2023-20023. Provisions To Improve the Supplemental Nutrition Assistance Program's Quality Control System

AGENCY:

Food and Nutrition Service (FNS), USDA.

ACTION:

Notice of proposed rulemaking.

SUMMARY:

The U.S. Department of Agriculture (the Department) is issuing this notice of proposed rulemaking to improve the Food and Nutrition Service's (FNS) Supplemental Nutrition Assistance Program (SNAP) quality control (QC) system as required in the Agriculture Improvement Act of 2018 (2018 Farm Bill). The proposed changes are intended to strengthen and improve the integrity and accuracy of the SNAP QC system and to better align SNAP with requirements in the Payment Integrity Information Act of 2019 (PIIA). These changes include a significant adjustment to the SNAP QC system that involves changes to Federal and State agency sampling processes, as well as changes to the active case review process. Quality Control case sampling and review processes are key aspects of the system used to annually assess SNAP payment error rates. The Department requests comment on this rule's proposed provisions.

DATES:

Written comments must be received on or before November 20, 2023 to ensure their consideration.

ADDRESSES:

The Food and Nutrition Service, USDA, invites interested persons to submit written comments on this proposed rule. Comments may be submitted in writing by one of the following methods:

— Federal eRulemaking Portal: Go to http://www.regulations.gov. Follow the online instructions for submitting comments.

— Mail: Send comments to John M., Branch Chief, Quality Control Branch, Program Accountability and Administration Division; Food and Nutrition Service; 1320 Braddock Place, 5th Floor; Alexandria, Virginia 22314.

— Email: Send comments to [e-mail]. Include Docket ID Number FNS–2020–0016, “Provisions to Improve the SNAP QC System” in the subject line of the message.

—All written comments submitted in response to this proposed rule will be included in the record and will be made available to the public. Please be advised that the substance of the comments and the identity of the individuals or entities submitting the comments will be subject to public disclosure. FNS will make the written comments publicly available on the internet via http://www.regulations.gov.

FOR FURTHER INFORMATION CONTACT:

John M., 703–457–7747, Food and Nutrition Service, 1320 Braddock Place, 5th Floor, Alexandria, Virginia 22314, [e-mail].

SUPPLEMENTARY INFORMATION:

I. Table of Contents

a. Acronyms or Abbreviations

b. Severability Clause

II. Background/History

a. SNAP QC—In General

b. Farm Bill Provisions

c. Improper Payment Determinations

d. Historical Information Supporting Congress' Request for Improvement and QC Reform

i. State QC Integrity Reviews (QCIRs)

ii. FNS Response

e. SNAP QC—Current Processes

f. New Proposed QC Approach

III. General (This section begins discussion of the proposed provisions)

a. Terminology Clean Up

b. Subpart A—Administration

i. Staffing Standards

ii. FNS Access to State Systems

iii. Federal Monitoring—Federal Subsampling

iv. Sampling—Federal Sub Sample

v. Federal Monitoring—Arbitration

c. Subpart C—Quality Control (QC) Reviews

i. Sampling—General

ii. Sampling Plan—Content

iii. Sampling Plan—Design

iv. Sample Size—Active and Negative Cases

v. Sample Size—Alternative Designs

vi. Sample Selection—Corrections

vii. Sample Frame—Active Cases

viii. Sample Universe—Active Cases

ix. Active Sample Allocation and Weighting

x. Review of Active Cases—General

xi. Review of Active Cases—Household Case Record Review

xii. Review of Active Cases—Field Investigation

xiii. Review of Active Cases—Personal Interviews

xiv. Review of Active Cases—Collateral Contacts

xv. Review of Active Cases—Variance Identification

xvi. Review of Active Cases—Variances Excluded From the Error Analysis

xvii. Review of Active Cases—Other Findings

xviii. Review of Active Cases—Reporting of Review Findings

xix. Review of Active Cases—Disposition of Case Reviews

d. Subpart E—Corrective Action

i. Corrective Action Planning—Negative Cases

ii. Corrective Action Planning—Incomplete Cases

e. Subpart F—Responsibilities for Reporting on Program Performance

i. Quality Control Review Reports—Mandating the Use of SNAP QC System

f. Subpart G—Program Performance

i. Determination of State Agency Program Performance—Determination of Payment Error Rates

ii. State Agency Error Rates—Completion Rate Penalty

iii. High Performance Bonuses

iv. Performance Measures—Program Access Index

g. Proposed Timeframe for Implementation of QC Sampling and Active Review Changes

IV. Procedural Matters

V. Amendatory Text

Acronyms or Abbreviations

Agriculture Improvement Act of 2018 (Pub. L. 115–334), the 2018 Farm Bill

Case and Procedural Error Rate, CAPER

Code of Federal Regulations, CFR

Corrective Action Plan, CAP

Department of Justice, DOJ

Federal Quality Control Reviewer, FQCR

Food and Nutrition Act of 2008, FNA

Food and Nutrition Service, FNS

Office of Management and Budget, OMB

Payment Error Rate, PER

Quality Control, QC

Quality Control Reviewer, QCR

Regional Office, RO

Request for Information, RFI

Supplemental Nutrition Assistance Program, SNAP

Supplemental Nutrition Assistance Program's Automated Quality Control System, SNAP–QCS

State Quality Control Reviewer, SQCR

U.S. Department of Agriculture, the Department or USDA

U.S. Department of Agriculture Office of Inspector General, USDA OIG

Severability: The Department proposes that certain individual components of this proposed rule are severable and seeks comment on that proposal. Specifically, the Department considers changes proposed in the following sections to be severable: Staffing Standards; FNS Access to State Systems; Federal Monitoring—Arbitration; Review of Active Cases—Other Findings; Review of Active Cases—Disposition of Case Reviews (with the exception of changes proposed to 7 CFR 275.12(g)(2)); Corrective Action Planning—Negative Cases; Corrective Action Planning—Incomplete Cases; State Agency Error Rates—Completion Rate Penalty; and Performance Measures. If a court were to find unlawful any or some combination of this rule as finalized, the Department still would intend any individual or combination of the above sections of this proposed rule to stand. The Department seeks comment regarding considerations about whether stakeholders consider those and any other provisions in this proposed rule severable or not.

Background

SNAP QC—In General

The Supplemental Nutrition Assistance Program (SNAP) is the nation's largest domestic food assistance program for Americans, reaching about 40 million people (approximately 12 percent of the nation's population) per month during fiscal year 2020.^[1]

Although the Federal government funds SNAP benefits, under 7 U.S.C. 2020(a)(1), State agencies are responsible for general program administration of SNAP within their States, including determining the eligibility of individuals and households to receive SNAP benefits and issuing monthly allotments of benefits.

However, given the large volume of SNAP cases, complexities of eligibility policies, and availability of State options, State agencies may issue overpayments or underpayments of SNAP allotments to participant households.

Pursuant to Section 16 of the Food and Nutrition Act of 2008, as amended (FNA), each State agency is responsible for monitoring and improving its administration of SNAP. A Quality Control (QC) system is necessary to help ensure State agencies measure improper payments and improve their administration of SNAP. SNAP QC reviews have four goals, identified at 7 CFR 275.10(b), which are to provide: (1) a systematic method of measuring the validity of the SNAP eligibility caseload; (2) a basis for determining all SNAP error rates; (3) a timely, continuous flow of information on which to base corrective action at all levels of administration; and (4) a basis for establishing State agency liability for payment errors that exceed the National performance measure pursuant to Section 16(c)(1)(C) of the FNA.

To comply with Section 16 of the FNA, State agencies conduct monthly reviews of a statistically representative sample of both participating SNAP households (active cases) and households for whom participation was denied, terminated, or suspended (negative cases). These reviews measure the accuracy of SNAP eligibility and ongoing allotment determinations and ultimately serve as the basis for the SNAP payment error rate (PER) and case and procedural error rate (CAPER).^[2] The results of these reviews provide feedback on State-by-State and national administration of the Program, including how State agencies' chosen policy options, waivers, and business processes affect the accuracy of their eligibility determinations. In short, the QC system allows FNS and the States to assess the integrity of SNAP by determining the extent to which the program is operating as required by statute and regulations. The system directly measures the accuracy of State actions to certify households as eligible for SNAP allotments and to determine the amount of those allotments, which are the actions that States should prioritize to ensure effective stewardship of taxpayer dollars and effective service to households in need.

Farm Bill Provisions

On December 20, 2018, the President signed Public Law 115–334, the Agriculture Improvement Act of 2018 (2018 Farm Bill). Section 4013(b) of the 2018 Farm Bill required the Department to issue an interim final rule to: (1) ensure the SNAP QC system produces valid statistical results; (2) provide for the oversight of contracts entered into by a State agency to improve payment accuracy; (3) ensure the accuracy of data collected in the QC system; and (4) provide for the evaluation of the integrity of the QC system for a minimum of two State agencies per fiscal year. Section 4013(e) of the 2018 Farm Bill also required that cost sharing for State computerization costs be, in part, contingent on State agencies granting FNS access to all State computer systems containing documentation and evidence related to SNAP eligibility. The Department determined that the most effective way to meet the statutory requirements was to issue two rules: (1) an IFR for the non-discretionary provisions, which FNS believed were necessary to comply with the 2018 Farm Bill and would be effective immediately, and (2) a proposed rule for the additional, discretionary provisions to improve the integrity and data quality of SNAP QC. By doing so, the Department ensures that major discretionary changes to SNAP QC go through the full notice and comment process, which allows stakeholders an opportunity to be part of the rulemaking process.

The Department codified all the non-discretionary requirements in Section 4013(b) in SNAP regulations by publishing the interim final rule (IFR) titled, “Supplemental Nutrition Assistance Program: Non-Discretionary Quality Control Provisions of the Agricultural Improvement Act of 2018” on August 13, 2021 (86 FR 44575.^[3] ) A correction to that interim final rule was published on September 2, 2021 (86 FR 49229 ^[4] ) and the final rule was published on April 18, 2023 (88 FR 23559 ^[5] ).

Improper Payment Determinations

The SNAP QC review process precedes Federal improper payment laws, including the Payment Integrity Information Act of 2019 (PIIA) (Pub. L. 116–117.^[6] ) These PIIA requirements include: identifying any case that a reviewer is unable to determine the accuracy of (known as an incomplete case) as an improper payment (31 U.S.C. 3352(c)(2)); and measuring technically improper payments, or payments in which a recipient was entitled to a payment but the payment failed to follow statutory or regulatory requirements (31 U.S.C. 3351(4)). With Congress' requirement to improve the SNAP QC system, the Department is including proposed changes in this rule that will better align SNAP with requirements in PIIA. In addition, the proposed shift in review focus for active cases (discussed later) will align the SNAP QC review process for determining SNAP payment errors more closely with the processes used to determine improper payments in other Federal programs that provide benefits to similar populations ( e.g., Medicaid and the Earned Income Tax Credit).

Historical Information Supporting Congress' Request for Improvement and QC Reform

The section that follow discusses challenges FNS faced and overcame that impacted the integrity of the QC system in the recent past; specifically, the challenges that occurred when State agencies introduced bias into the QC system's data. While FNS addressed the issues and has since regained confidence in the integrity of the data collected by State agencies, additional improvements and changes to the QC system will help ensure SNAP's QC system will be less susceptible to bias. The historical information below offers context and support for Congress' request for improvement and the reform proposals in this rule.

In 2015, an audit of SNAP QC's error rate determination process by the Department's Office of Inspector General (OIG)),^[7] and FNS's study on Enhancing Completion Rates for SNAP QC Reviews,^[8] both identified issues with the reliability of State-reported QC data. The OIG recommendations included that FNS amend and enforce its policies to ensure payment error rates are accurate and in compliance with regulations, and that FNS eliminate the two-comparison allotment test process, which is used to determine payment errors. FNS's Completion Rate study identified QC system issues including: the performance bonus system's creation of financial incentives for State agencies to underreport payment errors (which was subsequently eliminated by the 2018 Farm Bill), and State agencies' improper coding of QC cases as incomplete in order to reduce their payment errors.

State QC Integrity Reviews (QCIRs)

To identify the full extent of the QC data reliability issues, FNS began QCIRs in April 2015 and completed reviews of all 53 State agencies by September 2016. The purpose of the QCIRs was to validate that State agency QC systems did not include bias, as bias impacts the integrity of the data. During these reviews, FNS was able to validate only 11 State agencies' QC systems for FY 2015 and found data integrity issues in the remaining 42 States. As a result, FNS was unable to use the States' QC data to establish State or national SNAP payment error rates for FY 2015 and FY 2016.

The findings from the reviews fell into four categories related to integrity in the QC system and aligned with ways to mitigate errors. The first category centered on inadequate documentation. FNS found more than half of the State agencies were not properly documenting information in the QC case file, as required in the SNAP QC Review Handbook (the FNS–310).^[9] Inadequate documentation prevents FNS from having the information needed to do a thorough and independent validation of the cases completed by State agencies.

