into consideration the complex sampling design of the survey (see Gillcrist and Edson [2004] for a more detailed description of the NSCF’s survey design).  Unless otherwise noted, any differences described below in our descriptive comparisons are statistically significant at the 95 percent confidence level or above. 

Methods

We use the linked data to examine the characteristics of impairment subgroups of SSI recipients prior to age 18 and their adult SSI and employment outcomes after their age-18 redetermination.  These data allow us to analyze how the influences of the combination of impairment characteristics and individual and family characteristics cited in earlier studies influence program outcomes.   Our findings provide insights on how multiple factors, especially social factors not previously examined in this context, influence outcomes after age 18, and whether differences exist across subgroups of recipients.  In Appendix Table A.1, we provide a complete listing and definitions of the variables included in our analyses.

We begin by comparing the characteristics of child SSI recipients prior to age 18 across four SSA impairment subgroups, based on the primary impairment code that SSA used to award the original child SSI benefit.  One limitation of these codes is that the primary impairment that qualifies someone for benefits might not necessarily reflect their principal disabling condition or secondary conditions that affect their ability to function.  Nonetheless, these codes provide some important information on medical conditions of the youth, and are used by SSA to characterize and classify recipients.  A major advantage of this approach is that our findings can be used to inform policy options for subgroups that can be identified readily using administrative data.  This feature is especially important, given such recent SSA initiatives as the Mental Health Treatment Study, which target return to work services for adult SSA disability recipients with specific SSA impairment codes. 

We create the following four subgroups:  (1) mental retardation, (2) other mental and behavioral disorders, (3) systems disorders, and (4) other disabilities.  The mental retardation category is the largest administrative impairment subgroup within the child SSI population, followed by the other mental and behavioral disorders category.  The final two categories include youth whose impairments are more physical (see Table A1 for a full description of primary impairments in each category).  Rogowski et al. (2002) showed that differences exist across these impairment groups with respect to program outcomes, and findings from the special education literature indicate that differences likely will exist across these impairment groups with respect to employment and social outcomes.  Cameto (2005), for example, found differences in employment rates across impairments groups among former special education students, and Wagner (2005) found large differences in social outcomes across impairment groups, particularly in the probability of being arrested.   

To examine differences in program outcomes after age 18, we compare the characteristics of recipients who remained on SSI at age 19 (on-SSI) to those who were off at age 19 (off-SSI).  The on-SSI group included sample members who were receiving SSI benefits in January of the year they turned 19, while the off-SSI group included sample members who were not.  We have chosen this age because most age-18 redeterminations presumably should be completed by this birthday.[2]  Our findings after age 18 provide some indication of the immediate relationship between the post-age-18 program and employment outcomes and pre-age-18 characteristics that we use to characterize the transition process for different impairment groups.  Based on the findings from the special education literature and Rogowski et al. (2002), we anticipate that some substantial differences might exist in program and employment outcomes across our four impairment subgroups, especially for those with physical and mental impairments.

We further examined the relationship between the pre-age-18 characteristics of child SSI recipients and their SSI participation after age 18 using logistic regressions. This analysis will help policymakers better understand the complex underlying relationships between the various personal characteristics affecting post-age-18 program participation. We estimate several models in which we sequentially add various sets of control variables to examine how the probabilities of post-age-18 SSI participation changed for the four impairment subgroups identified in our descriptive analyses.  If impairment alone explains all of the variation in SSI outcomes, we would expect the coefficients for the impairment subgroups to remain unchanged across model specifications.  However, if other demographic, health, functioning, and social factors influence these probabilities, we would find that the effects of impairment should diminish as more controls are added to our model. 

We estimate the following logistic model:

(Equation:  The equation is a logistic regression where the dependent variable is equal to 1 if the child is receiving SSI at age 19.  The independent variables includes impairment, age of program entry, self report health status, a vector of human capital characteristics and a vector of demographic and family characteristics.)

 

Where:

·      Yi is an indicator for whether the child i is receiving SSI benefits at age 19.

·      is a set of dichotomous variables for a child i’s impairment status based on SSA program records in 2000, at the time the child was selected for inclusion in the NSCF.  The categories are mental retardation, other mental and behavioral disorders, and other disabilities. (The systems disorders category, including impairments of the nervous system and sense organs, is the excluded category).

·      Age of Program Entryik is a set of dichotomous variables for a child i’s age at first SSI receipt. The categories include ages 10-13 and 14-17 (less than 10 years old is the excluded category).

·      Demographic and Family Characteristicsin is a set of dichotomous variables that represent the demographic and family characteristics of child i at the time of the NSCF interview.  The variables include: (1) living in a two-parent family, (2) male, (3) white, and (4) age 18 at the time of the survey.

·      Self-Reported Healthil Characteristics is a set of dichotomous variables that represents the self-reported health status of child i based on NSCF questions.  The eight selected variables include: (1) needing or using more services than others the same age; (2) needing or using medicine prescribed by a doctor; (3) limited in ability to do the things others do at the same age; (4) needing or getting special therapy; (5) needing or getting mental health treatment or counseling; (6) using special medical equipment; (7) needing help with personal care; and (8) needing help with handling routine needs.

·      Human Capital Characteristics im is a set of dichotomous variables that represents the human capital characteristics of child i prior to age 18, based on survey responses and administrative data.  The variables include: (1) employment at ages 16 or 17; (2) ever being arrested; (3) ever being suspended or expelled from school; (4) currently enrolled in school; (5) ever in special education; (6) ever received vocational training; (7) ever participated in VR (as measured by having an Individual Written Rehabilitation Plan [IWRP]); and (8) ever earned more than $2,000 at ages 16 or 17.

We present the findings as the odds ratios of logistic regression results, rather than as coefficients. The odds ratio is the odds of an event occurring when the corresponding variable increases by one unit, divided by the odds of the event occurring had the corresponding variable not increased by one unit.

We conclude with an analysis of the employment and earnings experience after age 18 to determine whether former child SSI recipients are entering the labor market and, for those no longer on SSI, replacing their child SSI benefits with earnings.  Presumably, youth no longer on SSI should have higher employment and earnings relative to those on SSI after age 18, given that the off-SSI group needs an alternative source of income to replace SSI benefits.

Characteristics of the Research Sample

The research sample of child SSI recipients shows impairment, demographic, and family characteristics similar to those found in previous studies (Table 1).  The sample included 730 NSCF respondents representing a weighted total of 97,106 child SSI recipients who were age 17 in the 1999-2003 SSI cohorts.  At the time of the NSCF interview, 44 percent of our sample was age 17.  Consistent with Wittenburg and Loprest (forthcoming) and Rogowski et al. (2002), mental disorder was the primary impairment for more than 75 percent of child recipients in our sample; 50 percent had a primary diagnosis of mental retardation, and 26 percent had a primary diagnosis of another mental or behavioral disorder, including affective, anxiety, schizophrenia, and other mental and behavioral disorders.  Approximately two-thirds (63 percent) of our sample were male, less than half were white, and more than half lived in single-parent families.  Most child SSI recipients (84 percent) started benefit receipt prior to age 13, and the majority of those were receiving benefits before age 10.  Consequently, most of the research sample had been receiving benefits for at least four years prior to their NSCF interview, and the majority had more than seven years of SSI participation.

Characteristics of Child SSI Recipients Prior to Age 18 BY IMPAIRMENT STATUS

Potential barriers for child SSI recipients facing the transition to adulthood include challenges associated with their impairment (e.g., needed medications), social issues, access to a coordinated set of services and supports, and difficult choices regarding human capital decisions, especially work, that could jeopardize their long-term SSI eligibility.  These barriers could result in choices that compromise their ability to complete successful transitions, including an under-investment in human capital and poor social development.  Their transition outcomes may have serious long-term implications, limiting youth with severe disabilities to long-term dependence on SSA disability benefits.

The barriers faced by youth vary substantially by impairment category.  Below, we examine differences in demographic and family characteristics, functioning and disability characteristics and indicators of human capital development across SSI impairment subgroups that might influence short- and long-term outcomes.  These findings provide some context for the discussion of transition and post-age-18 outcomes discussed in later sections.

Demographic and Family Characteristics

There were large differences in the gender, racial, and family characteristics across the four impairment groups that potentially could influence post-age-18 program outcomes (Table 2).  Relative to the child SSI population as a whole, the systems disorders subgroup had fewer males (50 percent vs. 63 percent) and were more likely to be white (56 percent vs. 47 percent); the mental retardation subgroup was less likely to be white (41 percent vs. 47 percent); and the other mental and behavioral disorders group was less likely to be living in a two-parent family (18 vs. 23 percent). 

