Page:M-21-19 Memorandum for Heads of Executive Departments and Agencies.pdf/18

 3. OMB may determine on a case-by-case basis that certain programs may still be subject to the annual IP reporting requirements. If this occurs, OMB will notify the program.

B. Phase 2: Report IP Estimates for Identified Susceptible Programs with a Statistically Valid Sampling and Estimation Methodology Plan Programs reporting IPs for the first time and programs revising their current IP Sampling and Estimation Methodology Plan (S&EMP) should conform to the process and content described in this guidance.

1. The main purpose of an IP estimate is to reflect the annual estimated known IPs made by the program. When developing the S&EMP, errors identified for one payment should never exceed the amount of the payment. It is important to note that the S&EMP should have a mechanism for identifying, accounting for, and estimating the annual UPs and the annual IPs separately.

2. The head of each agency is ultimately responsible for producing a statistically valid estimate of the IPs and UPs made under the program that is in Phase 2. Each agency has the responsibility of designing and documenting a program’s S&EMP with the mindfulness that during their annual compliance review, their OIG will take into account the accuracy of the IP and UP estimates and whether the S&EMP used is adequate and appropriate given program characteristics. The program will be responsible for producing an IP and UP estimate that is accurate and appropriate given program characteristics and it will be the OIG’s responsibility to evaluate whether the explanation provided by the program and the S&EMP without point estimates and confidence intervals around those estimates warrants compliance during the annual OIG compliance review. For purposes of this guidance, S&EMPs will be considered statistically valid if they produce point estimates and confidence intervals around those estimates. Agencies must work with their statistician to determine the appropriate confidence interval given program characteristics, available resources, and whether the estimate is reliable. If a program is unable to develop a S&EMP that produces a point estimate and confidence interval around the estimate then it must include in their S&EMP a detailed explanation as to why it is not possible.

3. To the extent possible, data used for estimating IPs and UPs in a given program should coincide with the FY being reported (for example, the estimate reported in the FY 2021 Annual Data Call would be based on data from FY 2021 (October 2020 through September 2021). The 12-month timeframe represented in the reported estimate should be documented in the S&EMP submission. For consistency purposes, the agency should continue using the same timeframe (i.e. October through September) for subsequent reporting years, unless a different timeframe needs to be used. If the timeframe needs to change for subsequent reporting years then the agency should resubmit their S&EMP and accompanying checklist with certification with the updated 12-month timeframe. Rh