The second category concerned improper use of error review committees. State agencies may use error review committees after State quality control reviewers (SQCRs) transmit completed QC cases to FNS, to identify corrective actions that may be needed to improve the accuracy of eligibility and allotment determinations. However, State agencies are strictly prohibited from using error review committees prior to the SQCR transmission of completed QC cases to FNS. On August 1, 2005, after identifying concerns regarding the improper use of error review committees, FNS issued a policy memorandum ^[10] on the proper use of an error review committee to develop corrective action to prevent future errors. The memo reiterated that FNS prohibited the use of such committees to mitigate errors in cases actively under review. Despite this longstanding guidance, in the FY 2015 QCIRs, FNS discovered more than half of the State agencies continued to use error review committees inappropriately to discuss and mitigate the errors found in cases actively under review before submitting review results to FNS.

The third category of findings from the QCIRs related to the failure of more than half the State agencies to disclose complete case information to FNS, as required by 7 CFR 275.23(b). This finding included some State agencies refusing to give FNS access to their systems for auditing purposes, as required by Section 16(c)(4) of the FNA. Section 4013(a)(2) of the 2018 Farm Bill addressed this and required State agencies to give FNS access to their State systems for QC and oversight purposes. Section 4013(e)(2) of the 2018 Farm Bill required cost-sharing for computerization to be conditioned on State agencies being able to provide FNS access to their systems for audit and inspection purposes. The 2018 IFR codified the requirement to give FNS access, and this proposed rule includes the cost-sharing condition from Section 4013(e)(2) of the 2018 Farm Bill for State agencies to give FNS access.

The fourth category of findings from the QCIRs found that more than half of the State agencies were incorrectly interpreting and applying certification policy, sometimes intentionally, to mitigate QC errors. For example, during QC reviews, some State agencies were using an inappropriate number of weeks or months to determine a household's earned income. This approach led to the QC reviewer artificially increasing or decreasing the earned income amount of the household, with the goal of obtaining a result that more closely matched the amount used at the eligibility and allotment determination, thus mitigating the likelihood and potential impact of a payment error.

As a result of FNS's initial integrity findings in 2015, the Department of Justice (DOJ) began investigating State agencies for violations of Federal law in connection with the underreporting of SNAP QC payment errors. Between 2017 and 2021, DOJ settled with eight States for False Claims Act allegations of introducing bias into their QC processes. In addition, DOJ settled with a contracting company that provided QC support to all eight State agencies.

FNS Response

In January 2016, in response to early findings from the QCIRs, FNS issued a policy memorandum ^[11] to address concerns with the integrity of State agency QC systems. The memorandum reiterated and clarified FNS policies that are necessary to prevent bias from entering the QC system.

In September 2016, after completing all QCIRs, FNS issued revisions to its primary QC case review policy manual, the FNS Handbook 310,^[12] to help address FNS' and OIG's concerns regarding the integrity of the QC system. These revisions reinforced the January 2016 policy memorandum and addressed other issues affecting QC system integrity, including documentation and verification in QC reviews. By November 2016, FNS had trained all Federal quality control reviewers (FQCR) and SQCRs on the new manual and integrity-related provisions.

During the QCIRs, and while working with State agencies to address findings and resolve corrective action plans, FNS found the complex structure of QC reviews contributed to not only unintentional mistakes, but also to an environment that allowed for both the manipulation and mitigation of SNAP QC payment error findings. For instance, the QC review begins with an examination of the sample month circumstances for a household, which could be any month of the household's certification period. However, because of the complexity of Federal SNAP certification policy, as well as State SNAP policy options and waivers, reviews regularly involve a reviewer assessing household circumstances across multiple months. FNS determined this complexity creates an opportunity for both purposeful errors and unintentional review mistakes and, in some cases, can blur the connection between payment errors found during the QC review process and the relevant certification actions, which is where errors are occurring.

Often it is not possible for SQCRs to complete cases due to difficulties associated with obtaining verifications for past months and securing household cooperation. For instance, to simplify administrative procedures, some State agencies have chosen to implement a policy that allows them to forego verification of shelter expenses at certification if the expenses are not questionable. In the QC system, there are no comparable options, and these expenses must always be verified. While certification rules allow for a waiver of the requirement to verify expenses that are not questionable,^[13] these rules do not waive the requirement that correct expenses be used to determine program eligibility and allotments. SNAP households may then be skeptical of the SQCR's request for verifications of their shelter costs since it was not required for the certification worker to determine an allotment amount, increasing the odds that the household will not comply with the QC review. Additionally, the household may not be able to supply verification because they do not have a copy of the required record or did not retain it, not having needed it previously. In this circumstance, the SQCR would determine the case incomplete since the required verifications were unavailable.

The conclusion that the complexity of SNAP eligibility policy and the QC system's complexity played a major role in many of the integrity issues led FNS to explore ways to simplify and improve the QC review process. On June 1, 2018, the Department published a Request for Information (RFI) in the Federal Register (83 FR 25425 ^[14] ) to solicit input on how best to simplify and improve the QC system. FNS received 26 unique comments in response to the publication. This preamble references substantive comments from the 2018 RFI in the relevant sections that follow.

FNS also conducted focus groups with stakeholders who work in State SNAP QC and certification policy, social science research partners with government programs, and hunger advocates on conceptual ideas for the proposed provisions in this rule. These focus groups assisted FNS in identifying areas of concern, particularly with changing the review focus to a narrower eligibility scope.

Additionally, during national and regional conferences, State and Federal staff verbally expressed to FNS an interest in narrowing the QC review to focus solely on the eligibility action to simplify the QC review and increase compliance with QC review requirements while still maintaining a focus on measuring the validity of critical State agency decisions. This narrowed focus is consistent with how other social safety net programs ( e.g., Medicaid, earned income tax credit) measure improper payments. These other programs do not measure ongoing administration of benefits, as is done in SNAP's current QC review process. Creating such an alignment would allow Congress and others to compare improper payments more effectively across social safety net programs.

SNAP QC—Current Processes

To assist the reader in understanding the review process changes being proposed, this section discusses how the QC review process is currently structured. The SNAP QC system ^[15] consists of two tiers, a State tier and a Federal tier. At the State agency level, a statistician develops a sampling plan consistent with Federal regulations at 7 CFR 275.11 that the State agency then submits to FNS for approval. Each month, according to that sampling plan, State agencies select a sample of active cases (7 CFR 275.12) and negative cases (7 CFR 275.13) from the universe of their SNAP caseload and conduct reviews of the cases to determine the accuracy of the determination and allotment amount. Active cases are comprised of those households who are participating in SNAP and negative cases are comprised of cases in which the State agency acted to deny, terminate, or suspend a household or applicant (7 CFR 275.10(a)).

SQCRs must schedule and conduct face-to-face interviews with households selected for review in the active sample frame (7 CFR 275.12(c)(1)).^[16] There are few exceptions to a face-to-face interview for QC. For instance, State agencies may conduct telephonic interviews for hard-to-reach Alaskan households (7 CFR 275.12(c)) or in situations of a declared disaster using the SNAP waiver process (7 CFR 272.3(c)). SQCRs use the information gathered during interviews along with information in the case files, various databases, and documentation from collateral contacts, such as neighbors, banks, and employers, during the review (7 CFR 275.12(c)(2)).

The review itself consists of an examination of a random month during a case's certification period, called the sample month (7 CFR 275.12(a). SQCRs verify all factors of eligibility and allotment issuance for that sample month and calculate an allotment amount based off that information, which is referred to as the household budget calculation (7 CFR 275.12(e)). They then compare that calculated amount to the amount that was issued to the household for that month This is called Comparison I. If Comparison I results in the household budget calculation producing an allotment amount for the sample month that is either over or under the household's authorized issuance amount for that month, in an amount that exceeds the current fiscal year's threshold for excluding small errors,^[17] the SQCR will conduct a second comparison, called Comparison II. With Comparison II, the SQCR would then examine the month in which the most recent certification action occurred. The SQCR would correct any mistakes made at the time of this certification action, compute another budget, calculate an allotment amount based on those findings, and compare the SQCR budget for the certification action to the amount that was authorized at the certification action. If there is also a difference in the SQCR's budget and the authorized budget for Comparison II, the SQCR uses the lesser of the two determined error amounts as the case's payment error.^[18] After States transmit cases to FNS, any QC finding of an overissuance must be reported to the State office responsible for claims, and underissuance cases must be reported to the State offices responsible for supplements, to evaluate and address whether an action must be taken on the reviewed cases (7 CFR 275.12(f)). This requirement is not impacted by this rule.

For all active frame cases,^[19] SQCRs enter their data into either a paper or automated version of the form FNS–380, Worksheet for QC Reviews (OMB Control Number: 0584–0074; Expiration Date: 07/31/2025), to document the information needed to make a determination about the accuracy of the case (7 CFR 275.21(b)). State agencies also code the results of the reviews for all active cases onto the FNS–380–1, the QC Review Schedule (OMB Control Number: 0584–0299; Expiration Date: 07/31/2023; currently under review with OMB) (7 CFR 275.21(b)). State agencies then submit the completed forms through SNAP's Federally funded automated computer system, SNAP Quality Control System (SNAP–QCS), for transmission to FNS (7 CFR 275.21(b)(1)). For negative frame cases, State agencies complete form FNS–245, Negative Case Review Schedule (OMB Control Number: 0584–0034; Expiration Date: 1212/31/2024) (7 CFR 275.21(b)). and submit completed forms through SNAP–QCS (7 CFR 275.21(b)(1)).

FNS then works to validate the State agency findings. FQCRs, who are in FNS regional offices, review a subsample of the active cases reviewed by State agencies (7 CFR 275.3(d)(1)(i)). The subsample is comprised not only of cases State agencies complete, but also of all active cases the State agencies selected but were unable to complete, and all cases that were determined by the State agency to be “not subject to review.” ^[20] After FQCRs identify the cases for the subsample, they then request the relevant QC case file documents from the State agencies and conduct a comprehensive, independent review of each case (7 CFR 275.3(d)(1)(ii)).^[21] After case reviews are complete and final determinations made, FNS calculates two error rates: (1) the payment error rate (PER) for the active sampling frame (7 CFR 275.3(d)(1)); and (2) the case and procedural error rate (CAPER) for the negative sampling frame (7 CFR 275.3(d)(3)). FNS also calculates an annual application timeliness rate for State agencies with data from the active sampling frame.

The PER is based on the difference between the amount of allotments issued to households and the amount those households should have received had their cases been processed correctly (7 CFR 275.23(b)). The overall PER is the sum of two breakdowns: underpayment and overpayment error rates. For CAPER, the review of negative actions considers procedural aspects of case processing in addition to whether the action to deny, suspend, or terminate a household was accurate. Procedural components of this review include timeliness of the action and accuracy of the household notification, among other things. The negative review results in a determination of whether the negative actions were valid or invalid (7 CFR 275.13(c)(1)). The CAPER is based on what percentage of the negative actions reviewed were invalid compared to the total number of negative actions processed by the State agency.

New Proposed QC Approach

The rest of this preamble will share the Department's proposals for this rulemaking. The Department proposes to improve the SNAP QC active review process using lessons learned from the QCIRs, the OIG audit, the completion rate study on QC, the 2018 RFI ^[22] on QC, and other activities, such as technical assistance, that FNS has completed since FY 2015. As such, the Department is proposing a different active case review that: (1) would be less complicated, thus making it easier for State agencies and households to comply with, FNS to oversee, and less susceptible to bias; and (2) would still be capable of collecting detailed program information regarding State agency administration and the over and under issuance of SNAP allotments. In addition, the proposed active review would resolve an ongoing concern from OIG about how SNAP QC's use of Comparison I and Comparison II resulted in a measurement from two different points in time.

SNAP regulations require all SNAP State agencies to operate a QC system (7 CFR 275.10(a)). Although the 2018 IFR increased various reporting and recordkeeping requirements for State agencies to comply with the 2018 Farm Bill, the active review process for the SNAP QC system has remained relatively unchanged since February 1984 (49 FR 6292) ^[23] and can be improved upon. Since SNAP's QC system existed at least 25 years prior to the first Federal improper payment laws, such as the Improper Payments Information Act of 2002 (IPIA). SNAP's active QC review methodology and processes require adjustment to come into alignment with other Federal programs where the improper payment determination procedures were created after the first Federal improper payment laws and associated regulations and guidance came into existence.

The proposed process would change the active case review to focus on eligibility actions instead of a random, point-in-time review for each case. This proposed change would eliminate the two-comparison allotment tests (Comparison I and Comparison II) from the QC review and instead focus on eligibility determinations, consider cases that a State agency cannot validate as total dollar errors, and alter the sampling and sampling plan requirements to correspond to the changed review focus.

To help the reader understand the flow of the proposed provisions that follow, the provisions are organized in the same order as existing regulations found in the Code of Federal Regulations at 7 CFR part 275. It is important to note, this is a proposed rule that requires public comment before a final rule may implement any changes to SNAP's QC system. The Department proposes the following provisions for comment.

General

Prior to discussing the provisions, the Department would like to convey that this rule is very technical in nature. The provisions discussed include specific detail about QC review processes and detailed statistical formulas necessary to carry out the QC process. The Department will continue to explain technical information throughout this proposed rule when able; however, in areas where precision and specific vocabulary are necessary to accurately convey the proposal, such as with statistical formulae, an explanation in more colloquial terms could result in an inaccurate portrayal of the proposed concepts. Thus, the technical language will remain.