Health Needs and Disability Onset Characteristics

To capture differences in health needs that might influence the use of transition services, we analyze a set of screening questions included in the NSCF to identify children with special health care needs.  These questions were developed by the Foundation for Accountability (FACCT), and identify children who satisfy any one of five conditions where the condition is expected to last, or has lasted, 12 months or longer and is due to a medical, behavioral, or other health condition. The conditions are (1) need or use more services than others of the same age; (2) need or use medicine prescribed by a doctor; (3) are limited in their ability to do the things that others of the same age do; (4) need or get special therapy; or (5) need or get mental health treatment or counseling. We also analyze survey questions on the use of special medical equipment, and the need for help with personal care or handling routine needs because of a health condition.

Most child SSI recipients reported a special health need according to the FACCT screener (87 percent), which is consistent with the SSI program’s stringent disability definitions (Table 3).  More than two-thirds needed or used more services than others the same age, and almost as many felt limited in their ability to perform age-appropriate tasks (62 percent).  If the FACCT screener is a reasonable proxy for health severity, then the 13 percent of child SSI recipients who did not report any special health needs should be more likely to be off of SSI after age 18 than other groups.  In addition, many child SSI recipients reported the need for other health supports not covered in the FACCT screener.  Most needed help in handling a routine need (72 percent) and a minority reported more intensive needs for personal care (23 percent) or special medical equipment (11 percent).

There was variation in the types of specific health care needs within the impairment subgroups, although a large portion of all groups reported a need for mental health treatment.  Youth with other mental and behavioral disorders were most likely to report a need for mental health treatment, while those with physical disabilities (systems disorders or other disabilities) were more likely to use special medical equipment and personal care.  At least 40 percent of each impairment subgroup reported a need for mental health treatment/counseling, indicating that the need for mental health services extends beyond the mental retardation and other mental and behavioral disorders categories. 

The age at which individuals entered SSI also could influence post-age-18 program outcomes if recipients who enter the program earlier differ from those who enter at later ages.  Presumably, youth who enter SSI as young children have more easily identifiable impairments at early ages that make them program eligible and so are more likely to remain eligible after age 18 relative to groups identified at older ages.  Additionally, younger child SSI recipients also will have passed more CDRs than older children, possibly indicating more severe levels of impairment relative to those who enter the program at older ages. 

Our findings indicate that those with beneficiaries who had more physical disabilities and identifiable disorders—mental retardation, systems disorders, and other disabilities—had much earlier ages of first SSI receipt relative to the other mental disorders group (Table 3).  Only 39 percent of child SSI recipients with other mental and behavioral disorders qualified prior to age 10, compared to at least 60 percent of child SSI recipients in the other disability categories.  The later entry into the SSI system indicates that child SSI recipients with other mental and behavioral disorders have a briefer history of SSI participation, which could affect their decisions after age 18.

Human Capital Investment Characteristics

Youths’ participation prior to age 18 in human capital activities, including employment, education and training, likely plays a role in their decisions to pursue adult SSI benefits after age 18.  Presumably, youth who are more heavily invested in human capital activities will have a greater likelihood of no longer receiving benefits after age 18.  The importance of a progression of work-based experiences in preparing for independence, as well as the correlation of employment training and work experience with post-school employment success, is well documented for youth in special education (Wagner et al. 2006; Luecking and Fabian 2000).  However, as noted above, the structure, quality, and availability of education and training services can vary across states and localities so that not all individuals have equal access to services and supports to develop their human capital.  Negative social experiences prior to age 18, including dropping out of school, school problems, and arrests, also are likely to play a role in post-age-18 SSI employment and program decisions. 

Employment rates of child SSI recipients are substantially lower than those of other youth with and without disabilities, which indicates that many might not be developing sufficient human capital.  We find that 41 percent of child recipients facing the age-18 redetermination had some earnings between the ages of 16 and 17 (Table 4).  Comparatively, the National Research Council and Institute of Medicine (Committee on the Health and Safety Implications of Child Labor 1998) found that approximately 80 percent of all youth are employed at some point during their high school years, and Wagner et al. (2006) found that almost 60 percent of all youth in special education programs (ages 13 to 16) are employed during a one-year period.  The relatively low rates of employment among child SSI recipients may reflect their unique characteristics (e.g., youth with especially severe impairments), a reluctance to work based on fear of benefit reduction or loss of Medicaid, and/or lack of effective support systems for the transition to adulthood.

The number of recipients employed, and their levels of earnings, varied substantially across impairment groups, which indicates possible differences in human capital development within recipient subgroups.  Employment rates were lowest for the systems disorders group (31 percent) and higher for those with other mental and behavioral disorders and other disabilities (49 percent each).  Despite the overall employment differences, those with systems disorders had a relatively high concentration of youth earning more than $2,000 per year at ages 16 or 17 (17 percent).  By comparison, the other mental and behavioral disorders group only had 10 percent who earned over $2,000 at these ages.  This finding suggests that, while youth with other mental and behavioral disorders have higher employment rates compared to other groups, they tend to be in jobs that pay less, have fewer hours, or have shorter durations.

A substantial minority of child SSI recipients received vocational training (36 percent), which has been associated with positive employment outcomes after age 18 in the special education literature, although few SSI recipients access this training through VR programs.  The use of vocational training services was comparable across subgroups of recipients, although usage of VR services was substantially higher for those with systems disorders.  The similarities in the use of vocational training services could reflect the high rate of participation across all groups in special education services (82 percent).  The higher reports of work plans among the systems disorders group might reflect the greater need for physical accommodations that can be obtained through the VR system or it might be that individuals with less severe functional limitations are more likely to be served through the VR system.

While most child SSI recipients are in school prior to age 18, the high reported school dropout rates and social problems of child SSI recipients pose major concerns (Table 4).  Most child SSI recipients had received special education services at some point during their school years (82 percent).  However, at the time of the survey, 30 percent of all child SSI recipients had dropped out of school.  By comparison, the dropout rate among all youth ages 16 to 24 in the general population is 10 percent (Kaufman et al. (2001). A substantial share of child SSI recipients also reported a problem in school that resulted in suspension or expulsion (43 percent) or being arrested at some point prior to the survey (16 percent).  These findings are consistent with Wagner et al. (2006), who reported that special education students are more likely to experience social difficulties relative to other youth. 

The reported problems were especially high among child SSI recipients with other mental and behavioral disorders.  Of these recipients, 45 percent had dropped out of school at the time of the survey, 52 percent reported being expelled or suspended from school and 28 percent reported a prior arrest.  By comparison, youth with systems disorders had a much lower prevalence of reported problems in each of these categories. 

The findings on human capital investment characteristics suggest that, while some child SSI recipients appear to be preparing for a transition to independent living after age 18 through work, schooling, and training, a substantial minority face several risk factors that could limit their potential for long-term self-sufficiency.  Of particular concern are the high reported dropout and arrest rates, especially within the other mental and behavioral disorders subgroup.

Post Age-18 SSI Program OUTCOMES

In Table 5, we summarize the rates of remaining on and no longer receiving SSI at age 19 by individual and human capital characteristics to assess the extent that these experiences vary.  The findings above illustrate that the pre-age-18 characteristics, barriers, and experiences of child SSI recipients vary substantially across SSI impairment subgroups in ways that might influence their age-18 redetermination, particularly for the other mental and behavioral disorders population.  As in the prior tables, each of the characteristics was measured at the time of the NSCF interview (when the recipient was age 17 or 18), but the columns represent their adult SSI participation status at age 19.  Hence, the row values provide some indication of the correlation between pre-age-18 characteristics and age 19 program outcomes.  Our comparisons in this table show the proportion of each subgroup that remained on or no longer received SSI at age 19 (i.e., the row values sum to 100 percent).  Any subgroup that has a proportion substantially different from the total indicates a potentially different probability of remaining on or no longer receiving SSI for the group with this characteristic. 

Consistent with previous studies on the outcomes of the age-18 redetermination, we find almost two-thirds (64 percent) of the study sample remained on SSI at age 19 (Table 5).[3]  Also consistent with prior studies, we find those with other mental and behavioral disorders prior to age 18 were least likely (47 percent) and those with systems disorders were most likely to be on SSI at age 19 (83 percent), relative to youth in other SSI impairment groups.

