National Beneficiary Survey

ANNUAL NATIONAL BENEFICIARY SURVEY TO OCCUR IN 2004-2007

The National Beneficiary Survey (NBS) is being sponsored by the Social Security Administration (SSA) to collect information from SSDI and SSI beneficiaries about their health and well-being, service use, and employment. SSA will use this information to evaluate how well its programs are meeting the needs of disability beneficiaries. This survey is being conducted as part of SSA's evaluation of the Ticket to Work (TTW) program. SSA has contracted with Mathematica Policy Research (MPR), an independent research company, to collect data from SSA disability beneficiaries and Ticket to Work participants. Independent nationally representative samples of beneficiaries will be selected and interviewed in 2004, 2005, 2006, and 2007. In addition, cohorts of Ticket participants will be followed longitudinally for several years. In all, SSA will conduct four rounds of interviews. The survey has very strict privacy protections that have been approved by SSA and the Office of Management and Budget (OMB NO.: 0960-0666).
The survey has five key objectives:

• To provide critical data on the work-related activities of SSI and SSDI beneficiaries, particularly as they relate to the implementation of Ticket to Work
  
• To describe the characteristics and program experiences of beneficiaries who use their Tickets
  
• To gather information about beneficiaries who do not use their Tickets and the reasons they do not
  
• To evaluate the employment outcomes of Ticket users and other SSI and SSDI beneficiaries
  
• To collect data on service utilization, barriers to work, and perceptions about Ticket to Work and other SSA programs designed to help disability beneficiaries find and keep jobs

The survey data will be combined with SSA administrative data to provide critical information on access to jobs and employment outcomes for disability beneficiaries, including those who participate in the TTW program and those who do not. Survey data will be used by SSA for policy making and program planning, and by external researchers interested in disability and employment issues. All information collected will be strictly confidential and will not be reported in any way that identifies survey respondents. Further, the information collected will not be used by the Social Security Administration to determine the continuing eligibility of respondents for disability benefits.

SAMPLE DESIGN

The target population for the NBS includes all SSI and DI beneficiaries between the ages of 18 and 64 who are receiving benefits at the time the sample is drawn. The target population will include beneficiaries in all 50 states and the District of Columbia.

The NBS will use a three-stage sample design. In the first stage, SSA administrative records will be used as a list frame from which 80 primary sampling units (PSUs) will be selected. PSUs are defined as a single county or multiple adjacent counties. In the second stage (used in the two largest PSUs), smaller areas within the PSUs, called secondary sampling units (SSUs) will be formed based on ZIP codes and selected for the study. In the final stage, beneficiaries will be selected from within either PSUs or SSUs using stratified sampling. The initial sample sizes for the NBS will be 10,250 in 2004, 9,653 in 2005, 7,715 in 2006, and 6,320 in 2007.

QUESTIONNAIRE

The questionnaire collects data on a rich array of topics about employment, disability, and experience with the Ticket to Work program. Topics include disability and current work status, current and former employment, awareness of Ticket to Work and other SSA work incentive programs, experience with the Ticket to Work program (for participants), reasons for non-participation (for nonparticipants), employment services used in the past year, health and functional status, health insurance, income and other assistance, and sociodemographics.

The NBS questionnaire will take approximately 45-60 minutes to administer. In cases where the respondent is unable to answer the survey due to disability, the respondent will be offered an in-person interview to accommodate their disability. To promote response among Hispanic populations, the questionnaire will be available in Spanish.

DATA COLLECTION

Extensive participation in this study is very important since it will help SSA better understand the opinions, needs, and experiences of Social Security beneficiaries with a broad range of disabilities. For this reason SSA and MPR are working hard to be sure that all beneficiaries selected for this survey have an opportunity to be heard. The NBS will be a dual-mode survey. Initial interview attempts will be made by phone using computer-assisted telephone interview (CATI) software. Beneficiaries who request an in-person interview, who cannot located by phone, and those who are unable to complete a telephone interview due to their disability, will be contacted in person by an MPR interviewer for a computer-assisted personal interview. The target response rate for the NBS is 80 percent.

MPR will send an advance letter to sampled beneficiaries prior to the interview. The advance letter will explain the purpose of the survey, offer assurances of confidentiality, and include a toll-free number for respondents to call with questions or to complete the interview at their convenience. Participants will be given a check for $10.00 as a token of appreciation for completing the interview.

DATA PREPARATION

At the conclusion of each round of data collection, MPR will code and edit the survey data. Where feasible, responses that are missing will be imputed and an imputation flag will be created to indicate imputed values on the final data files. MPR will also create nonresponse-adjusted sample weights to enable unbiased survey estimates for the Nation and for primary subpopulations of interest. Finally, MPR will develop the variance estimation parameters and specifications to be used to estimate standard errors that will account for the complex sample design.

DATA PRODUCTS

Data Files. Two files will be produced for each survey round. The survey data file will include all survey data and data items from SSA administrative records that were used to implement the sample design. The sample file will include the full sample that was selected for the survey and the final disposition of each sample case (for example, completed interview, refused, or not located). At the conclusion of the study, the survey databases from all four years will be merged into a single database.

Codebooks. A machine-readable codebook will be developed to accompany each data file. For each data item, the codebook will identify the variable name, the number of positions occupied by the variable, the variable's location in the data record, the question from which the variable was derived, the population who was asked the question (for example, Ticket participants), and value codes and labels.

User's Manual. MPR will develop a User's Manual for data users that provides essential information for analysis. The User's Manual will include the final codebooks for the data files, means and frequency distributions, variance estimation parameters, and a discussion of survey, weighting, and variance estimation procedures.

Public Use File. MPR will also develop a public use file for use by researchers external to SSA. Specifications for these files will adhere to SSA guidelines for protecting the privacy and confidentiality of respondents. A disclosure analysis will be performed to prevent variables from being used to identify respondents This analysis will also examine other public data sets to assure that they cannot be used in combination with NBS to identify respondents. Appropriate techniques will be applied for masking identity, for example, data blanking and imputation, data modification or blurring, and data swapping. The public use file will also include the variance estimation parameters.

SCHEDULE

Data collection for the first round of the NBS occured between February 2004 and September 2004. Preliminary data files will be available in early 2005. Final first round data files and the first round public use file are expected to be complete in early 2005. Subsequent survey rounds will follow a similar schedule. The final combined data and documentation including all four survey rounds will be completed in the summer of 2008.