Terminology Updates

In 7 CFR parts 271 and 275, the Department proposes to update or remove outdated and duplicative terminology and update currently applicable terminology to reflect the proposed review process. The following terms are no longer current: “negative case error rate” (or “negative error rate”) and “Immigration and Naturalization Services (INS).” Therefore, the Department proposes to change these to “case and procedural error rate (CAPER)” ^[24] and “United States Citizenship and Immigration Services (USCIS),” respectively. The Department proposes to update the definition of “overissuance” to remove reference to paper food stamp coupons. The Department proposes to revise the definitions for “active case,” “error,” “review date,” and “sample month” in 7 CFR 271.2 to reflect changes made by this proposed rule. Additionally, the term “active case error rate” is currently used interchangeably with the term “payment error rate” throughout 7 CFR 275. To avoid confusion, the Department proposes to use “payment error rate” consistently throughout 7 CFR 275.

Subpart A—Administration

7 CFR 275.2(b)—Staffing Standards

The Department proposes to add a provision in the Staffing Standards section under 7 CFR 275.2(b) to specify the expectation that State agencies ensure the independence and objectivity of the merit staff performing QC case reviews. The QC integrity reviews found instances where State agency managers were putting undue influence on SQCRs to find that the eligibility worker's initial determination was correct, as opposed to independently focusing on the accuracy of the case, which is the purpose of the QC review. A separate audit completed by USDA OIG further supported these findings, as well as action by DOJ to settle allegations of violations of Federal law with eight State agencies. These findings provide substantial evidence that SQCRs reporting in the organizational structure to the same individuals responsible for overseeing eligibility determinations has an adverse effect on their ability to objectively review cases and determine errors. To address these issues and ensure the accuracy of the PER and CAPER estimates, the Department proposes to require a new provision in 7 CFR 275.2(b) that State agencies take proactive measures to ensure SQCRs work independently and are free from undue influence by ensuring the staff used to conduct QC reviews operate independently from those responsible for overseeing the eligibility determination process to ensure objective and accurate assessments of the Performance Reporting System. For example, QC staff would not have the same immediate supervisor or director as the eligibility staff.

In addition, the Department proposes to clarify the components that constitute prior knowledge for purposes of staff disqualification under 7 CFR 275.2(b). The focus on integrity and bias in SNAP's QC system generated concerns in State agencies about SQCRs unintentionally biasing the QC process. FNS has heard from State agencies, anecdotally, that QC workers were concerned that a staff person might know information about a household one of their colleagues was reviewing but felt that disclosing that information would be considered a prohibited practice. The Department proposes to address this issue by removing the current language at 7 CFR 275.2(b) and adding a modified provision to QC's Staffing Standards at 7 CFR 275.2(b)(2) clarifying that even though State agency staff must disqualify themselves from directly working on a QC review if they have prior knowledge of a household, they are allowed to participate in a QC review as a collateral contact ^[25] of the household. In such situations, the staff must follow the governing rules regarding collateral contacts at 7 CFR 273.2(f)(4)(ii).

7 CFR 275.2(d)—FNS Access to State Systems

Section 4013(e)(2) of the 2018 Farm Bill—which amends Section 16(g)(1)(A) of the Food and Nutrition Act of 2008 (codified as 7 U.S.C. 2025(g)(1)(A))—specifies that cost sharing for computerization for systems is an allowable SNAP administrative expense only when, among other criteria, the Secretary determines the systems to be accessible to the Department for review and audit purposes. Therefore, to implement this statutory provision, in the paragraph at 7 CFR 275.2(d), the Department is requiring cost sharing for State agency system costs to be conditioned on, in part, FNS having access to all State agency records and systems in which those records are contained.^[26] The Department believes this will incentivize State agencies to make the necessary changes to come into compliance with the requirement of allowing remote access to State agency computer systems so that Federal QC staff have access to the full case record for all QC sampled cases to ensure the accuracy of the collected data. FNS will continue to work with State agencies collaboratively to establish data sharing and system integrity agreements to facilitate the required systems access. This new sentence in 7 CFR 275.2(d) will reference existing regulations and procedures for the suspension or disallowance of administrative funds found at 7 CFR 277.16 if State agencies do not comply with the requirement.

7 CFR 275.3(d)—Sampling—Federal Sub Sample

To be consistent with the changes proposed for State agency sampling in 7 CFR 275.11, discussed later in this rule, the Department proposes at 7 CFR 275.3(d)(1)(i)(A) to only have one minimum Federal subsampling calculation table to determine the Federal subsampling pull for the re-review of active cases. In addition to the change in Federal subsample size, the Department also proposes the Federal subsample be allocated across five strata in order to capture cases from each of the points in time where the State agency makes a determination to authorize or re-authorize benefits and to be consistent with the weighting that is proposed to determine the State agency's sampling frame. The five strata reflect all the possible action types and reporting periods. Since the reporting system and certification period length assigned at certification impact the number of reports that will end up in the sampling universe under the proposed sampling procedures, the five strata as well as weighting are used to ensure that all types of households are included in the sampling universe and to even out selection probabilities so that all cases have a chance to be selected for sampling in determining the SNAP PER. The five strata are based on the combination of action types ( i.e., certification, recertification, and when a required monthly, quarterly, or periodic report is due and an allotment is issued in the following month) and reporting period ( i.e., less than six months, exactly six months, and more than six months). FNS chose certifications, recertifications, and required reports with allotments in the following month because they are points in which the State agency must make an eligibility determination or terminate the household from the Program. FNS based the proposed strata on an analysis of the FY 2017 SNAP QC data (see table).

Strata	Action type	Reporting period	Frequency	Percentage
A	Certification/recertification	Less than 6 months	365	4.9
B	Redetermination	Less than 6 months	94	1.3
C	Certification/recertification	Exactly 6 months	3,873	52.2
D	Redetermination	Exactly 6 months	1,628	21.9
E	Any *	More than 6 months	1,457	19.6
* Includes the three actions: certification, recertification, and redetermination.

The proposed changes to determine the Federal subsample would be necessary to determine State agency PERs based on the new proposed methodology. Under the proposed changes, the Federal subsample size would be determined using the following table:

Average monthly reviewable caseload (N)	Federal subsample target (n')
60,000 and over	n′ = 400.
10,001 to 59,999	n′ = .005 N + 100.
10,000 and under	n′ = 150.

Once the Federal subsample size n' is determined using Table 1 above, the Federal subsample must be allocated across the five strata proportionally to the State final weights to ensure the Federal subsample is self-weighting and there is no loss of precision due to differential sampling probabilities. The Federal subsample size for each stratum shall be determined as follows:

In the table formulas above: N is the sampling universe/monthly caseload; F is the sampling frame; n is the state sampling size; n' is the Federal subsample size; W is the state sampling weights; and W' is the Federal sampling weights. For stratum i, the Federal subsample size n'_i shall be proportional to the sum of the final state weights for that stratum, ni×W_i = N_i, in other words

This means that the final weight for the cases selected for the Federal subsample from stratum i, W'_i, is given by:

for every i, which is a constant across the strata.

7 CFR 275.3(d)(4)(i)(A)—Federal Monitoring—Arbitration

Currently, when there is a dispute between a State agency and FNS regarding a finding (whether a case was correct, overissued, underissued, or was ineligible as of the review date) or disposition (whether a case was complete, not subject to review, incomplete, or the case was deselected) of a QC case, the State agency may request arbitration from the FNS Arbitrator. FNS's Arbitrator is the SNAP Administrative and Judicial Review Branch. This Branch includes administrative review officers who serve as a neutral third party, as they do not directly work with SNAP certification policy or quality control. The disputes subject to arbitration are limited to disagreements over finding or disposition only and the arbitrator's decision on a case is considered final and not subject to subsequent appeal. Any other disagreements should be handled through an informal resolution process, which is separate from the arbitration process.

FNS issued a policy memorandum dated November 7, 2003, that provided procedures for arbitrating application processing timeliness (APT) disagreements when the APT measure was first introduced. FNS intended the memorandum to be temporary, covering FY 2003 only, but did not identify an expiration date. The preamble in the final rule, Quality Control Provisions of the Mickey Leland Childhood Hunger Relief Act (62 FR 29652 ^[27] ), published June 2, 1997, detailed the reasoning for why arbitrations were only to be used for finding and disposition disagreements. The Department noted that arbitrating “agree cases,” which encompasses disagreements solely on APT, would adversely impact the accuracy and timeliness of the arbitration process. FNS discovered in 2017 that some FNS regional offices were still allowing their State agencies to arbitrate APT-only disagreements. The FNS national office provided clarification through policy guidance, but to clarify the Department's original intent in the 1997 Rule, the Department proposes to update language in 7 CFR 275.3(d)(4)(i)(A) to clarify that, other than those circumstances specified in regulations, other disagreements would not be provided arbitration rights.

In addition, current regulations instruct State agencies to send arbitration requests to their FNS regional office, addressed to the attention of the FNS Arbitrator. Over time, FNS found that by providing the Arbitrator's contact information for State agencies to make direct arbitration requests, FNS was not only able to provide better customer service, but also have a more efficient request process. Therefore, the Department proposes to update its regulations at 7 CFR 275.3(d)(4)(iv) to require that a State agency send its request for arbitration directly to the FNS Arbitrator and copy the appropriate FNS regional office.

Subpart C—Quality Control (QC) Reviews

7 CFR 275.11—Sampling—General

Currently, the universe (all cases with the possibility of being selected for the QC sample) for SNAP's active frame includes all households receiving SNAP allotments in any given sample month. Under this proposed rule, the active frame sample universe would no longer include all households receiving SNAP allotments in any given sample month. Instead, only those households that experienced an eligibility action— i.e., certification for SNAP, recertification for SNAP, or requirement to submit a required monthly, quarterly, or periodic report in the sample month and an allotment is issued in the following month—would constitute the universe for the active frame. The following paragraphs in this preamble explain the proposed new sampling plan in 7 CFR 275.11, which would then be reviewed according to the proposed sampling review procedures in 7 CFR 275.12.

7 CFR 275.11(a)(1)–(2)—Sampling Plan—Content and Criteria

Currently, at 7 CFR 275.11(a), State agencies have two options for designing their minimum QC sampling size: a standard or a reduced sample size. The Department proposes to remove the language about alternative sampling designs at 7 CFR 275.11(a)(2) since, unlike the current methodology, the proposed methodology would not be conducive to State agencies utilizing a reduced sample size due to issues with reliability of the estimates that would result from the sample size reduction. This would include the removal of alternative sampling related rules in subsections 7 CFR 275.11(a)(2)(iii) and 7 CFR 275.11(a)(2)(iv). In addition, the Department proposes to add language in section 7 CFR 275.11(a)(2)(ii) about the sample size selection to help State agencies ensure they follow the procedures set forth in subsections (b), (c), (d), (e), and (f).

7 CFR 275.11(a)(3)—Sampling Plan—Design

Computer programs and systems that assist in selecting probability samples have progressed considerably since the publication of the current QC regulations in 1977. Now, systematic sampling is but one of the many possible ways to select a probability or random sample, and there are new statistical software tools available that can easily draw a random sample without using systematic sampling. The Department proposes to amend the recommendation at 7 CFR 275.11(a)(3) that State agencies should primarily utilize systematic sampling to also recommend that State agencies be open to considering other sampling software tools.

In addition, the Department proposes to revise the language at 7 CFR 275.11(a)(3), describing the proposed active sampling design. This revised language provides specific details about how State agencies would need to design their new sampling procedures to align with the new review process, including how to construct the new sampling frame and which variables to use to create the proposed strata (discussed further below). The revision would also clarify that each month, State agencies must select a sample size equal to one twelfth of the annual sample size specified in 7 CFR 275.11(b) (rounded to the next whole number) to ensure the sample accurately reflects the entire year of SNAP cases.

7 CFR 275.11(b)(1)–(2)—Sample Size—Active Cases and Negative Cases

The Department proposes to simplify and standardize the active case review process by (1) eliminating the current allowance for choosing between two different possible sample sizes and (2) increasing the current sample size for active cases specified in 7 CFR 275.11(b)(1)(iii) by 30 percent. Increasing the overall sample size ensures that the new sampling design would deliver the same level of precision as the current sampling design by providing for sufficient sample sizes within each stratum. Without this increase in overall sample size and proper allocation among the strata, the new design would over-represent cases with shorter reporting periods and under-represent those with longer reporting periods. Having a larger overall sample size is necessary to ensure statistical validity of the sample, which results in a more precise PER estimate. The Department proposes to remove the choice of the other sample size option, currently described at 7 CFR 275.11(b)(1)(ii), because in order for the new sampling methodology to provide a basis for calculating a national PER, all States must follow the same sampling design; therefore, a second option is no longer appropriate. Given the proposed deletion of 7 CFR 275.11(b)(1)(ii), the Department proposes to re-designate the remaining paragraphs of 7 CFR 275.11(b)(1) that follow paragraph 7 CFR 275.11(b)(1)(i).