We find some large differences in program outcomes at age 19 across several program, health, and human capital characteristics of the youth at the time of the survey, and relatively small differences in program outcomes across demographic and family characteristics.  Child SSI recipients who entered the program between ages 10 and 13 were substantially less likely to be on SSI (53 percent vs. 64 percent for the total), particularly relative to groups that come on prior to age 10 (71 percent vs. 64 percent for the total). In addition, youth with certain health care needs, especially those needing or using medicine prescribed by a doctor, and those needing help with personal care, were more likely to be on SSI at age 19.  The SSI participation rates at age 19 were especially high for youth who reported needing help with routine needs (91 percent), needing/using medicine (88 percent), and needing special therapy (72 percent).  Finally, we find relationships between human capital and social measures—such as employment, school dropout rates, and arrest rates—and receipt of SSI at age 19.  Interestingly, youth with negative social experiences prior to age 18 (e.g., arrests and high school dropouts) were less likely to be on SSI at age 19 compared to the total, as were youth who had more employment experiences prior to age 18.  It is possible that the former group might lack the necessary social connections after age 18 to continue pursuing SSI benefits, or they might have other outcomes, such as institutionalization, which preclude them from receiving benefits.  Conversely, as expected, those with more employment experiences were less likely to report being on SSI than the average child recipient (51 vs. 64 percent of the total).

These descriptive comparisons indicate that the probabilities of remaining on SSI after age 18 vary substantially by disability subgroup.  Additionally, other factors, such as social behaviors (e.g., arrests), also are strongly correlated with programmatic outcomes.  While it is difficult to assess whether characteristics, such as prior work history, directly affect the outcomes of the redetermination, the findings do further illustrate the diversity of needs and outcomes across a child SSI population at the time of redetermination.  Whether these bivariate relationships remain after controlling for other factors is the focus of the next section.

Multivariate Analysis of Factors that Influence Probability of Being On SSI At Age 19

To further assess differences in program outcomes across child SSI subgroups, we present our findings from a multivariate analysis that includes many of the subgroup indicators in Tables 2 through 5.  We take an iterative approach, where we first assess the probability of no longer receiving SSI by impairment subgroup, and then sequentially add in information on the youth’s demographic, health, and, finally, human capital characteristics.  We have chosen this ordering because we want to assess whether the probability of no longer receiving SSI changes for any of these impairment subgroups when more information about the recipient’s characteristics is added.  We start with administrative characteristics because these can be readily identified using SSA data; we add in the respondents’ characteristics at age 17 to examine whether the probabilities for any of these impairment groups change. If the probabilities do not change, we conclude that the other factors play a relatively minimal role in program outcomes after age 18.  We are particularly interested in the roles that education, employment, and other social indicators play in influencing program outcome probabilities. While these characteristics are not considered formally in SSA’s redetermination process, they may play important roles in youths’ decisions to pursue SSI benefits further. 

Column 1 of Table 6 presents finding from a logistic regression of SSI program participation on the impairment groups that were introduced earlier.  It presents the odds-ratios for program participation for youth with mental retardation, other mental and behavioral disorders, and other disabilities, relative to youth with systems disorders.[4] The systems disorders group is the excluded category. While the odds of each group being on SSI are less than those with physical disabilities, those with other mental and behavioral disorders have the lowest odds of remaining on SSI.  The odds of those with other mental and behavioral disorders remaining on the program are 72 percent lower than those with physical disabilities.  (Note that, while the odds of those with “other disabilities” being on SSI are lower than for those classified as mentally retarded, the heterogeneity of this group makes it difficult to interpret why this would be so.)  This confirms previous results from Rogowski et al. (2002) indicating that recipients with mental disorders (including mental retardation) are more likely to be terminated from SSI as a result of the age-18 CDR.

In column 2, we add the age of entry into the program, where the excluded category is under age 10.  The odds of remaining on SSI at age 19 for children who entered SSI between ages 10 and 13 are almost half those of children who entered SSI at younger ages.  Children with entry ages of 14 to 18 are indistinguishable from children with younger entries at the 5 percent level. The findings with respect to age at first program entry could indicate that program duration plays an important role in the transition process.  The younger the recipients when they first receive SSI payments, the more likely they are to have more severe disabilities, especially relative to those who came on between ages 10 and 13. 

There are two potential explanations for why youth who first receive payments between ages 10 and 13 are at greater risk of program exit than those in other age groups.  First, the nature of their disabilities is different from both the youngest and oldest group.  Young children who come onto SSI are more likely to have conditions that can be identified at early ages, such as congenital anomalies (SSA 2006). These conditions tend to have similar definitions for children and adults, and are likely to last a lifetime.  Second, children who enter SSI at young ages are likely to have gone through several CDRs by age 18, which “weed out” those with less severe or diminished impairments; these individuals’ disabilities are less likely to meet the adult requirements.  In fact, previous research has shown that those who have received at least one CDR before their age-18 redetermination are more likely to continue in the program (SSA 2003).  It is unclear why these same conditions do not lead to a significant difference in program continuation between the youngest and oldest age groups.  One explanation may be that children who are first determined to be eligible at older ages may have impairments that are more likely to meet the adult eligibility criteria compared to those who enter during their pre- and early-teen years. 

Our odds ratio estimates for the SSI impairment groups are relatively unchanged from the first column, indicating that age of entry does not have a disproportionate effect on program continuation for any subgroup. This is not surprising, given our assumption that any youth who enter the program at earlier ages—regardless of their impairment—likely have a more readily diagnosable condition that was identified in early childhood. 

Adding personal demographic characteristics from the NSCF to the model does little to change the above findings (column 3).  Even with the addition of a wide range of personal characteristics, the odds-ratios for the impairment groups change relatively little. Unlike previous studies (e.g., Loprest and Wittenburg 2007, SSA 2003), we find males are more likely to remain on SSI after controlling for these other characteristics though we believe this finding might be related to our sampling construction.[5] Whites also were more likely to remain on SSI at age 19. 

In column 4, we add health need indicators from the NSCF to the analysis of SSI participation.  As Loprest and Wittenburg (2007) found, most of these have highly significant effects on SSI participation at age 19.  Individuals needing or using more services than others of the same age are much more likely to be on SSI at age 19 than those who do not.  Those using special equipment, needing help with personal care or handling routine needs, and needing mental health treatment also have much higher odds of SSI participation at age 19.  Many of these indicators suggest a direct connection with the medical system, including service providers, that might be able to help the youth during their age-18 redetermination.  The odds of the mentally retarded group remaining on SSI is no longer significantly different from the odds for those with physical disabilities, suggesting some overlap in what these variables measure. The probabilities for the youth in the “other disabilities” impairment group remain relatively unchanged with the inclusion of these variables, which could suggest that the NSCF health and functioning questions do not fully capture the types of health difficulties facing SSI youth in this group.  There are no differences across race; however, coming from a two-parent home is significantly associated with remaining on SSI.  While the odds of males remaining on SSI increase relative to females when health characteristics are added to the model, the effect of entering SSI between ages 10 and 13 lessens a bit.  These findings suggest that demographic characteristics play a more modest role in SSI decisions after age 18.

In column 5, we add indicators of human capital development and social problems.  We find that the probability of remaining on SSI is substantially lower for those who were employed prior to age 18 and those who had earnings in excess of $2,000.  The odds of child SSI recipients who had worked being on SSI at age 19 are 42 percent lower than for recipients without employment experience.  We also find those who had been arrested are 41 percent less likely than those who had not been arrested to be on SSI at age 19.  Interestingly, vocational training increases the odds of SSI participation at age 19 by 34 percent.  It is possible that a connection to a vocational training program suggests a broader connection to a system of supports that youth can use to maintain their SSI benefits after age 18.  Indeed, youth still in school, with access to the accompanying support system, have 19 percent higher odds of remaining on SSI at age 19, although this is significant only at the 10 percent level.

The probabilities for the SSI impairment subgroups change minimally with the addition of these covariates, suggesting that the effects of human capital development and social problems on program outcomes are relatively similar across the impairment groups.  Additionally, there are only minor changes in the relationship between the health characteristics and supports and remaining on SSI. Coming from a two-parent family is no longer significant, however, and the relationship between age at program entry and remaining on SSI is lessened further.  Males are even more likely to remain on SSI when human capital characteristics are included in the model.

These results confirm the previous analyses regarding the importance of the impairment category on post-age-18 SSI participation (Rogowski et al. 2002).  Even after controlling for measures of disability severity, duration, and human capital, youth with other mental and behavioral disorders are much less likely to receive SSI at age 19.  Combined with previous evidence on the prevalence of school-related and social problems among this group, these youth appear to have a particularly difficult time in adjusting to adulthood (Wagner 2005). 

