Supporting Statement Part B (Final)

Supporting Statement Part B (Statistical Information) for

RSI/DI Quality Review Case Analysis - Sampled Number Holder; Auxiliaries/Survivors; Parent; Stewardship Annual Earnings Test

OMB: 0960-0189

2. Collections of Information Employing Statistical Methods

1. Statistical Methodology

We selected a statistically valid sample of all Old-Age, Survivors, and Disability Insurance Program (OASDI) beneficiaries in current payment status from a respondent universe of approximately 48 million. The sample involves approximately 1,800 cases. Each case in the sample may consist of the number holder (SSA-2930) and the auxiliary/survivor (SSA-2931), or the stewardship Annual Earnings Test (AET) workbook (SSA-4659). We use the stewardship review cases to measure the impact of the AET on benefits paid in the closed year (2 years prior to the sample month). The sample involves 325 cases. Based on experience, we anticipate an estimated 96% response rate. Before we make a determination to exclude a case because the beneficiary did not respond, we first attempt to verify their identity using other pieces of data found in SSA records, or by reviewing any recent record of contact in our Customer Service Response system. If we are able to verify their identity, then we attempt to contact any representative payee that exists to try to do an interview. If this is not successful, our manual has certain guidelines that exist that allow us to complete cases in certain circumstances. If the case does not meet those guidelines, then we exclude the case.

2. Procedures for Collecting the Information

We selected the sample from the Master Beneficiary Record through use of a random number generator. The ongoing stewardship review provides a dollar payment error rate of + .2 percent at the national level, with a 95 percent confidence level. The Stewardship sample is designed to give a precision of +/- 0.2% for the entire national OASDI sample. While there will be year to year variance, because the individual samples are smaller, on average the precision levels for the individual programs will be lower, with larger confidence intervals, in particular for DI and foreign records.

3. Methods to Maximize Response Rates

We check data for accuracy and consistency, and verify documents used to establish entitlement to benefits are at their source of origin. The sample yields reliable data that can be generalized to the universe.

SSA has no evidence to suggest that the four percent of selected individuals who do not respond and are subsequently excluded from the case review bias or undermine the validity of the RSI/DI Quality Review Case Analysis. This determination is in part based on demographic comparisons between the excluded case population and the general OASDI population. This comparison has found no statistically significant difference in demographics that might indicate that exclusions are non-random or introducing bias. SSA anticipates conducting additional research in FY21-22 to further evaluate potential exclusion biases among the OASDI Stewardship sample.

4. Tests of Procedures

Proven tests of the sampling procedure, since the stewardship reviews have been in existence for several years. We have gathered valuable information by refining the RSI/DI programs.

5. Statistical Agency Contact for Statistical Information

Danielle B. Harris, (410) 966-1547, SSA, Office of Quality Review.