In the final rule published on June 11, 2010, titled, “Supplemental Nutrition Assistance Program: Quality Control Provisions of Title IV of Public Law 107–171” (75 FR 33422),^[28] the regulations pertaining to SNAP negative case reviews (the review of cases that were denied, terminated, or suspended) were changed to emphasize customer service. This rule required State agencies to determine whether the action on the negative case under review was correct and whether the State agency correctly notified the household or participant ^[29] in clearly understandable language of the adverse action on their case.

The Department implemented this change in FY 2012, and later began informally referring to the associated error rate as the case and procedural error rate, as opposed to the negative error rate, to reflect the change in focus. This practice of referring to the error rate for negative case reviews as CAPER continues today and is the commonplace term State agencies and FNS use. As such, the Department proposes to amend 7 CFR 275.11(b)(2)(i) to include the explanation that negative case reviews result in determining the “case and procedural error rate.”

The Department found State agencies typically utilize the minimum annual sampling size for their negative case samples. Just as with the proposed removal of the alternative active case sampling design discussed earlier, the Department proposes to amend 7 CFR 275.11(b)(2)(i) to remove the alternative design option for the negative sample size so that sample sizes are standard, provide more precision for error rate estimates, and align with the language used for the selection of active cases in 7 CFR 275.11(b)(1)(i).

7 CFR 275.11(b)(4)—Sample Size—Alternative Designs

Consistent with the proposed removal of alternative design options for both the active and negative sampling frames, the Department also proposes to remove the regulatory provisions at 7 CFR 275.11(b)(4) through 7 CFR 275.11(b)(4)(iii) since those paragraphs discuss the options for State agencies concerning alternative QC sampling sizes.

7 CFR 275.11(c)(2)—Sample Selection—Corrections

To ensure they select enough cases to review annually, State agencies often pull supplemental samples when they find they do not have enough cases to meet FNS requirements. FNS refers to this act of pulling a supplemental sample in current regulations at 7 CFR 275.11(c)(2) as a `correction.' Corrections can be necessary for many reasons but most often occur because some cases in the sample were later determined to be not subject to QC review or because an increase in the average monthly reviewable caseload necessitated an increase in monthly sample size. The Department proposes to amend the provisions regarding corrections at 7 CFR 275.11(c)(2) to clarify that the new procedures for sample size, sample selection, sample frame, and sample allocation in proposed paragraphs 7 CFR 275.11(b), (c), (e), and (g) are also applicable to corrections, or in other words, when State agencies pull additional cases to compensate for under sampling.

7 CFR 275.11(e)—Sample Frame

Current regulations at 7 CFR 275.11(e) allow State agencies a `choice' in what their sampling frame must include. The Department proposes to remove language concerning the `choice' of the sampling frame since the Department's proposed methodology requires State agencies to use a specific frame for sampling.

7 CFR 275.11(e)—Sample Frame—Active Cases

The Department proposes to change its active case sampling frame description at 7 CFR 275.11(e)(1) to include only households that either experienced an initial certification action, a recertification action, or were required to submit a required monthly, quarterly, or periodic report in the sampling month and an allotment is issued in the following month. As noted earlier, this simplified QC process would shift the review focus to when an eligibility action occurred rather than a review of a random month within a household's certification period.

The simplification of only reviewing the eligibility action would ease compliance with QC requirements for both recipients and SQCRs by eliminating the complex aspects of the current review process and by making the SNAP review consistent with other benefit programs' improper payment review procedures. The complexities of the current process can contribute to State agencies having low QC case review completion rates and to, the integrity issues discussed throughout this proposed rule.

7 CFR 275.11(f)—Sample Universe—Active Cases

Regulations at 7 CFR 275.11(f)(1) inform State agencies which cases they must exclude from their sampling frames. Consistent with proposed changes in this rule to the sampling frame, the Department is also proposing corresponding changes for those cases the State agency must exclude. The Department proposes to amend one condition and remove one condition from the current list of excludable households for review at 7 CFR 275.11(f)(1).

The Department proposes to amend the provision at 7 CFR 275.11(f)(1)(iv) that addresses the exclusion of a household appealing an adverse action when the review date falls within the time period covered by continuing participation pending a fair hearing, or in other words, the time the household continues to receive benefits while they await their fair hearing. Instead of the condition being “when the review date falls within the time period covered by continuing participation pending the hearing,” the Department proposes to exclude all cases under review where the household is appealing an adverse action for any of the sampled actions, since the benefit determination is subject to change based on the result of the appeal.

The Department also proposes to remove the condition for exclusion at 7 CFR 275.11(f)(1)(v) that indicates a household can be excluded when the household is receiving restored benefits but is not participating based upon an approved application. This condition no longer applies to the new proposed sampling method because only approved applications will be included in the sample. As such, only the instruction that, “Other households excluded from the active case universe during the review process are identified in 7 CFR 275.12(g)” would remain in this paragraph.

7 CFR 275.11(g)–(h)—Active Sample Allocation and Weighting

In order for the new sampling method to be reflective of the SNAP caseload, the Department proposes to add two new paragraphs at 7 CFR 275.11(g), entitled “Active sample allocation” and 7 CFR 275.11(h), entitled “Weighting,” which will describe how State agencies would allocate the sample across the strata defined in the new sampling design and compute the weights needed to make the sample representative of the sample universe/target population.

As described earlier, the Department proposes to add requirements at 7 CFR 275.11(g) for the active sample allocation so that active cases are allocated to five pre-defined strata. These five strata are based on the combination of action type ( i.e., certification, recertification, and when a required monthly, quarterly, or periodic report is due and an allotment is issued in the following month) and reporting period ( i.e., less than six months, six months, and more than six months). This proposed change would ensure the resulting allocation will contain sufficient sample sizes within each stratum and cases from each stratum are represented in the sample.

In addition, the Department proposes to add a section, 7 CFR 275.11(h) “Weighting” to illustrate how to compute the weights needed to make the sample representative of the universe/target population. The new sampling design samples cases with unequal probabilities across the strata and, therefore, some cases have higher selection probabilities than others. State agencies must provide FNS the data necessary to compute the weights. To compensate for the unequal probabilities of selection across the strata, the Department is proposing the following formula for the weights W_i:

W_i = (N_i /F_i) × (F_i /n_i) = N_i /n_i for i = a,b,c,d,e,

Where:

—W_i is the weight for cases sampled from stratum i (the new proposed sampling design has five strata);

—N_i is the size of stratum i in the sampling universe/target population;

—F_i is the size of stratum i in the sampling frame;

—n_i is the number of cases sampled from stratum i.

—F_i /n_i is in the inverse of the sampling probability for cases sampled from stratum i (also called sampling weight), while N_i /F_i represents the post-stratification weights and corrects for the biases in the sampling frame. This ensures the sample represents the sampling universe, i.e., the SNAP caseload.

State agencies would be responsible for providing to FNS the counts N_i, F_i, and n_i (for i = a,b,c,d,e).

7 CFR 275.12—Review of Active Cases—General

By focusing only on eligibility actions, the Department anticipates the information obtained through the modified QC review process will be more useful for program improvement purposes, as it will be focused on the main touchpoints of a SNAP case and will be collected closer in time to those touchpoints. Currently, SNAP's error rates are a combination of errors measured at random points of time within a household's certification period and at points of eligibility. By narrowing the review, FNS would issue an error rate derived from the time of most recent eligibility action for all cases reviewed. This would allow FNS to more accurately identify where errors are occurring and allow State agencies to correct more effectively the causes of error at certification or the point of most recent eligibility action. FNS would also be collecting information at a point in time that is closer to when many actions and their corresponding errors are occurring, making the monitoring of State agency performance timelier than is possible under the current QC system.

Under the new proposed methodology, the PER would be representative of points in time when households have an obligation to report or confirm their circumstances and attest the information is accurate. In addition, at those points in time, the State agency also has an obligation to verify those household circumstances and accurately determine the household's eligibility and benefit level. With this new approach, payment errors would specifically reflect errors made directly as a result of the benefit determination process, a connection that has proven challenging with the current review system due to the distance in time between the selected review month and the action causing the errors. By identifying errors in a timelier manner, FNS could raise deficiencies in program operations more quickly and effectively, and State agencies and FNS would be able to address the root causes of errors more quickly than under the current QC system.

Another significant impact of moving from a random “point in time” review of the present QC system to this modified “eligibility action-only” review is that FNS would no longer require SQCRs to re-create all of a household's sample month circumstances for a random month within a certification period. Currently, to review the sample month (which can be at any point in time during a household's certification period), SQCRs must collect and validate the household's circumstances for that new month, which most of the time does not exist in the case record, to determine whether the allotment they received that month is accurate for what they should have received based on those circumstances. In order to determine whether the sample month's allotment determination by the SQCR is accurate, the SQCR must often request information about household circumstances for the past months in between the sample month and the most recent certification month, if they are not one in the same, making the review process very complex and prone to inaccurately applying QC policy since it relies on the ability to collect information from SNAP households for, in some instances, a year or more in the past. The reason these are so challenging to obtain is that SNAP households do not always have stable incomes, bank accounts, and contracts with cellular phone carriers or other utility companies, for example, that would make verifications easy to recall months later.^[30]

In the new approach, SQCRs would still be required to verify the household's circumstances for the sample month; however, the sample month would be the month the action took place, eliminating the need to look at previous months. As a result, the process would require fewer months to review, and much of the information a SQCR needs to verify the household's circumstances and determine the allotment's accuracy should already be available in the case record. In addition, if verifications do not already exist in the case record and SQCRs must either obtain this information from the SNAP household or collateral contacts, it would be less challenging for them to produce verification from the previous month or two than it would be for them to produce verification issued at some point in the previous four to twelve months, as is necessitated under the current system. Since this change would make obtaining verifications less challenging, compliance from households and collateral contacts will likely increase and SQCRs would be able to complete more of their QC case reviews, resulting in a more accurate PER.

Given the proposed change in review focus, FNS considered whether an ex parte review, in which the SQCR relies solely on the case record and does not contact the household, would be appropriate. FNS determined that relying solely on an ex parte review for QC reviews would not be appropriate for several reasons. First, the purpose of the QC review is to determine the accuracy of the allotment authorized by the State, and to accomplish that purpose, SQCRs must verify all household circumstances in the sample month. However, using allowable administrative flexibilities, many States choose to ease verification requirements at certification to reduce the administrative burden on applicants. As a result, verifications of all household circumstances may not be included in the case record and available for QC reviewers. Generally accepted government auditing standards (GAGAS) require reviewers to obtain sufficient, appropriate evidence to provide a reasonable basis for addressing review objectives and supporting their findings and conclusions.^[31] If SQCRs rely solely on the information in the case record, they may not have sufficient evidence from which to assess the accuracy of eligibility and benefit determinations. In addition, GAGAS require reviewers to have a certain level of independence when conducting quality control evaluations and audits, and to avoid situations that could lead reasonable and informed third parties to conclude they are not independent and thus are not capable of exercising impartial judgment.^[32] If SQCRs were to rely solely on the information in the case record, they may not be able to independently review all elements of eligibility and benefit determinations.

This proposed review would still require a complete and independent review of the household's circumstances to assess whether the State agency accurately determined the household's eligibility and SNAP benefit amount. By focusing on the eligibility action and not on a random sample month, FNS hopes to place greater emphasis for corrective action on making more accurate benefit determinations through better documentation, verification, and accountability.

The change in review procedure would also eliminate the current SNAP QC Comparison I and Comparison II process when determining a QC case's reportable error (an error that is above the current fiscal year threshold for excluding small errors).^[33] As described earlier, SQCRs and FQCRs conduct up to two assessments, Comparison I and Comparison II, to determine the final error amount in a case. In Comparison I, the reviewer determines the accuracy of the benefit received by the household based on the household's sample month circumstances. If the reviewer finds an error above the national error tolerance threshold (presently $54) in Comparison I, the reviewer completes Comparison II by examining the accuracy of the certification action that authorized the sample month's benefits. The final error amount is the lesser error amount of the two comparisons. In OIG's audit report titled, “FNS Quality Control Process for SNAP Error Rate,” ^[34] OIG asserted that these comparisons measure two different points in time and suggested FNS should take action to increase consistency in the PER measure. FNS has determined the improvements to the quality and consistency of QC data inherent with the proposed approach would address OIG's concerns regarding the Comparison I and II processes. Along with these changes, the Department also proposes to eliminate the concept of “as of the review date” throughout 7 CFR 275.12, since eligibility actions would now be the critical focus of the QC review and this phrase would no longer be meaningful.

7 CFR 275.12(b)—Review of Active Cases—Household Case Record Review

The process of reviewing a case in QC has several distinct components, including the household case record review, where the reviewer gathers information and evidence from the case record to determine what occurred in the case and to plan for the next fact-finding phase.