The findings also suggest that non-health factors, particularly education, employment, and social indicators, play an important role in the probability of a child SSI recipient being on adult SSI after age 18.  The employment findings indicate that child SSI recipients with early employment experiences are less reliant on adult SSI benefits.  In part, these findings might be related to the work-based definition of disability used in the adult definition.  However, they also might provide some indication that early employment experiences, especially for those who had earnings above $2,000, lead to more adult employment, and thereby reduce the odds of being on SSI after age 18.  Somewhat surprisingly, we also find that several negative social indicators reduce the probability of remaining on adult SSI benefits after age 18.  It is possible that these indicators, especially arrests, capture those youth with limited social connections, which reduce their prospects of following up or providing the appropriate information during their age-18 redetermination.  Finally, across all model specifications, we find that the probability of being on SSI at age 19 is substantially lower for recipients in the other mental disorder and other disabilities groups, and for those who entered the program at ages 10 to 13.  The relatively large differences in the probabilities of transition to adult SSI benefits at age 18 indicate that policymakers might want to research options that would prepare these groups for eventual non-participation, or to provide supports other than SSI, given that a substantial share of these recipients do not receive benefits after age 18. 

Post Age-18 Employment and Earnings

We conclude with an analysis of post age-18 employment and earnings to determine how many youth were working and whether they earned amounts sufficient to offset the loss of their child SSI benefits.  These data provide limited evidence of whether some youth are succeeding in the workforce following their age-18 redetermination.

Young adults who no longer receive SSI benefits must find alternative sources of income to maintain their standard of living.  To fully replace their SSI benefits with earnings from employment, they need to earn approximately $5,500 per year on average to offset their average monthly benefit amount (which averaged about $460 across our sample at the time of the survey), assuming that the income of other family members remained the same before and after the adult redetermination process.  They would likely need to earn more, since most children in fact move out on their own at this time and their family’s income is no longer disposable to them.  In addition, this calculation does not factor in the potential loss of Medicaid.

Across all former child SSI recipient groups, 41 percent were employed at age 19 and 14 percent were earning more than the annual SSI payments they received in the calendar year in which they turned age 17 (Table 7).  The relatively low employment rates across the entire population are not surprising, given that approximately 64 percent remain on adult SSI benefits.  Among those no longer receiving SSI at age 19, 61 percent were employed, in comparison to 29 percent who were on SSI.  Additionally, average earnings among those employed were substantially higher for the off-SSI group, in comparison to the on-SSI group ($5,428 vs. $3,236).

A concern is that only 61 percent of those no longer receiving SSI were working and only 25 percent earned enough to replace their childhood SSI benefits.  This finding indicates that most former child SSI recipients are not earning enough money at age 19 to offset their childhood SSI benefits and so must find another source of income to maintain their economic well being. 

In Table 8, we more closely examine the pre-age-18 characteristics of those no longer receiving SSI at age 19 to identify whether there are correlations between these characteristics and age 19 program-earnings outcomes.  We stratify the table into youth who did not have any employment earnings, youth who had earnings less than their age 17 (calendar year) SSI payments, and youth who had earnings greater than these payments.  Presumably, youth in the third category were successfully making a transition into employment to offset their SSI benefits.  Youth in the other categories were not earning enough to offset their SSI benefits, although it is possible they were involved in other activities aimed at developing their human capital (e.g., education, training, or work for pay that is not reported to SSA or the Internal Revenue Service).  While we do not observe human capital activities after age 19, the pre-age-18 human capital activity characteristics provide a reasonable proxy for the types of activities in which these youth might have engaged after age 18.  For example, correlations between poor human capital indicators before age 18, such as dropping out of high school and no subsequent earnings, suggest that some youth might not successfully transition to self-supporting employment.  However, if youth who did not have any earnings were disproportionately likely to be in school prior to age 18, we could infer that they might have continued their schooling or training after age 18. 

Our findings indicate that a substantial portion of those no longer receiving SSI who were not employed or earned less than their child SSI payments at age 19 had relatively limited human capital investments relative to those earning more.  Based on their pre-age-18 activities, a major concern is that many youth who are no longer receiving SSI at age 19 and not working may not be pursuing any human capital activity after age 19. Forty-seven percent of these youth dropped out of school, 26 percent had arrest records, and more than half had been suspended or expelled from school.  Low-earning former recipients had similar dropout and arrest rates and higher rates of suspension or expulsion.  In comparison, the rates of dropout, arrest, and suspension or expulsion were lower among those who were earning above their child SSI payments (23, 19, and 41 percent, respectively).  It is possible that some of those not earning enough to offset their previous SSI benefits were engaged in other behaviors in lieu of developing their human capital, such as criminal activity or parenting, or were incarcerated.[6]

Early work experience also appears to correlate with post-age-18 employment outcomes. High earners at age 19 were most likely to have had high earnings pre-age 18, and those not employed were least likely to have high pre-age-18 earnings.  While 71 percent of low-earning 19 year olds not receiving SSI worked at ages 16 or 17, only 13 percent were earning more than $2,000 dollars at the time.  Comparatively, only 66 percent of high-earning 19 year olds not receiving SSI worked at ages 16 or 17, but 40 percent earned more than $2,000.

These findings suggest that, while some former child SSI recipients who were no longer receiving SSI at age 19 appear to have made a successful transition into adult life, a large majority of them were not earning enough to offset their child SSI benefits and appear to have few prospects for long-term independent living.  This seems evident prior to the age-18 redetermination, as those with high earnings post-redetermination also had high earnings pre-redetermination.

Conclusion

Approximately one-third of child SSI recipients facing the age-18 redetermination will need to search for alternative sources of support after that critical point.  However, the probability of remaining on or no longer receiving SSI after this point varies across child recipients with different primary impairments and human capital characteristics.  Those with mental and behavioral disorders (other than mental retardation) are least likely to be on SSI at age 19, relative to recipients with other impairments.  In addition, those with early employment experience and those with a history of problem behaviors (arrest, suspension or expulsion from school, dropped out of school) are less likely to be on SSI at age 19, relative to others.  Of particular interest are the relatively poor outcomes of child SSI recipients with other mental and behavioral disorders who, relative to others, have both higher rates of problem behaviors prior to age 18 and lower rates of adult SSI participation after age 18.  While we cannot assess the efficacy of the age-18 redetermination in influencing outcomes, our findings do indicate that many former child SSI recipients do not develop their human capital sufficiently to make a successful transition to adulthood.  Few of those no longer receiving SSI at age 19 had enough earnings to offset their pre-redetermination SSI benefit.

The findings suggest that additional research is needed to determine whether effective intervention options can be developed to better prepare and support those who do not continue receiving benefits after the age-18 redetermination.  A substantial share of child beneficiaries will not qualify for adult benefits.  Many of these youth have limited employment prospects and high rates of reported social problems. 

In considering additional supports to smooth the transition to adulthood, policymakers may wish to look at the characteristics and needs of child recipients with other mental and behavioral disorders in particular, and specifically target interventions toward this vulnerable group.  SSA currently is implementing the Mental Health Treatment Study, which provides employment support and mental health treatments for SSA disability beneficiaries with mental disorders.  Policymakers may wish to consider developing a similar initiative for SSI recipients nearing or just completing the age-18 redetermination.

In addition, policymakers may want to review the important role the SSI program plays in influencing the human capital decisions of child SSI recipients.  Specifically, they may want to consider expanding opportunities to promote work, such as outreach to increase knowledge of current SSI work incentives, since early employment experiences have important implications for later outcomes.  Currently, SSA is implementing extensions to Section 301, which will allow SSI recipients to continue receiving benefits after age 18 if they stay in school or vocational rehabilitation activities.  Additional initiatives could be considered for targeting segments of the SSI population that need social and other employment supports (such as those being implemented in the Youth Transition Demonstration) to ensure that they do not fall through the cracks of the system as adults, although these alternative supports may extend beyond the purview of the SSI program.

Finally, additional research on longer-term outcomes would be useful in further describing the transition process from childhood to adulthood for child SSI recipients and informing future intervention efforts.  Linking additional years of SSA administrative data to the NSCF data, for instance, could provide insight on the extent to which (1) those who no longer are receiving SSI after the age-18 redetermination return to the program; (2) those who no longer are receiving SSI but are employed maintain employment and increase their earnings; and (3) those who no longer are receiving SSI and are not employed immediately after the age-18 redetermination develop additional human capital (through schooling, for example) in lieu of employment, or are floundering.  Research also could provide insight into whether concerns about the lower socioeconomic prospects of child recipients with other mental and behavioral disorders are borne out, and could explore potential service strategies for former child recipients who become disconnected from financial, human capital development, and other support systems. 