The Department proposes to amend the regulations at 7 CFR 275.12(b) to update the case record review provisions to reflect the new proposed review process. The proposed change specifies that the case record must include the initial certification or recertification application or, the monthly, quarterly, or periodic report, and QC worksheets applicable to the sample month determination. Documentation contained in the eligibility case record is allowable as verification, but only if the evidence used in the eligibility determination meets or exceeds QC verification standards found in the FNS Handbook 310. The Department also proposes to eliminate the requirement at 275.12(b) that reviewers examine the household issuance record for pertinent information if they cannot find the household's case record, as the case record review would be an essential component of the active review process. Under the proposed rule, situations where the case record could not be located would result in an incomplete case and, therefore, the total dollar amount issued would be reported as an error. Further discussion on incompletes becoming total dollar errors is in this preamble under the heading, Review of active cases—Disposition of case reviews.

The purpose of this proposed change is two-fold. First, including dropped/incomplete cases in the error amount would strengthen integrity by acting as a deterrent against the types of dropped-case manipulation that previously led to unreliable State PER data reporting, discussed in the history part of this preamble. Second, by including dropped/incomplete cases now excluded from the improper payment rate calculation, the proposed change would align SNAP with PIIA and other Federal programs, as noted in an audit released on July 6, 2016, by the General Accounting Office.^[35]

7 CFR 275.12(c)—Review of Active Cases—Field Investigation

The field investigation is another fact-finding phase of the QC review and includes interviewing the household and collateral contacts. Collateral contacts can include, for example, individuals, organizations, governmental agencies, and businesses that hold relevant information on the household's circumstances.

While conducting QCIRs, FNS staff found that State agencies do not widely use technology to collect documentary evidence during the field investigation phase. For example, while some SQCRs have access to digital scanners or cellular phone cameras to make electronic copies of verifications, many SQCRs have no access to these technologies during the field investigation or were not aware the use of electronic devices was a permissible way of obtaining documentary evidence that could not be brought back to the office.

Additionally, in the 2018 RFI,^[36] FNS asked for recommendations to encourage greater use of technology that could enhance the accuracy of case reviews. A prevalent response was that more funding and grants from the Federal Government would assist in State agencies using additional technology. The Department emphasizes here that, under current regulations at 7 CFR 277.3, expenditures for technology to aid in program administration, including in association with QC reviews, qualifies for reimbursement up to 50 percent.

FNS hopes that promoting the use of technology in regulations to collect documentary evidence from both households and collateral contacts would encourage State agencies to offer households better customer service by limiting the need to send in a physical copy of documentary evidence, a step that can prove challenging for many. Therefore, the Department proposes to amend paragraph 7 CFR 275.12(c) to provide that the use of technology is not only permitted, but encouraged, to obtain verification, including copies of documentary evidence from households and collateral contacts, so long as the privacy of the household and the information gathered are protected pursuant to applicable Federal and State privacy laws.

7 CFR 275.12(c)(1)—Review of Active Cases—Personal Interviews

QC regulations currently require SQCRs to conduct a face-to-face personal interview for all households. Under current regulations at 7 CFR 275.12(c)(1), only specific circumstances allow telephone interviews to be used in lieu of face-to-face interviews for the QC field investigation. Seventy percent of State agency responses to the RFI ^[37] question about the interview process included requests that telephone interviews be an acceptable interview method instead of face-to-face, citing reports of QC staff regularly traveling long distances only to find households failing to meet for the scheduled interview times, resulting in lost time and wasted administrative funds. During FNS's QCIRs, SQCRs raised concerns about staff safety while conducting the required face-to-face interviews outside of the local office. Finally, comments on the RFI from non-profit organizations presented that face-to-face interviews may be more challenging for SNAP's working households to comply with.

In considering the various comments provided by different stakeholders, the Department determined that switching to telephone interviews as the primary interview method will continue to meet the needs of the proposed QC review changes in this rule. As a result, the Department proposes to revise the QC personal interview regulation at 7 CFR 275.12(c)(1) to require that telephone interviews be the default interview format, and require that State agencies inform households that a face-to-face interview is an option available to them by request. This change will not only address the concerns brought forth by commenters but may also have a positive impact on State agency completion rates since the increased flexibility that telephone interviewing allows will provide more households the opportunity to comply with the QC review process. The reason the Department proposes standardizing the telephonic interview as the default interview mode as opposed to making it another interview option is to ensure fair treatment for all sampled households during the QC review process. In addition, because of this change, the Department proposes to eliminate the exception at 7 CFR 275.12(c)(1) available to remote, isolated households in Alaska because this proposed change will cause the exception to be unnecessary.

Regulations at 7 CFR 275.12(c)(1)(iii) require that, during the personal interview, reviewers must review with the household all documentary evidence in the household's possession and secure information about collateral sources of verification. The Department proposes to amend the provisions at 7 CFR 275.12(c)(1)(iii) to codify that during the personal interview, reviewers would also be required to review with the household all documentary evidence that is already in the case file and request new documentary evidence from the household, as needed. This is a best practice of State agencies and, by codifying this requirement, FNS seeks to ensure greater consistency in the interview process as well as accountability for the quality and scope of interviews, a vital component in the QC review process.

7 CFR 275.12(c)(1)–(2)—Review of Active Cases—Collateral Contacts

During, prior to, or after the personal interview, the SQCR may need to contact a collateral contact of the household who can be used as a source to verify household circumstances. Because State law might require an SQCR to obtain consent from the head of household to contact collateral contacts FNS codified the provision at 7 CFR 275.12(c)(1)(iv), which currently reads, “If required by the State, the reviewer shall obtain consent from the head of the household to secure collateral information.” However, FNS has since discovered the language of 7 CFR 275.12(c)(1)(iv) is being interpreted differently by States. For example, some States without such laws are still mandating consent be obtained, which has created roadblocks for SQCRs and resulted in preventing the SQCR from being able to complete cases. It is imperative that SQCRs accurately complete as many QC cases as possible without any unnecessary burdens. Therefore, the Department proposes to amend the language at 7 CFR 275.12(c)(1)(iv) to clarify its intent by linking obtaining consent to the presence of a State law.

Currently, regulations at 7 CFR 275.12(c)(2) require SQCRs to obtain verification from collateral contacts in all instances when adequate documentation is not available from the household. The current regulatory provision does not address situations when there is inadequate documentation in the case file. Therefore, the Department proposes to amend 7 CFR 275.12(c)(2) to require the reviewer to obtain verification from collateral contacts in all instances when adequate documentation is not available.

7 CFR 275.12(d)—Review of Active Cases—Variance Identification

According to Section 2(b) of the PIIA,^[38] an improper payment includes “any payment that should not have been made or that was made in an incorrect amount, under a statutory, contractual, administrative, or other legally applicable requirement.” Per Section 2(c), when an agency cannot discern whether a payment was proper because of lacking or insufficient documentation, the payment shall be treated as an improper payment. The Department proposes to amend and reorganize 7 CFR 275.12(d), as discussed in the three sections that follow, to reflect the new review focus and align it with requirements in PIIA. In addition, consistent with what was noted earlier, all references under 7 CFR 275.12(d) regarding the “sample month” would either be changed to “action under review” or eliminated, where applicable.

7 CFR 275.12(d)(2)—Review of Active Cases—Variances Excluded From Error Analysis

Regulations at 7 CFR 275.12(d) define a `variance' as a review finding that policy was applied incorrectly or that information verified as of the review date differs from the information used at the most recent certification action. The Department is also proposing to allow fewer opportunities for variance exclusion from the error analysis. There are two reasons for this proposal: (1) a more restrictive review focus in accordance with PIIA specifications on which errors are and are not considered payment errors; and (2) the change in review focus to reviewing only the household's eligibility determination versus a point in time analysis. As such, the Department proposes to eliminate regulations at 7 CFR 275.12(d)(2)(i), (iii), and (ix). These provisions are related to variances associated with countable income unable to be verified at the time of certification, changes in household circumstances that were not yet required to be reported as of the review date, and changes to child support orders that occurred after the most recent certification action, respectively, and no longer align with the new review focus. Provisions (ii) and (viii) regarding variances resulting from postponed verification for expedited service households and from incorrect written Federal policy, respectively, would be revised, as further discussed below, and references to Immigration and Naturalization Services (INS) would be updated.

FNS found in its QCIRs that inadequate documentation in household eligibility case records presents challenges for reviewers to complete cases, particularly in expedited service cases where verification of certain elements of eligibility is postponed. In these cases, the QC reviewer must rely solely on the eligibility case record to determine what verification was postponed and if any variances in benefits resulting from the postponed verification qualify to be excluded from the error rate calculation, called a variance exclusion. In these cases, a lack of required documentation should result in forfeiture of the variance exclusion since the reviewer cannot validate if regulatory requirements regarding verification were followed accurately. Current regulations are not clear that a lack of documentation means the exclusion does not apply, and some State agencies have used this ambiguity to apply the variance exclusion inappropriately. Therefore, the Department proposes to amend the language at 7 CFR 275.12(d)(2)(ii) to clarify that, for this exclusion to apply, the case record must include clear documentation indicating which elements of verification were postponed. Otherwise, if an eligibility worker does not sufficiently document an element to indicate they properly postponed it, the exclusion would not apply and any variances arising from errors related to the element would then be included in the error determination process.

The Department proposes to amend the current regulatory language at 7 CFR 275.12(d)(2)(viii) [re-designated as 7 CFR 275.12(d)(2)(vii)] regarding policy memoranda for clarity. Instead of specifying categories of policy memoranda, as exists now, the Department proposes to generalize the concept to be all-inclusive of SNAP policy memoranda issued.

7 CFR 275.12(d)(3)—Review of Active Cases—Other Findings

The regulations at 7 CFR 275.12(d) define a variance as a review finding that policy was applied incorrectly or that information verified as of the review date differs from the information used at the most recent certification action while the regulations at 7 CFR 275.12(d)(1) and (2) further describe which variances are included in and excluded from the error analysis. Regulations at 7 CFR 275.12(d)(3) describe findings other than variances; however, there is currently no requirement to report these findings to a local office. When a SQCR does report a finding other than a variance, they do so through a notification comment that is sent along with the case's official QC results. The lack of a requirement to report these findings can lead SQCRs to bypass those issues, thereby reducing the local eligibility office's ability to determine whether a change in the case record is warranted following the QC review. This is important to address, because currently it is possible for a SQCR to learn information about a household's current circumstances that do not affect the sample month under review but could affect the household's ongoing SNAP allotment after the sample month. For example, a SQCR conducts an interview with a household in May for a case selected for the March sample month. During the interview, the SQCR finds that a household member recently started a new job that increased the household's earned income over the income limit for its household size but failed to report this information to the State agency within the required time frames. The new earnings are excludable from the review because they occurred after the sample month, but the household is required to report this change and it may impact the household's ongoing SNAP allotment in future months.

As a result, the Department is proposing to amend 7 CFR 275.12(d)(3) by adding that the State QC office would be required to notify the local office of these other findings in all cases, regardless of the error impact those deficiencies may have on the case. The Department believes this notification is essential to good management of the Program, because it provides information about inaccuracies in the case file to eligibility staff, enabling them to correct the issue, prevent future errors from occurring, and potentially provide improved customer service to households. The Department also proposes to remove the following examples from the same paragraph since the household would not have been eligible for benefits according to SNAP rules at 7 CFR 273, and therefore, any benefits issued to such households are considered improper payments under PIIA: ^[39] an overdue subsequent certification and no current application on file. Additionally, the Department proposes to remove the following examples from the same paragraph since the change in review focus would no longer exclude them from variances cited in the QC review: insufficient documentation and incorrect application of the verification requirements specified in part 273.

Currently, at 7 CFR 275.12(d)(3), State agencies have the discretion to act on findings other than error variances discovered during the review, which are pertinent to the SNAP household or the case record. Such findings may include, for example, the incorrect age of a household member that is unrelated to an element of eligibility and deficiencies in work registration procedural requirements, among others. The Department proposes to maintain this State agency discretion to act or not act on additional information discovered during the QC review.

7 CFR 275.12(f)—Review of Active Cases—Reporting of Review Findings

Consistent with the changes made at 7 CFR 275.12(d)(3), where SQCRs would be required to document all variances discovered in the review and not just those causing a reportable error, the Department proposes to revise regulations at 7 CFR 275.12(f)(1) and (f)(2) to require reviewers include all variances in their individual review findings reports for both eligibility and basis of issuance variances, respectively. Changes in both sections would also remove the reference to `sample month', as this would be consistent with the shift in review focus from a point in time to the eligibility action under review.

7 CFR 275.12(g)—Review of Active Cases—Disposition of Case Reviews

As noted, proposed updates to the active review would be more reflective of improper payment requirements for PIIA, which requires a more thorough reporting of errors. Currently, FNS “charges” State agencies a penalty for having a low QC completion rate. This penalty is part of the calculation that determines the State agency's PER. This method is unique to SNAP. Other programs that report for PIIA, follow a different process, whereby any case they are unable to complete for QC results in a total dollar error. In SNAP, reviewers drop cases for which they cannot verify an element of eligibility or basis for the benefit amount. For instance, if a case indicates the household received earned income, and the household is unable to provide verification or a collateral contact that can validate the information, the reviewer drops the case as incomplete.

To be consistent with PIIA, the Department proposes to amend 7 CFR 275.12(g)(1) so that, as with other Federal programs, all cases that cannot be completed, regardless of the reason, would result in the reporting of an error for the total allotment issued for the action under review. The Department requests public comment for considerations of circumstances it should consider in implementing this policy.