References

D. Auxter, W. Halloran, H. Berry, and S O’Mara. 1999. “The Precarious Safety Net: Supplemental Security Income and Age 18 Redeterminations,” Focus on Autism and Other Developmental Disabilities, Winter Vol. 14 No (4, pp. 194-203.

R. Cameto, Employment of Youth with Disabilities after High School in Chapter 5 of After High School: A First Look at the Postschool Experiences of Youth with Disabilities, April 2005, available at http://www.nlts2.org/reports/2005_04/nlts2_report_2005_04_ch5.pdf (access May 25, 2007).

Committee on the Health and Safety Implications of Child Labor, National Research Council and Institute of Medicine, Protecting Youth at Work: Health, Safety, and Development of Working Children and Adolescents in the United States, Washington, DC: National Academy Press, 1998.

J. Gillcrist and D. Edson, National Survey of SSI Children and Families User’s Manual for Restricted and Public Use Files, Washington, DC: Mathematica Policy Research, Inc., 2004.

H. Ireys, D. Kasprzyk, A. Takyi and J. Gillcrist, Estimating the Size and Characteristics of the SSI Child Population: A Comparison Between the NSCF and Three National Surveys, Mathematica Policy Research Report No. 8761-980. Washington, D.C.: Mathematica Policy Research, Inc., 2004.

P. Kaufman, M. N. Alt and C. Chapman, Drop Out Rates in the United States: 2000, Report to the National Center for Education Statistics, Statistical Analysis Report, 2001. Available at http://nces.ed.gov/pubs2002/droppub_2001/ (accessed April 25, 2007).

P. Loprest and D. Wittenburg, Post Transition Experiences of Former Child SSI Recipients, Social Service Review (volume forthcoming) (2007).

R. Luecking and E. Fabian, Paid Internships and Employment Success for Youth in Transition, Career Development for Exceptional Children 23(2) (2000), pp. 205-221.

J. Rogowski, L. Karoly, J. Klerman, M. Inkelas, M. Rowe and R. Hirscher, Final Report for Policy Evaluation of the Effect of the 1996 Welfare Reform Legislation on SSI Benefits for Disabled Children, DRU-2559-SSA, Santa Monica, CA: The Rand Corporation, March 2002.

K. Rupp, P. Davies, C. Newcomb, H. Iams, C. Becker, S. Mulpuru, S. Ressler, K. Romig and B. Miller. Profile of Children with Disabilities Receiving SSI: Highlights from the National Survey of SSI Children and Families. Social Security Bulletin 66(2), 2005/2006, pp. 21-48.

Social Security Administration, Age-18 Redetermination Profiling Model Development Report, 2003.

Social Security Administration, Children Receiving SSI, 2005, Baltimore, MD: Author, 2006.

Social Security Advisory Board, Statement on the Supplemental Security Income Program, at http://www.socialsecurity.gov/OACT/SSIR/SSI06/board.html, 2006.

SSA Office of Policy, at http://www.ssa.gov/policy, 2007.

R. Stancliffe and C. Lakin, Costs and Outcomes of Community Services for People with Intellectual Disabilities, Baltimore, MD:  Paul H. Brookes Publishing Company, 2005.

U.S. Code 42, Sec. 1382c, available at http://www.law.cornell.edu/uscode/uscode42/ usc_sec_42_00001382---c000-.html (accessed August 7, 2007).

M. Wagner, The Leisure Activities, Social Involvement, and Citizenship of Youth with Disabilities After High School, in After High School: A First Look at the Postschool Experiences of Youth with Disabilities, 2005, Chapter 7, available at http://www.nlts2.org/reports/2005_04/nlts2_report_2005_04_ch7.pdf

M. Wagner, L. Newman, R. Cameto, P. Levine and N. Garza, An Overview of Findings From Wave 2 of the National Longitudinal Transition Study-2, Report to the U.S. Department of Education, Institute of Education Sciences, August 2006. Available at http://ies.ed.gov/ncser/pdf/20063004.pdf.

D. Wittenburg and P. Loprest, Early Transition Experiences of Transition Age Child SSI Recipients: New Evidence from the National Survey of Children and Families, Journal of Disability Policy Studies, forthcoming. 

D. Wittenburg and P. Loprest. “Policy Options for Assisting Child SSI Recipients in Transition” Report Submitted to SSA Ticket Afadvisory Panel, 2003, available on-line at http://www.ssa.gov/work/panel/panel_documents/SSI%20Kids-Final.pdf (accessed October 30, 2007).

 


TABLE 1

 

DEMOGRAPHIC AND FAMILY CHARACTERISTICS OF AGE 17 CHILD

SSI RECIPIENTS FROM THE 1999-2003 COHORTS

 

 

Percent

SE

Gender

 

 

Male

62.9

1.0

Age at Time of NSCF Interview

 

 

17

43.9

1.1

Race

 

 

White

46.7

1.9

Education

 

 

In school (6th – 12th grade, special education)

58.6

1.4

Graduated high school

11.0

0.6

Dropped out/out of school/other

30.4

1.4

Living Arrangements

 

 

Two-parent family

23.4

1.2

Single-parent family

55.1

1.3

Other

21.5

0.9

SSA Primary Impairment at Age 17

 

 

Mental retardation (MR)

49.6

1.1

Mental and behavioral disorders (other than MR)

25.8

0.9

Systems disorders

14.8

0.9

Other disabilities

9.8

0.8

Age at Program Entry

 

 

Under 10

57.1

1.2

10-13

27.0

1.0

14-17

15.9

1.1

Sample Size

730

Weighted sample size

97,106

 

Source:    Authors’ tabulations using linked 2001-2002 NSCF and SSA data files. 

 

Notes:     The sum of individual categories may not equal the total because of individual rounding.  Tabulations are based on five cohorts of SSI recipients (1999 through 2003) who were receiving payments in January of the year in which they turned 17.  Sample members were ages 17 or 18 at the time of their NSCF survey interview.  All other characteristics are based on the recipients’ responses to the NSCF interview.  See Appendix Table 1 for a full description of all variables.  Standard errors for all estimates are based on balanced repeated replicate weights included with the NSCF. 

 


TABLE 2

 

DEMOGRAPHIC AND FAMILY CHARACTERISTICS OF AGE 17 CHILD SSI RECIPIENTS

FROM THE 1999-2003 COHORTS, BY SSA IMPAIRMENT STATUS

 

 

Total

Mental Retardation

Other Mental Disorders

Systems Disorders

Other Disabilities

 

%

SE

%

SE

%

SE

%

SE

%

SE

Total

 

 

 

 

 

 

 

 

 

 

Percent of Population (weighted)

100

-

49.6

1.10

25.8

.89

14.9

.92

9.8

.78

Gender

 

 

 

 

 

 

 

 

 

 

Male

62.9

0.95

62.9

1.60

68.1

2.16

50.3

3.71

68.2

3.45

Race

 

 

 

 

 

 

 

 

 

 

White

46.7

1.91

40.9

2.46

50.8

2.91

55.7

4.30

51.3

4.22

Living Arrangements

 

 

 

 

 

 

 

 

 

 

Two-parent family

23.4

1.23

25.3

1.71

17.7

1.42

23.5

2.94

28.4

3.51

Single-parent family

55.1

1.27

52.5

1.77

55.1

2.36

62.6

3.32

57.1

3.42

Other

21.5

0.88

22.2

1.08

27.3

2.22

13.9

1.61

14.5

3.02

Sample Size

730

372

205

90

63

Weighted sample size

97,106

48,133

25,072

14,419

9,483

 

Source:    Authors’ tabulations using linked 2001-2002 NSCF and SSA data files. 

 

Notes:     The sum of individual categories may not equal the total because of individual rounding.  Tabulations are based on five cohorts of SSI recipients (1999 through 2003) who were receiving payments in January of the year in which they turned 17.  Sample members were ages 17 or 18 at the time of their NSCF survey interview.  Impairment and age at program entry are based on SSA administrative records. All other characteristics are based on the recipients’ responses to the NSCF interview.  See Appendix Table 1 for a full description of all variables.  Standard errors for all estimates are based on balanced repeated replicate weights included with the NSCF. 

 

 

 

 


TABLE 3

 

HEALTH CHARACTERISTICS AND PROGRAM ENTRY DATE OF AGE 17 CHILD SSI RECIPIENTS

FROM THE 1999-2003 COHORTS, BY SSA IMPAIRMENT STATUS

 

 

Total

Mental Retardation

Other Mental Disorders

Systems Disorders

Other Disabilities

 

%

SE

%

SE

%

SE

%

SE

%

SE

Special Health Care Needs (FACCT Screener)

 

 

 

 

 

 

 

 

 

 

Any need

86.8

1.1

83.7

1.3

88.4

1.3

87.8

4.5

97.1

0.8

Needs/uses more services    than others same age

67.0

1.4

61.8

2.1

65.2

2.7

74.7

4.3

87.0

2.6

Needs/uses medicine prescribed by a doctor

51.4

1.3

39.5

1.7

63.3

2.1

60.7

4.0

66.6

4.4

Limited in ability to do things others of same age do

62.2

1.5

56.5

1.9

55.2

2.4

77.2

4.7

86.8

2.2

Needs/gets special therapy

33.6

1.3

26.9

1.4

30.5

2.4

49.8

4.2

51.6

4.4

Needs/gets mental health treatment/counseling

53.1

1.1

46.1

1.4

76.3

1.8

39.9

3.5

47.8

4.0

Other Health Supports

 

 

 

 

 

 

 

 

 

 

Uses special medical equipment

10.9

0.7

3.7

0.5

2.0

0.4

38.6

3.5

28.4

3.3

Needs help with personal care

22.6

1.0

18.4

1.1

16.4

1.7

47.0

4.2

23.5

3.2

Needs help with handling routine needs

72.3

1.2

74.2

1.4

71.1

1.8

71.1

4.5

67.4

4.4

Age at SSI program entry

 

 

 

 

 

 

 

 

 

 

Under 10

57.1

1.2

61.4

1.6

39.1

2.3

72.1

4.4

59.9

4.2

10-13

27.0

1.0

28.5

1.6

34.1

2.2

8.9

1.5

28.2

4.2

14-18

15.9

1.1

10.1

1.1

26.8

2.4

18.9

4.3

11.9

2.1

Sample size

730

372

205

90

63

Weighted sample size

97,106

48,133

25,072

14,419

9,483

 

Source:   Authors’ tabulations using linked 2001-2002 NSCF and SSA data files. 

 

Notes:     The sum of individual categories may not equal the total because of individual rounding.  Tabulations are based on five cohorts of SSI recipients (1999 through 2003) who were receiving payments in January of the year in which they turned 17.  Sample members were ages 17 or 18 at the time of their NSCF survey interview.   Impairment and age at program entry are based on SSA administrative records. All other characteristics are based on the recipients’ responses to the NSCF interview.  See Appendix Table 1 for a full description of all variables.  Standard errors for all estimates are based on balanced repeated replicate weights included with the NSCF. 

 

 


TABLE 4

 

HUMAN CAPITAL CHARACTERISTICS OF AGE 17 CHILD SSI RECIPIENTS

FROM THE 1999-2003 COHORTS, BY SSA IMPAIRMENT STATUS

 

 

Total

Mental Retardation

Other Mental Disorders

Systems Disorders

Other Disabilities

 

%

SE

%

SE

%

SE

%

SE

%

SE

Employment

 

 

 

 

 

 

 

 

 

 

Percent ever employed age 16-17

40.7

1.3

37.7

1.7

48.9

2.5

31.2

4.3

48.8

4.9

Percent ever earned over $2,000 at age 16 or 17

11.2

0.8

8.6

0.8

10.0

1.2

17.1

4.5

18.0

4.2

Training

 

 

 

 

 

 

 

 

 

 

Ever received vocational training

35.8

1.1

36.9

1.4

36.2

2.3

34.8

4.3

30.6

3.6

Ever had a work plan

13.2

1.1

9.8

1.1

14.3

2.5

22.7

4.1

13.2

2.2

Education

 

 

 

 

 

 

 

 

 

 

In school (6th- 12th grade, special education)

58.6

1.4

63.8

1.3

46.9

2.7

73.6

3.0

40.0

3.5

Graduated high school

11.0

0.6

9.6

0.6

8.6

0.9

9.2

1.2

26.9

4.2

Dropped out/out of school/other

30.4

1.4

26.6

1.4

44.5

2.6

17.1

2.6

33.1

4.8

Special education

81.5

0.8

82.1

1.2

85.7

1.5

81.2

2.5

67.6

4.2

Social Problems

 

 

 

 

 

 

 

 

 

 

Ever suspended or expelled from school

42.7

1.4

44.7

2.0

51.7

2.6

26.8

4.6

32.6

4.9

Ever arrested

16.4

0.7

13.3

1.0

28.4

1.9

5.5

1.5

16.5

2.8

Sample size

730

372

205

90

63

Weighted sample size

97,106

48,133

25,072

14,419

9,483

 

Source:    Authors’ tabulations using linked 2001-2002 NSCF and SSA data files. 

 

Notes:     The sum of individual categories may not equal the total because of individual rounding.  Tabulations are based on five cohorts of SSI recipients (1999 through 2003) who were receiving payments in January of the year in which they turned 17.  Sample members were ages 17 or 18 at the time of their NSCF survey interview.  Impairment and age at program entry are based on SSA administrative records. All other characteristics are based on the recipients’ responses to the NSCF interview.  See Appendix Table 1 for a full description of all variables.  Standard errors for all estimates are based on balanced repeated replicate weights included with the NSCF.  Ever having a work plan is identified in the NSCF by having an individualized work rehabilitation plan, now known as an individualized plan for employment.

 

 

 


 TABLE 5

 

FORMER CHILD SSI RECIPIENTS ON AND OFF OF SSI AT AGE 19, BY SELECTED IMPAIRMENT, DEMOGRAPHIC, FAMILY, EMPLOYMENT, AND HUMAN CAPITAL CHARACTERISTICS

 

 

On SSI at Age 19

Off SSI at Age 19

 

%

SE

%

SE

Total

64.1

1.1

35.9

1.1

SSA Primary Impairment at Age 17

 

 

 

 

Mental retardation (MR)

67.5

1.5

32.5

1.5

Mental and behavioral disorders (other than MR)

46.8

2.3

53.2

2.3

Systems disorders

82.8

2.4

17.2

2.4

Other disabilities

64.9

3.2

35.1

3.2

Age at SSI Program Entry

 

 

 

 

Under 10

71.0

1.4

29.0

1.4

10-13

52.5

2.0

47.5

2.0

14-18

59.5

4.0

40.5

4.0

Demographic and Family Characteristics

 

 

 

 

Male

65.5

1.4

34.5

1.4

White

67.3

1.4

32.7

1.4

Two-parent family

69.7

2.0

30.3

2.0

Health Characteristics

 

 

 

 

Special Health Care Needs

 

 

 

 

Any Need

66.7

1.1

33.3

1.1

Needs/uses more services than others same age

70.2

1.2

29.8

1.2

Needs/uses medicine prescribed by a doctor

87.9

1.4

12.1

1.4

Limited in ability to do things others of same age do

66.7

1.6

33.3

1.6

Needs/gets special therapy

71.8

1.4

28.2

1.4

Needs/gets mental health treatment/counseling

70.5

1.3

29.5

1.3

Other Health Supports

 

 

 

 

Uses special medical equipment

78.7

1.8

21.3

1.8

Needs help with personal care

63.5

1.6

36.5

1.6

Needs help with handling routine needs

90.8

1.6

9.2

1.6

Employment, Education, and Social Characteristics

 

 

 

 

In school/graduated

69.5

1.3

30.5

1.3

In special education

65.29

1.4

34.7

1.35

In vocational training

66.5

1.8

33.5

1.8

Ever had a work plan

63.6

4.3

36.4

4.3

Percent ever employed ages 16-17

50.9

2.0

49.1

1.998

Ever earned over $2,000 at age 16 or 17

42.7

5.3

57.3

5.3

Ever suspended or expelled from school

57.2

1.9

42.9

1.889

Ever arrested

49.2

2.7

50.8

2.686

Sample size

462

 

263

 

Weighted sample size

61,891

 

34,628

 

 

Source:    Authors’ tabulations using linked 2001-2002 NSCF and SSA data files. 