The Department is also seeking feedback on whether there should be a threshold applied to completion rates in the proposed error rate calculation methodology, similar to the current percent completion threshold which requires a penalty be applied to a State agency's error rate if it fails to complete at least 98 percent of its minimum QC case load.

Current regulations at 7 CFR 275.12(g)(1)(ii) list instances in which the household's unwillingness to cooperate in completing a QC review has the effect of a refusal to cooperate. Those instances include when the household fails to respond to mail sent Certified Mail-Return Receipt Requested; when the household fails to attend an agreed upon interview; and when the household does not return a signed release of information after agreeing to do so or after receiving one through Certified Mail-Return Receipt Requested. FNS received input from State agencies through various meetings with the American Public Human Services Association (APHSA) that the use of certified mail is prohibitively expensive and that delivery and service issues with the United States Postal Service (USPS) have presented challenges even when using Certified Mail-Return Receipt Requested.

APHSA members requested reviewers be provided an opportunity to use a process similar to the Request for Contact (RFC) process used for unclear information, as found in 7 CFR 273.12(c)(3)(i)(A). FNS determined the use of a modified RFC process; whereby households that fail to respond to the request are suspended from SNAP for one month, with opportunity to verify their circumstances during that time, prior to having their SNAP participation terminated; is appropriate for QC purposes. This new process would be referred to as a request for quality control contact or RFQCC. Allowing a RFQCC will help State agencies complete cases and reduce the number of incomplete cases that, under the new process, would count as total dollar errors. The availability of this process will ensure integrity in the Program by encouraging households to cooperate with the QC process. It will also protect access to the Program for those households that do cooperate as required, as current regulations do not provide an additional month for the household to cooperate before the State must pursue termination of their participation. In addition, commonly known issues with mail delivery necessitate other allowable processes for States to utilize in gaining household cooperation. For that reason, the Department proposes to revise 7 CFR 275.12(g)(1)(ii) by reordering the examples and adding the previously mentioned RFQCC process as an alternative way to respond to households that either refuse or fail to ever respond to communication from State QC to cooperate with the QC review. Similar to the RFC process outlined in 7 CFR 273.12(c)(3)(i)(B)(2), if the household fails to respond to the RFQCC, the reviewer will inform the State and the State will send a notice of adverse action that suspends the household for one month to allow the household an opportunity to cooperate with QC prior to termination. If the household does not cooperate with QC by the deadline provided in the notice of adverse action, the reviewer must notify the State agency of the household's refusal and the State must follow through with terminating the household as stated at 7 CFR 275.12(g)(1)(ii).

The proposed review process would also require an update to regulations at 7 CFR 275.12(g)(2) regarding active cases that are not subject to review. The Department proposes to eliminate the current provisions at 7 CFR 275.12(g)(2)(iv) and 7 CFR 275.12(g)(2)(ix) pertaining to households receiving restored benefits and households not receiving benefits in the sampled month, because the provisions would no longer be relevant to the way cases would be sampled. In addition, 7 CFR 275.12(g)(2)(x) would be removed because all cases in which the household is unable to be reached for the QC review would result in a total dollar error amount for the eligibility action under review, as mentioned above, and, therefore, no longer be considered a case not subject to review. As a result of the proposed changes, the Department would also reorganize the section to accommodate the removals and additions.

Subpart E—Corrective Action

7 CFR 275.16(b)(2)—Corrective Action Planning—Negative Cases

In 2012, the Department changed the negative case review process in the final rule titled, Supplemental Nutrition Assistance Program: Quality Control Provisions of Title IV of Public Law 107–171 (75 FR 33421),^[40] from a review of the accuracy of a denial, termination, or suspension to a process that incorporated a customer service review of those actions, including whether the State agency accurately informed the household of the reason for the action and whether the State agency's procedures were correct. Since that change, the requirement at 7 CFR 275.16(b)(2) for State agencies to implement a corrective action plan (CAP) whenever its CAPER rose above one percent has become impractical, as exceeding the one percent threshold became routine because of the new review procedures. State agencies have informed FNS of their concerns in various ways, including through conferences and advisory group meetings, since the first CAPER release in calendar year 2013. In the 2018 RFI,^[41] FNS requested feedback from commenters regarding the factors FNS should consider in revising the current CAP requirement for negative cases. FNS received ten comments about changes to CAP requirements and three suggested a threshold change for CAPs required on CAPERs. However, those three commenters overwhelmingly agreed the threshold should be increased. FNS based the current threshold on the previous negative case review process and now agrees that the threshold should be adjusted to better accommodate the process implemented in FY 2012.

Consequently, the Department is proposing that a State agency would be required to implement a CAP when its CAPER is higher than the national CAPER for a given fiscal year. To do so, the Department proposes to amend 7 CFR 275.16(b)(2) by removing the phrase “of 1 percent or more in negative cases” and adding in its place the phrase “above the national average CAPER.” However, because FNS has received only limited and mainly anecdotal feedback from State agencies since the changes, the Department is seeking additional comments through this rulemaking regarding when State agencies would be required to implement a CAP to improve their CAPER.

7 CFR 275.16(b)(4)—Corrective Action Planning—Incomplete Cases

The CAP requirement at 7 CFR 275.16(b)(4) for incomplete cases is another area that State agencies have questioned. A CAP is currently required whenever a State agency is unable to complete more than 5 percent of its annual QC caseload. The Department received comments to the 2018 RFI ^[42] on factors to consider in revising the current CAP requirement in this area. Ten commenters provided comments on factors to consider if revising the current CAP requirements. The commenters, which included eight State agencies, provided the following to consider: adjust the completion rate because a 95 percent completion rate is unreasonable and unattainable for many State agencies; utilize the Federal regression rate penalty only if State agencies are not completing cases in accordance with Federal rules and regulations; adjust acceptable levels of performance before corrective action occurs based on trends and current data; and initiate CAPs only when FQCRs demonstrates that incomplete cases were completed by FQCRs and SQCRs had the ability to complete the case in the same manner. The Department considered these comments but determined that, due to the proposed new handling of incomplete cases as complete dollar amount errors, the most appropriate revision to this threshold would be to eliminate the CAP requirement altogether for active cases.

Since this rule proposes to count the full allotment for incomplete cases as an error, the Department has concluded that the development of a separate CAP for active cases would be duplicative and unnecessary. This is due, in part, to the fact that an excess number of dropped cases would result in higher PERs and the Department has procedures, namely the liability and sanction process, to respond to high PERs.

For negative cases, current rules related to dispositioning a case as incomplete in the negative sample frame stipulate that State agencies may list as incomplete only those cases where the whole case file cannot be located. This largely stems from long-standing requirements that eligibility workers fully document their case files in sufficient detail to permit a reviewer to determine the reasonableness and accuracy of the determination. As a practical matter, the incidence of incomplete negative cases by State agencies is extremely low. For example, for FY 2019, there were no dropped cases in the negative sample frame.

For these reasons, the Department is proposing to remove the provision at 7 CFR 275.16(b)(4). As a result, current regulations at 7 CFR 275.16(b)(5) will be redesignated as 7 CFR 275.16(b)(4).

Subpart F—Responsibilities for Reporting on Program Performance

7 CFR 275.21(b)(1)—Quality Control Review Reports—Mandating the Use of SNAP QC System

In 2013, FNS successfully implemented a web-based electronic QC review system (SNAP–QCS) for State agencies to input their QC case review data, including the documentary evidence to support case reviews. Therefore, the Department is proposing to require State agencies to use the Federally funded SNAP–QCS.

The primary purpose of SNAP–QCS is to provide a central location to house QC review cases which can be accessed at both the State and Federal levels. SNAP–QCS provides complete audit and status tracking for each case. All changes, beyond drafts, are available (where authorized) for comparison to prior iterations. SNAP–QCS contains edit check rules that prevent the submission by any user of information that violates business logic or other policy/regulatory guidance. In all cases, SNAP–QCS informs users of the nature of the rule violation in a manner that allows for correction of any such violation.

Since its inception, SNAP–QCS has replaced numerous State agency legacy systems. The system provides the following functionality: the creation of worksheets and review schedules (FNS 380, FNS–245 and FNS–380–1) ^[43] by the States, workflow management for State review worksheets and schedules; the Federal subsampling of QC review cases, the creation of review notes and findings by FQCRs, detailed workflow management for the Federal review process; tracking of case-related information, reporting tools for workflow and case characteristics, and analysis tools for advanced QC finding reports.

FNS requested comments in the 2018 RFI ^[44] as to whether the Department should mandate SNAP–QCS for all QC Worksheets. Ten State agency respondents commented on such a mandate. Four State agencies expressed support for mandated use of SNAP–QCS because they currently use it and asserted that it would allow for improved consistency in the submittal of required information to FNS across States. One State agency indicated it had no concerns about the mandate because State agencies would retain the option to maintain their own internal automated QC system and upload the results into SNAP–QCS. However, four State agencies objected to the mandated use of SNAP–QCS. One State agency noted SNAP–QCS does not account for special budgeting rules, such as waivers, and other commenters expressed concerns related to State system variability resulting in system modifications, increased work, and inefficient use of State agency staff time ( e.g., double data entry).

The Department appreciates all comments submitted. Most State agency commenters who opposed the requirement to use SNAP–QCS did not have experience using the automated version of the system. The Department understands there is a learning curve when a new reviewer begins using SNAP–QCS. FNS anticipates that mandating the use of SNAP–QCS will improve data integrity and expects that the edit checks would assist in accurate data collection. Therefore, the Department proposes to mandate the use of SNAP–QCS at 7 CFR 275.21(b)(1). This includes using the Auto-FNS Form 380, FNS Form 380–1, and FNS Form 245, and upload all documentation necessary to understand the disposition and findings for each sampled case.

Subpart G—Program Performance

7 CFR 275.23(b)(2)—Determination of State Agency Program Performance—Determination of Payment Error Rates

To accurately apply the new sampling method to determine State agency program performance, the Department proposes to amend the language at 7 CFR 275.23(b)(2)(i)(A) to clarify what the quantities y₁, b₁, x₁, and X₁ in the PER determination formula are and that X₁ should be weighted (using W_i) to account for the differential selection probabilities of the new sampling design. The Department also proposes to amend the language at 7 CFR 275.23(b)(2)(i)(B) to clarify what the quantities y₂, b₂, x₂, and X₂ in the formula are and that X₂ should be weighted (using W_i) to account for the differential selection probabilities of the new sampling design. The Department also proposes to remove language about stratified sample designs covered by both provisions because it would no longer apply when using the proposed sampling method.

To avoid confusion with the notation introduced in the new paragraph 7 CFR 275.11(g), Active sample allocation, the Department proposes to change the notation of “u” to “Z” in paragraph 7 CFR 275.23(b)(2)(i)(C). In addition, the Department proposes to remove and rephrase language about the quantity “Z” to make its definition clearer.

The assignment of error rates by FNS, whether for active or negative cases, has occurred only infrequently in the past. Typically, FNS made assignments because of extraneous circumstances beyond a State agency's control, such as when a large disaster impedes its ability to complete cases. However, situations do arise in which FNS must assign rates to State agencies due to other reasons. FNS assigned multiple State agency error rates in FY 2017 due to data integrity issues and found that greater clarity in regulations at 7 CFR 275.23(b)(2)(ii) is necessary. First, the Department proposes to amend the regulation to reflect the name change of the error rates for negative cases and clarify that assigned rates could be used in the determination of liability status, subject to 7 CFR 275.23(d). Second, the Department proposes to amend the paragraph by adding a statement at the end of the section to indicate that under no circumstance, would an assigned rate be eligible for appeal unless the rate resulted in a liability amount, per 7 CFR 275.23(d). While these clarifications are consistent with current practice and law, the Department maintains that the additional language adds emphasis.

7 CFR 275.23(b)(2)(iii)—State Agency Error Rates—Completion Rate Penalty

The Department proposes to change the current penalties applied to State agencies that fail to complete 98 percent of their active QC caseload as described in 7 CFR 275.23(b)(2)(iii) by changing the application of the penalty to apply to any State agency that fails to sample the required minimum annual sample size, while also increasing the impact of the penalty. Since State agencies would report all active cases with a disposition of incomplete as a total dollar error for the sampled action under review, the Department believes there is a sufficient deterrent to past State agency practices of dispositioning cases as incomplete to avoid errors. Still, situations exist where a State agency may not sample the minimum sample size. For FY 2022, a total of seven States did not sample enough cases to meet the minimum regulatory requirement for active and negative sample frames. This could happen in situations beyond the control of the State agency, such as when a natural disaster impedes the State agency's ability to sample and complete its required QC reviews. This could also occur in situations that are within the State agency's control to prevent, such as when a State agency fails to sufficiently staff the QC unit and is subsequently unable to complete the required minimum sample size.

Specifically, the Department seeks input on the proposal to apply an adjustment penalty to a State agency's error rate when the State agency fails to sample enough cases to complete the minimum sample size in any given sample month, including the following: (1) if the changes proposed to the current penalty formula are sufficient and (2) whether FNS should distinguish in how it applies the penalty based on whether a State agency cannot sample its minimum sample size due to situations beyond its control.

7 CFR 275.24—High Performance Bonuses

Section 4013(d) of the 2018 Farm Bill removed the requirement for the Secretary to award performance bonus payments to State agencies and prohibited the Secretary from awarding performance bonuses in calendar year 2019 for FY 2018 performance. The 2018 Farm Bill also required the Department to establish performance criteria relating to actions taken to correct errors, reduce rates of error, improve eligibility determinations and any other indicators of effective administration determined by the Secretary.

To ensure SNAP performance bonuses are removed from regulations throughout 7 CFR 275.24, the Department proposes to amend current regulations at 7 CFR 275.24 with this rulemaking and replace that language with performance criteria, as directed by Congress pursuant to Section 4013(d)(3) of the 2018 Farm Bill. The performance criteria mirror current language at 7 CFR 275.24 but removes all references to bonuses and adjusts the language, grammar, and structure of the provision accordingly. The Department proposes these performance criteria because they mirror prior performance bonus criteria, which the Department believes, based on prior experience with performance bonuses, are informative measures of performance.

7 CFR 275.24(a)(3)—Performance Measures—Program Access Index

As one of the technical changes, the Department proposes to remove the fourth sentence in subparagraph 7 CFR 275.24(a)(3)(i) as it is no longer necessary to allow an exception to calculating the program access index rate for the State of California as the State converted its cash out program to allow households to receive SNAP. Therefore, the Department can use actual SNAP participant numbers for their calculation and the sentence is unnecessary.

Proposed Timeframe for Implementation of QC Sampling and Active Review Changes

The Department recognizes the significant sampling and active review changes will require lead time and technical assistance for State agencies to come into compliance for implementation. The Department is considering establishing an implementation date of two full fiscal years after publication of the final rule. For example, if a final rule were published in March of 2027, the Department would establish an implementation date of October of 2029. The Department seeks input on this implementation timeline.

To summarize, the simplification of the QC review is expected to benefit several stakeholders. Households would benefit from no longer having to provide verifications for multiple months in the certification period if their case is selected for QC review, SQCRs and FQCRs would have an equally robust but less complicated active case review to conduct, and the Department and Congress would benefit from both an improvement in the quality of information that comes out of the reviews and by being able to compare improper payments more effectively across government social safety net programs.

In addition, simplifying the QC review in the proposed manner would also allow State agencies to implement more timely corrective actions. Timely corrective actions should reduce the compounding impact of improper payments as State agencies would be able to correct systemic errors and implement policies or processes more immediately, thus improving customer service to recipients, State agencies, and the American taxpayer.

The Department acknowledges some of the most impactful changes for State agencies administering SNAP and households sampled for review include (1) the shift in review of the action that authorized the allotment; (2) the handling of cases that cannot be validated, and (3) the introduction of the RFQCC process. The Department asks for and welcomes comment on all provisions in this proposed rulemaking, and we also request comments on potential impacts, direct and indirect, of these changes on State agencies and SNAP households. The Department reiterates this is a proposed rule, and we will consider all comments provided before a final rule implements any changes to SNAP's QC system.

Procedural Matters

Executive Order 12866, 13563, and 14094

Executive Orders 12866, 13563, and 14094 direct agencies to assess all costs and benefits of available regulatory alternatives and, if regulation is necessary, to select regulatory approaches that maximize net benefits (including potential economic, environmental, public health and safety effects, distributive impacts, and equity). Under Executive Order 12866, as amended, OMB's Office of Information and Regulatory Affairs (OIRA) determines whether a regulatory action is significant and, therefore, subject to OMB review. This proposed rule has been determined to likely be significant under E.O. 12866, as amended, and is being reviewed by the Office of Management and Budget (OMB) in conformance with Executive Order 12866.

Executive Order 13563 emphasizes the importance of quantifying both costs and benefits, of reducing costs, of harmonizing rules, and of promoting flexibility. The analysis below outlines the impacts that the Department anticipates may result from this proposed rule, if finalized, and was prepared pursuant to the above-mentioned executive orders.

Summary of Total Cost Impacts

A regulatory impact analysis (RIA) must be prepared for rules which are determined to be significant under Section 3(f)(1) of E.O. 12866, as amended ($200 million or more in economic effects in any one year). USDA does not anticipate this proposed rule is likely to have an economic impact of $200 million or more in any one year, and therefore, does not meet the definition of significant under Section 3(f)(1) of E.O. 12866 as amended. This proposed rule will not impact SNAP participants' benefit levels. Overall household burden will increase due to the increase in case samples required by the new process, not by adding additional burden to individual households.

The Department expects several of the proposed rule provisions to impact State Administrative Expenses (SAE) or FNS administrative costs. The rule includes the following changes expected to have measurable impacts:

—Section 275.11(b)(1)(iii) modifies the active case sample size formulas for State reviewers and Federal re-reviewers, increasing the sample size by 30 percent, to maintain the current level of precision.

—Section 275.11(3)(1) changes the active case sampling frame from all households to those with an action, including initial certification, recertification, and the submission of monthly, quarterly, or periodic reports.

—Section 275.12(b) clarifies that reviewers will only be required to examine households' circumstances for the month that the action took place, eliminating the need for comparing households' circumstances at the month of action versus the month of the review. As a result of the changes to the active case reviews, the shift to only reviewing eligibility actions will result in collecting less information about the household since only the point of eligibility will be reviewed, rather than a random point in time during a household's certification, which requires the collection of a new month's circumstances, not one that's already in the case record.

—Section 275.11(g) specifies that the active sample is divided into five strata, allowing the sample to be representative of the SNAP caseload.

—Section 275.12(d)(3) requires SQCRs to notify local offices of all non-error causing variances found in the review.

—Section 275.21(b)(1) mandates that SQCRs use the SNAP–QC system for QC reporting, rather than using their State systems.

Below in Table E.1 is a summary of the combined impacts of these provisions on both State Agencies and the Federal Government. The Annual Baseline column shows the current annual costs for each row. The columns with FY headers are the difference between the annual baseline and new procedures resulting from this rule change in that specific fiscal year, with FY 2024 being the first implementation year and the first year in which the discount rate is applied. The Total column shows the sum of the five FY columns: the costs over the first five years of implementation. As noted previously, there are no anticipated impacts on SNAP allotments, but there are some expected costs for SNAP households as a result of the increased number of sampled households. Cost savings are anticipated after the implementation year, in FY 2026, due to expected reductions in the time needed for a caseworker to perform a QC review in under the new rule from 8.98 to 6.33 hours.

Table E.1—Summary of Total Cost Impacts **

Impacts on state administrative expense (cost in nominal FY 2024 $000s)	Baseline:	Annualized cost differences from pre-rule baseline
	FY 2024	FY 2025	FY 2026	FY 2027	FY 2028	FY 2029	Total
New Sampling Procedures	$11,370.3	$3,393.4	−$950.9	−$950.9	−$950.9	−$950.9	−$410.2
Training	0.0	974.2	0.0	0.0	0.0	0.0	974.2
Reporting all variances	0.0	0.0	411.5	411.5	411.5	411.5	1,646.0
Record keeping	59.8	0.0	17.9	17.9	17.9	17.9	71.6
Help Desk **	70.0	0.0	210.0	210.0	210.0	210.0	840.0
Require all States use QCS **	175.0	0.0	525.0	525.0	525.0	525.0	22,100.0
Total	11,675.1	4,367.6	213.5	213.5	213.5	213.5	55,221.6
Discounted Cost Impact
7 percent	11,675.1	4,081.9	186.5	174.3	162.9	152.2	4,757.7
3 percent	11,675.1	4,240.4	201.2	195.4	189.7	184.2	5,010.9
Impacts on Household Burden (cost in nominal 000s)
Increase in Household Burden for Newly Sampled Cases	0.0	49.5	49.5	49.5	49.5	49.5	247.5
Discounted Cost Impact
7 percent	0.0	46.3	43.2	40.4	37.8	35.3	203.0
3 percent	0.0	48.1	46.7	45.3	44.0	42.7	226.7
* Totals may not sum due to rounding.
** These costs are incurred only by the Federal Government.

The net present value of costs to State and Federal governments over five years is $4,76 million at a seven percent discount rate and $5,01 million at a three percent discount rate. The net present value of costs to SNAP participants over five years is $203.0 thousand at a seven percent discount rate and $226.7 thousand at a three percent discount rate. Annualized in perpetuity, the government costs are $333.04 ($150.33) thousand per year at a seven (three) percent discount rate, and participant costs are $14.21 ($6.80) thousand per year at a seven (three) percent discount rate.

This rule change will result in substantive benefits for QC reviewers. By changing the active case review to focus on eligibility determinations (certifications, recertifications, and submission of required reports), the rule will result in the SNAP QC process being better aligned with how other Federal programs measure payment accuracy and with PIIA. The proposed changes are also expected to simplify QC reviews, which in turn will improve the quality of the reviews and of the information collected. The time spent by each caseworker in QC review is anticipated to drop from 8.98 to 6.33 hours per case, which will result in −$950.9 thousands in savings per year, even accounting for the expected increase in 13,649 QC reviews per year (see Table E.3). Simplifying the process and focusing on eligibility actions will also allow for more timely corrective actions.

Implementation Costs (State and Federal Administrative Expense)

Developing New Sampling Plans. Statisticians from each of the 53 State Agencies will need to develop new sampling plans for active cases. It will take an estimated 40 hours for each State Agency to develop a plan. Given an hourly rate of $47.81 (the median hourly wage for a statistician, according to the Bureau of Labor Statistics (BLS)), the total one-time cost is $101,400 (Table E.2). Because these are State Administrative Expenses for which the Federal Government reimburses States, $50,700 will be borne by the Federal Government and $50,700 will be borne by State Agencies.

Training SQCRs on new review procedures. SQCRs will need to be trained in the changes to the active case QC review process. There are approximately 634 SQCRs. Training will take an estimated 40 hours. At an hourly rate of $27.83 (the median hourly rate for a social worker, according to the BLS at time of analysis), the total cost for training will be $705,800. Because these are State Administrative Expenses for which the Federal Government reimburses States, $352,900 will be borne by the Federal Government and $352,900 will be borne by State Agencies.

Training SQCRs on entering data into the SNAP QC system. All SQCRs will be required to enter QC review data into the SNAP QC system. Currently, 263 SQCRs of an estimated 634 SQCRs nationwide are entering QC data into the SNAP QC system. Therefore, based on this estimate, an additional 371 SQCRs will need to be trained on how to enter data into SNAP–QCS by making the system mandatory. This training will take an estimated 26 hours. At an average social worker rate of $27.83, the cost for this training will be $268,400. Because these are State Administrative Expenses for which the Federal Government reimburses States, $134,200 will be borne by the Federal Government and $134,200 will be borne by State Agencies.

Longer time initially processing cases. During the first year, reviewers are expected to take an estimated 2 hours longer to review each case as they become accustomed to the changes in QC. With an estimated 59,149 cases, at $27.83 an hour, the one-year cost is $3,292,100. Because these are State Administrative Expenses for which the Federal Government reimburses States, $1,646,100 will be borne by the Federal Government and $1,646,100 will be borne by State Agencies.

These one-time implementation costs will total an estimated $4,367,700, of which $2,183,900 will be borne by the Federal Government and $2,183,900 borne by State Agencies.

Table E.2—Implementation Costs (State and Federal Administrative Expense) *

	Develop sampling plan	Train SQCRs on new review procedures	Train QC revisers to use QC system	Additional time for SQCRs to process active cases	Total implementation cost
Number Impacted	53.0	634.0	371.0	59,146.1
Time (hours)	40.0	40.0	26.0	2.0
Hourly rate	$47.8	$27.8	$27.8	$27.8
Total Cost ($000)	$101.4	$705.8	$268.4	$3,292.1	$4,367.7
Federal Share ($000)	$50.7	$352.9	$134.2	$1,646.1	$2,183.9
State Share ($000)	$50.7	$352.9	$134.2	$1,646.1	$2,183.9
* Totals may not sum due to rounding.

Annual Operational Costs and Cost Savings (State Administrative Expense)

Impact of changes in sample size and review procedures. Currently, SQCRs review 45,497 active cases per year.^[45] It takes an average of 8.908908 hours per review. The median hourly rate for a social worker is $27.83. Thus, the cost of performing QC reviews is currently an estimated $11,370,300. (Table E.3)) Under the proposal, SQCRs will review about 59,146 cases per year, an increase of 30 percent. However, after the first year, the average time per review is expected to drop from 8.908908 hours to 6.33 hours. This is because reviewers will only be required to examine households' circumstances for the month that the action took place, eliminating the need for comparing households' circumstances at the month of action versus the month of the review. The cost of performing QC reviews is estimated to be $10,419,400, a savings of $950,900. Because these are State Administrative Expenses the savings will be equally split between the Federal Government and State Agencies, with each saving $475,450.