 

Notes:     The sum of individual categories may not equal the total because of individual rounding.  Tabulations are based on five cohorts of SSI recipients (1999 through 2003) who were receiving payments in January of the year in which they turned 17.  Sample members were ages 17 or 18 at the time of their NSCF survey interview.   Impairment and age at program entry are based on SSA administrative records. All other characteristics are based on the recipients’ responses to the NSCF interview.  The sample size for this table is smaller in comparison to Tables 1 through 4 because five sample members died before age 19. SSI status at age 19 is based on positive payments in January of the year they turned 19.  Standard errors for all estimates are based on balanced repeated replicate weights included with the NSCF.  Ever having a work plan is identified in the NSCF by having an individualized work rehabilitation plan, now known as an individualized plan for employment.


TABLE 6

 

ODDS RATIOS FROM MULTIVARIATE MODELS PREDICTING THE SSI PARTICIPATION OF AGE 17 CHILD SSI RECIPIENTS BEING ON SSI AT AGE 19

(1999-2003 COHORTS)

 

 

(1)

(2)

(3)

(4)

(5)

Administrative Variables

 

 

 

 

 

 

SSA Primary Impairment at Age 17

 

 

 

 

 

Mental retardation (MR)

0.43***

0.47***

0.47***

0.70

0.58*

 

(0.08)

(0.09)

(0.10)

(0.20)

(0.18)

Mental and behavioral disorders (other than MR)

0.18***

0.21***

0.20***

0.29***

0.26***

 

(0.04)

(0.05)

(0.05)

(0.09)

(0.09)

Other disabilities

0.38***

0.42***

0.39***

0.37***

0.35***

 

(0.08)

(0.09)

(0.09)

(0.11)

(0.12)

Age at Program Entry

 

 

 

 

 

Age 10-13

 

0.54***

0.54***

0.67***

0.72**

 

 

(0.06)

(0.06)

(0.09)

(0.10)

Age 14-18

 

0.73*

0.75*

1.05

1.11

 

 

(0.12)

(0.12)

(0.18)

(0.21)

 
Survey Variables

 

 

 

 

 

 
Demographic and Family Characteristics

 

 

 

 

Male

 

 

1.34**

1.47***

1.66***

 

 

 

(0.16)

(0.20)

(0.23)

White

 

 

1.33***

1.04

0.99

 

 

 

(0.12)

(0.10)

(0.11)

Age 18 at the time of the NSCF Interview

 

1.07

1.10

1.09

 

 

 

(0.08)

(0.09)

(0.11)

Two-parent family

 

 

1.25*

1.34**

1.25

 

 

 

(0.15)

(0.17)

(0.18)

Health Characteristics and Supports

 

 

 

 

 

Needs/uses more services than others same age

 

 

1.61***

1.72***

 

 

 

 

(0.21)

(0.22)

Limited in ability to do the things others of same age do

 

 

0.95

0.90

 

 

 

 

(0.11)

(0.10)

Needs/gets special therapy

 

 

 

1.31*

1.30*

 

 

 

 

(0.19)

(0.19)

Uses special medical equipment

 

 

 

1.59***

1.48**

 

 

 

 

(0.23)

(0.23)

Needs help with personal care

 

 

 

0.83

0.92

 

 

 

 

(0.11)

(0.11)

Needs help with handling routine needs

 

 

 

2.24***

2.02***

 

 

 

 

(0.55)

(0.54)

Needs/uses medicine prescribed by a doctor

 

 

2.96***

2.45***

 

 

 

 

(0.58)

(0.48)

Needs/gets mental health treatment/counseling

 

 

1.71***

1.65***

 

 

 

 

(0.19)

(0.20)

 

Employment, Education, and School Characteristics

 

 

 

In school

 

 

 

 

1.19*

 

 

 

 

 

(0.12)

In special education

 

 

 

 

0.85

 

 

 

 

 

(0.11)

Ever had vocational training

 

 

 

 

1.34***

 

 

 

 

 

(0.14)

Ever had a work plan

 

 

 

 

0.86

 

 

 

 

 

(0.15)

Ever earned over $2,000 at age 16 or 17

 

 

 

 

0.64**

 

 

 

 

 

(0.14)

Ever employed at age 16 or 17

 

 

 

 

0.58***

 

 

 

 

 

(0.06)

Ever arrested

 

 

 

 

0.59***

 

 

 

 

 

(0.07)

Ever suspended or expelled from school

 

 

 

 

0.88

 

 

 

 

 

(0.10)

Sample size

96,519

Weighted sample size

725

 

Source:   Authors’ tabulations using linked 2001-2002 NSCF and SSA data files. 

 

Notes:    The sum of individual categories may not equal the total because of individual rounding.  Tabulations are based on five cohorts of SSI recipients (1999 through 2003) who were receiving payments in January of the year in which they turned 17.  Sample members were ages 17 or 18 at the time of their NSCF survey interview.   Impairment and age at program entry are based on SSA administrative records. All other characteristics are based on the recipients’ responses to the NSCF interview.  The sample size for this table is smaller in comparison to tables 1 through 4 because five sample members died before age 19. SSI status at age 19 is based on positive payments in January of the year they turned 19.  Standard errors for all estimates are based on balanced repeated replicate weights included with the NSCF.   Ever having a work plan is identified in the NSCF by having an individualized work rehabilitation plan, now known as an individualized plan for employment.

 

*Statistically significant at 10% level

**Statistically significant at 5% level

***Statistically significant at 1% level


TABLE 7

 

EMPLOYMENT EXPERIENCES AT AGE 19 OF FORMER CHILD

SSI RECIPIENTS FROM THE 1999-2003 COHORTS

 

 

Total

On SSI at Age 19

Off SSI at Age 19

 

%

SE

%

SE

%

SE

Percent employed at age 19

40.8

1.3

29.3

1.7

61.2

1.7

Percent employed at age 19 and earning at least the equivalent of their actual SSI payments in the year that they were 17 as of January

13.6

1.2

7.3

1.4

25.0

2.2

Average earnings among those employed

$4,421

$232

$3,236

$358

$5,428

$312

Sample size

713

454

259

Weighted sample size

93,963

60,140

33,823

      

 

Source:   Authors’ tabulations using linked 2001-2002 NSCF and SSA data files. 

 

Notes:    The sum of individual categories may not equal the total because of individual rounding.  Tabulations are based on five cohorts of SSI recipients (1999 through 2003) who were receiving payments in January of the year in which they turned 17.  Sample members were ages 17 or 18 at the time of their NSCF survey interview.   Impairment and age at program entry are based on SSA administrative records. All other characteristics are based on the recipients’ responses to the NSCF interview.  The sample size for this table is smaller in comparison to earlier tables because five sample members died before age 19 and 12 sample members had missing administrative data on earnings. SSI status at age 19 is based on positive payments in January of the year they turned 19.  Standard errors for all estimates are based on balanced repeated replicate weights included with the NSCF.  

 


TABLE 8

 

EMPLOYMENT AND EARNINGS AT AGE 19 OF FORMER CHILD SSI RECIPIENTS WHO ARE OFF OF SSI AT AGE 19 (1999-2003 COHORTS)

 

 

Off SSI at Age 19

 

Not Employed

Employed and Earning Less than Age 17 SSI Benefits

Employed and Earning at Least the Equivalent of Age 17 SSI Benefits

 

%

SE

%

SE

%

SE

 

Employment

 

 

 

 

 

 

 

Ever employed at age 16 or 17

35.1

2.7

71.2

2.8

66.4

5.0

 

Ever earned over $2,000 at age 16 or 17

6.7

1.2

12.6

1.5

40.1

4.8

 

Training

 

 

 

 

 

 

 

Ever received vocational training

25.8

2.7

33.7

2.5

47.0

5.1

 

Ever had a work plan

14.5

1.9

7.7

1.1

16.1

6.1

 

Education

 

 

 

 

 

 

 

Dropped out/out of school/other

46.9

2.5

39.8

2.6

22.9

3.6

 

Social Problems

 

 

 

 

 

 

 

Ever suspended or expelled from school

50.3

3.5

60.6

2.8

40.5

4.3

 

Ever arrested

26.1

2.6

24.6

2.7

18.8

3.0

 

Total (Row Percentages)

38.8

1.6

36.3

1.9

25.0

2.2

 

Sample size

100

99

60

 

Weighted sample size

13,107

12,277

8,439

 

 

Source:   Authors tabulations’ using linked 2001-2002 NSCF and SSA data files. 

 

Notes:    The sum of individual categories may not equal the total because of individual rounding.  Tabulations are based on five cohorts of SSI recipients (1999 through 2003) who were receiving payments in January of the year in which they turned 17.  Sample members were age 17 or 18 at the time of their NSCF survey interview.   Impairment and age at program entry are based on SSA administrative records. All other characteristics are based on the recipients’ responses to the NSCF interview.  SSI status at age 19 is based on positive payments in January of the year they turned 19.  Standard errors for all estimates are based on balanced repeated replicate weights included with the NSCF.   Ever having a work plan is identified in the NSCF by having an individualized work rehabilitation plan, now known as an individualized plan for employment.

 

 

 


Table A.1

 

Summary of Linked nscf-SSA Variables included in descriptive and multivariate analyses

 

NSCF Self Reported Variables

Description

Gender

Male/female.

Race

Categories include white, black and other.

Education

Categories include in school (6th – 12th grade, special education), graduated high school, and dropped out/out of school/other.

Living Arrangements

Categories include two-parent family, single-parent family, and other.

Has Special Health Care Need

 

Categories include needs/uses more services than others same age, needs/uses medicine prescribed by a doctor, limited in ability to do the things others of same age do, needs/gets special therapy, and needs/gets mental health treatment/counseling.

Other Health Needs

Categories include uses special medical equipment, needs help with personal care, and needs help with handling routine needs.

Training

Categories include ever received vocational training and ever had a work plan.

Social Problems

Categories include ever suspended or expelled from school and ever arrested.

SSA Administrative Variables

 

SSA Primary Impairment at Age 17

Categories include mental retardation (MR), mental and behavioral disorders (other than MR), systems disorders, and other disorders.  SSA impairment is based on SSA program records in 2000, when recipients were selected for inclusion in the survey.  Impairments for sample members whose SSI status was uncertain at the time of selection were updated based on more recent administrative data.  The other mental disorders category includes affective, anxiety, schizophrenia, and other mental disorders.  The systems disorders category includes diseases of the nervous system and sensory organs, and systemic disorders.   The other disabilities category includes recipients with infectious and parasitic diseases; neoplasms; endocrine, nutritional, and metabolic diseases; diseases of the blood and blood-forming organs; pregnancy problems; diseases of the skin and subcutaneous tissue; congenital anomalies; perinatal diseases; injuries; and non-codable diseases.

Age at Program Entry

Categories include under age 10, age 10-13, and age 14-17.

Employment

Percent with any earnings, from the SSA summary earnings files (SER).

Earnings above $2,000

Percent with any earnings above $2000 within a calendar year, from the SSA SER.

On/Off SSI at 19

SSI status at age 19 is based on positive SSI payments in January of the year they turn age 19.

TABLE A.2

 

outcome of age-18 redetermination AMONG AGE 17 CHILD SSI RECIPIENTS (1999–2003 COHORTS), AS RECORDED in SSA administrative data

 

 

Number

Percentage

Standard Error

Continuances and Reasons

 

 

 

Meets listing

269

37.3

1.24

Functionally equals listing

12

1.4

.24

Residual functioning/vocational factors

60

7.6

.65

Unknown

16

2.1

.37

Total

357

48.5

1.32

 

 

 

 

Cessations and Reasons

 

Non-disability

219

30.7

1.09

Able to engage in SGA

6

.8

.20

Failure to cooperate

23

2.6

.32

Total

248

34.1

1.04

 

 

 

 

Missing Data

125

17.4

1.21

 

 

 

 

Sample size

730

100.0

-

 

Source:   Authors’ tabulations using linked 2001-2002 NSCF and SSA data files. 

 

Notes:    The sum of individual categories may not equal the total because of individual rounding.  Tabulations are based on five cohorts of SSI recipients (1999 through 2003) who were receiving payments in January of the year in which they turned 17.  Sample members were ages 17 or 18 at the time of their NSCF survey interview.  Age-18 redetermination results are based on SSA’s 833 file, which records all continuing disability review outcomes. Standard errors for all estimates are based on balanced repeated replicate weights included with the NSCF. 

 

 

 



[1] Recipients facing the age-18 redetermination may leave the program of their own volition.  Rogowski et al. (2002) estimate that only 7 percent of individuals truly unaffected by the age-18 redetermination (because of their specific disability type) would have left SSI on their own.  This was estimated as part of an analysis of the effect of PRWORA, and it is not clear what the true normal exit rate is for cohorts unaffected by the change in program rules.  As a baseline comparison for pre-age-18 redetermination exit rates, however, we find from NSCF data that only 9 percent of individuals on SSI at age 15 are off of SSI at age 17, and 15 percent of individuals age 14 in the NSCF are off of SSI at age 16.

[2] To be considered “timely,” age-18 redeterminations should be started within one year of an individual’s eighteenth birthday, so it is possible that some redeterminations may not be completed until the youth reach age 19.  We analyzed SSI status at ages 20 and 21 and obtained similar results to the analysis at age 19.  However, after age 19, the sample size drops because individuals die or not enough time has elapsed for our sample to reach these ages.  As a result, we present results of the analysis at age 19 only.  Additionally, a large fraction of the sample does not have earnings to analyze after age 19 due to a lag in the reporting of earnings. 

[3] Those no longer receiving SSI at age 19 consist of those who did not meet the adult disability criteria at redetermination or the income or resource eligibility criteria, did not complete the redetermination process (including those who failed to cooperate), or left the program on their own.  We cannot reliably distinguish among these reasons in the data.  However, among a group of child SSI recipients who were redetermined at age 18, Rogowski et al. (2002) found that 70 percent of those who did not continue to receive SSI benefits after age 18 failed to meet the adult disability criteria.  Appendix Table A.2 presents data on the actual outcome of the age-18 redetermination from SSA’s 832 administrative data file. The sample for our analysis of post-age-18 redetermination outcomes is slightly smaller than for our analysis of pre-age-18 characteristics because some sample members either did not reach age 19 within the follow-up period or died before reaching age 19..

[4] These are obtained by exponentiating the estimated coefficients (βj). If X1 increases by 1, the odds increase by the odds ratio—exp(β1).  Unlike ordinary least squares regressions, the break between a positive and negative effect is not at 0, but 1.  If the odds ratio is greater than one, then the odds of the event are greater when that variable is increased by one unit. To compute how much more likely an event is, subtract 1 from the odds ratio. For example, if the coefficient on the Female variable is 0.5596, the odds ratio is 1.75 and the odds of the SSI participation are 75 percent greater when Female equals 1 than when it equals 0 (Male).  Similarly, if the odds ratio is less than one, the event is less likely to occur when the variable increases by one unit. If the odds ratio were 0.75, then the odds of the event are lower for females. In this example, the odds of SSI participation are 25 percent lower for females.

[5] When we rerun our estimates using other cohorts of 17 year olds (i.e., those who were ages 14-16 at the time of the survey but who were on SSI at age 17 according to SSA administrative data), we find males were less likely to remain on SSI (the estimates are available upon request).  The general directions of the remaining coefficients are the same as those reported in Table 6, though the magnitude of the coefficients vary.  We do not have an explanation for the difference in our findings reported above for gender and those in the alternative model and in the Loprest and Wittenburg and SSA studies other than that sample selection and our definition of outcomes (“on SSI” at age 19) could be driving these differences.  For example, the Loprest and Wittenburg (2007) sample included former child SSI recipients age 19 and 23 in 2000 who received child SSI benefits in 1996. These young people were between the ages of 14 and 18 in 1996.  It is possible that the outcomes of the larger age cohort differs from the smaller cohort of 17 year olds in our sample. Additionally, Loprest and Wittenburg did not include impairment status in their equations because the linked data were not available for their analysis.   The SSA (2003) findings are based on a sample of beneficiaries at age 17, though they use a different outcome (age-18 redetermination decision) than our study (on SSI at age 19).  One possible explanation is that our estimates of being on-SSI are influenced by other administrative factors, such as the continuation of payments that former child SSI recipients can receive under Section 301 provisions (e.g., participating in vocational rehabilitation) or while appealing their initial cessation under the Goldberg-Kelly provisions (due process). 

[6] Some of these individuals may have only been working part of the year because they were awaiting their redetermination.  Thus we would expect to see those with early birthdays reporting higher earnings than those with later birthdays.  However, only 3 percent of high earners not on SSI were born in the first 3 months of the year, compared with 9 percent of those not on SSI with no earnings and 8 percent of those not on SSI with low earnings.  Those appealing their decision may also not be working, which would bias the results downward.  They would be more likely to report lower earnings.  However, this hypothesis is un-testable with the current data.