New requirement to report non-error causing variances. Currently, States are not required to report non-error causing variances back to local SNAP offices. Section 275.12(d)(3) will require SQCRs to notify the local office of deficiencies in all cases, regardless of the error impact on the case. Little is known about how many variances of these kinds are currently reported, how many non-error causing variances will be reported under the requirement, or how long it takes to report such variances. Some cases will have no non-error causing variances, whereas others may have multiples of these variances. FNS is assuming that the additional time will average 15 minutes per case. This average is not meant to assume that all cases will have variances to report; rather, it is an average to balance cases with many variances to report against cases that have no additional variances to report. Using the hourly rate of $27.83 for a social worker times 59,146 cases (this is equal to 30 percent more than the FY 2017 actives QC case load, as required by the proposed sampling methodology) times 15 minutes per case yields a cost estimate of about $411,500, of which nearly $205,750 will be borne by the Federal Government and $205,750 will be borne by State Agencies.

Record keeping for more cases. States are required to keep records of all State QC reviews. It is estimated that the record keeping takes about 3 minutes per case, and that will be unchanged under the proposed system. However, because the number of cases is increasing by 13,649, using the average hourly rate of $27.83 for a social worker, the cost will increase by an estimated $17,900. This cost will be split between the Federal Government ($8,950) and State Agencies ($8,950).

New requirement to attest the validity of the sampling plan. State Agencies will now be required to provide an annual statement attesting to the validity of the sampling plan. The time to provide the statement is estimated to be 12 minutes. Each of the 53 State Agencies will have a SNAP manager provide this statement. According to the Bureau of Labor and Statistics, the median rate for a general manager is $55.41 at time of analysis. Thus, the cost is $600 per year, $300 borne by the Federal Government and $300 borne by State Agencies.

Increased QC help desk support. FNS will be required to provide increased help desk support. Currently, FNS contracts for one staff member to provide this support, at an annual cost of $70,000. FNS anticipates needing to contract for three additional people, for an additional cost of $210,000. This cost will be entirely borne by the Federal Government.

Additional QC data storage. FNS currently stores the SNAP QC data on six servers, at an annual cost of $27,333.33 per server, or $164,000 total. In addition, the storage costs are currently $11,000 per year. With all States using the SNAP QC data system and the increased sample size, FNS anticipates that these costs will quadruple, to $656,000 for the servers and $44,000 for storage, for a total of $700,000. The additional cost will total $492,000 for the servers and $33,000 for storage, for a total of $525,000. This cost will be entirely borne by the Federal Government.

The total annual administrative cost of the changes to the QC review process is estimated to be $214,100. Because some costs, such as help desk support and additional data storage, are not shared with State Agencies, the estimated cost to the Federal Government is $474,600. State Agencies are expected to save $260,500 annually due to the reduced case processing time.

Annual Household Administrative Burden

As discussed previously, the changes to the case sampling procedures will result in an overall increase in the number of cases sampled each year, from 45,497 annually to 59,146 (an increase of 13,649 cases). As described in the Paperwork Reduction Act section of this proposed rule, the burden to an individual household selected for review is not expected to change and will remain about 30 minutes per household. However, since the number of cases selected will increase, overall household burden will also increase. The increase is expected to cost $49,477.6 annually (13,649 cases × .5 hours per case × minimum wage of $7.25).

Uncertainties

While this proposed rule is expected to improve SNAP program integrity, it is unclear whether the rule provisions will result in additional sanctions or liabilities being imposed on State Agencies as a result of these proposed changes.

Table E.3—Annual Operating Costs and Savings (Changes From Current Procedures) *

Annual variable costs	Current units	Proposed units	Difference	Current hours	Proposed hours	Difference ($)	Current hourly rate ($)	Proposed hourly rate ($)	Difference ($)	Current cost ($000's)	Proposed cost ($000's)	Difference ($000's)
Processing Active Cases	45,497.00	59,146.10	13,649.10	8.98	6.33	−2.65	27.83	27.83	0.00	11,370.30	10,419.40	−950.90
Reporting all non-error causing Variances	0.00	59,146.00	59,146.00	0.25	0.25	0.00	27.83	27.83	0.00	0.00	411.50	411.50
Record Keeping	45,497.00	59,146.10	13,649.10	0.05	0.05	0.00	27.83	27.83	0.00	59.80	77.70	17.90
Attest Sample Plan	53.00	53.00	0.00	0.00	0.20	0.20	55.41	55.41	0.00	0.00	0.60	0.60
Annual Fixed Costs							Current unit cost ($)	Proposed unit cost ($)
Help Desk Support	1.00	4.00	3.00	N/A	N/A	N/A	70,000.00	70,000.00	0.00	70.00	280.00	210.00
Servers	6.00	24.00	18.00	N/A	N/A	N/A	27,333.33	27,333.33	0.00	164.00	656.00	492.00
Storage	1.00	4.00	3.00	N/A	N/A	N/A	11,000.00	11,000.00	0.00	11.00	44.00	33.00
Total Annual Costs										11,675.10	11,889.20	214.10
Federal Share										5,95960.05	6,434.60	474.55
State Share										5,75715.05	5,454.60	−260.45
* Totals may not sum due to rounding.

Regulatory Flexibility Act

The Regulatory Flexibility Act (5 U.S.C. 601–612) requires Agencies to analyze the impact of rulemaking on small entities and consider alternatives that would minimize any significant impacts on a substantial number of small entities. The entities impacted by this rule are State SNAP agencies that conduct QC reviews, which are not considered small entities for purposes of this analysis.

Pursuant to our review, the Department certifies that this rule will not have a significant impact on a substantial number of small entities.

Unfunded Mandates Reform Act

Title II of the Unfunded Mandates Reform Act of 1995 (UMRA), Public Law 104–4, establishes requirements for Federal agencies to assess the effects of their regulatory actions on State, local and Tribal governments, and the private sector. Under section 202 of the UMRA, the Department generally must prepare a written statement, including a cost benefit analysis, for proposed and final rules with “Federal mandates” that may result in expenditures by State, local or Tribal governments, in the aggregate, or the private sector, of $100 million or more in any one year. When such a statement is needed for a rule, Section 205 of the UMRA generally requires the Department to identify and consider a reasonable number of regulatory alternatives and adopt the most cost effective or least burdensome alternative that achieves the objectives of the rule.

This rule does not contain Federal mandates (under the regulatory provisions of Title II of the UMRA) for State, local and Tribal governments, or the private sector of $100 million or more in any one year. Thus, the rule is not subject to the requirements of sections 202 and 205 of the UMRA.

Executive Order 12372

SNAP is listed in the Catalog of Federal Domestic Assistance under Number 10.551. For the reasons set forth in the Final Rule codified in 7 CFR part 3015, subpart V and the related Notice (48 FR 29115), this Program is excluded from the scope of Executive Order 12372, which requires intergovernmental consultation with State and local officials.

Federalism Summary Impact Statement

Executive Order 13132 requires Federal agencies to consider the impact of their regulatory actions on State and local governments. Where such actions have federalism implications, agencies are directed to provide a statement for inclusion in the preamble to the regulations describing the agency's considerations in terms of the three categories called for under Section (6)(b)(2)(B) of Executive Order 13121.

The Department has considered the impact of this proposed rule, with comment, on State and local governments and has determined that this rule does not have federalism implications. Therefore, under Section 6(b) of the Executive Order, a federalism summary is not required.

Executive Order 12988, Civil Justice Reform

This proposed rule has been reviewed under Executive Order 12988, Civil Justice Reform. This rule is intended to have preemptive effect with respect to any State or local laws, regulations, or policies which conflict with its provisions or which would otherwise impede its full and timely implementation. This rule is not intended to have retroactive effect unless so specified in the Effective Dates section of the final rule. Prior to any judicial challenge to the provisions of the final rule, all applicable administrative procedures must be exhausted.

Civil Rights Impact Analysis

FNS has reviewed the proposed rule, Provisions to Improve the Supplemental Nutrition Assistance Program's QC System, in accordance with the Department Regulation 4300–004, Civil Rights Impact Analysis to identify and address any major civil rights impacts the proposed rule may have on participants on the basis of race, color, national origin, sex, and disability. A comprehensive Civil Rights Impact Analysis (CRIA) was conducted on the proposed rule, including an analysis of data and provisions contained in the proposed rule. The CRIA outlines outreach and mitigation strategies to lessen any possible civil rights impacts. The CRIA concludes the provisions of the proposed rule will impact the statistical design and active case review process, as well as clarify and update current regulations. The proposed rule would result in more SNAP households being selected for QC review in the active frame. The demographic profile of SNAP participants includes minorities, persons with disabilities, and the elderly; thus, program participants in these groups may be selected for QC review in the active frame. Additionally, the proposed rule will require State agencies to revise their review procedures, possibly resulting in less onerous reviews for a larger number of cases. The Department finds that the implementation of mitigation strategies and monitoring by the FNS Civil Rights Division and FNS SNAP may lessen these impacts. If necessary, the FNS Civil Rights Division will propose further mitigation and outreach strategies to alleviate impacts that may result from the implementation of the proposed rule.

Executive Order 13175

Executive Order 13175 requires Federal agencies to consult and coordinate with Tribes on a government-to-government basis on policies that have Tribal implications, including regulations, legislative comments or proposed legislation, and other policy statements or actions that have substantial direct effects on one or more Indian Tribes, on the relationship between the Federal Government and Indian Tribes, or on the distribution of power and responsibilities between the Federal Government and Indian Tribes. This regulation has possible Tribal implications, so consultation is required. FNS attended a Tribal consultation meeting on May 1, 2019, in Washington, DC and virtually to a Nevada meeting on December 6, 2022, where the changes to this rule were explained. No questions or concerns were brought to FNS's attention about this rule by any members of either meeting. If further consultation is requested, the Office of Tribal Relations will work with FNS to ensure quality consultation is provided.

Paperwork Reduction Act

The Paperwork Reduction Act of 1995 (44 U.S.C. Chap. 35; see 5 CFR 1320), requires that the Office of Management and Budget (OMB) approve all collections of information by a Federal agency from the public before they can be implemented. Respondents are not required to respond to any collection of information unless it displays a current, valid OMB control number. We are seeking a new OMB Control Number for these new, existing, and changing provisions in this rule and, once OMB approves the information collection request burden associated with this rulemaking, we will submit a request to merge the burden hours into their respective OMB Control Numbers. Once the merge is approved the newly assigned OMB control number can be discontinued. The current burden inventories for this collection are found in OMB–0584–0074, Expiration Date: 07/30/2025; 0584–0299, Expiration Date: 07/31/2023, and 0584–0303, Expiration Date: 1/31/2024. These changes are contingent upon OMB approval under the Paperwork Reduction Act of 1995. Additionally, when the information collection requirements have been approved, FNS will publish a separate action in the Federal Register announcing OMB's approval.

Comments on the information collection in this proposed rule must be received by November 20, 2023.

Send comments to the Office of Information and Regulatory Affairs, OMB, Attention: Desk Officer for FNS, Washington, DC 20503. Please also send a copy of your comments to John McCleskey, Branch Chief, Quality Control Branch, Program Administration and Nutrition Division, 1320 Braddock Place, 5th Floor; Alexandria, Virginia 22314. For further information, or for copies of the information collection requirements, please contact John McCleskey at the address indicated above. Comments are invited on: (1) whether the proposed collection of information is necessary for the proper performance of the Agency's functions, including whether the information will have practical utility; (2) the accuracy of the Agency's estimate of the proposed information collection burden, including the validity of the methodology and assumptions used; (3) ways to enhance the quality, utility and clarity of the information to be collected; and (4) ways to minimize the burden of the collection of information on those who are to respond, including use of appropriate automated, electronic, mechanical, or other technological collection techniques or other forms of information technology.

All responses to this request for comments will be summarized and included in the request for OMB approval. All comments will also become a matter of public record.

Title: Provisions to Improve the Supplemental Nutrition Assistance Program's Quality Control System.

OMB Number: 0584–NEW.

Expiration Date: Not Yet Determined.

Type of Request: New collection.

Abstract: Section 16 of the Act provides the legislative basis for the operation of the QC system. Part 275, Subpart C, of SNAP regulations implements the legislative mandates found in section 16. Regulations at 7 CFR 275.1, 275.14(d) and 275.21(a) and (b)(1) provide the regulatory basis for the QC reporting requirements. Section 11(a) of the Act provides the legislative basis for the recordkeeping requirements. Existing SNAP regulations at 7 CFR 275.4 specifically address record retention requirements for QC including form FNS–380, FNS–380–1, and the sampling plans found in 7 CFR 275 of the regulations.

Component (1) Form FNS–380 [OMB Control Number: 0584–0074], is a SNAP worksheet used to determine eligibility and benefits for households selected for review in the QC sample of active SNAP cases. This form provides a systematic means of aiding the State's Quality Control Reviewer in analyzing the case record, planning and doing field investigation and gathering, comparing, analyzing and evaluating data. FNS estimates that while this rule will require thirty percent more cases (households) be reviewed for QC, the rule does not change the existing burden on households, and will effectively decrease the ongoing burden for 53 State agencies by 20,151 hours annually. In addition, in order to implement the changes of the rule in the first year, the rule will add 196,915.17 startup burden hours for State agencies. These startup hours include 40 hours of training for 263 State QC reviewers on just the new review procedures for active cases, 64 hours to train 371 reviewers on using both SNAPQCS's automated FNS 380 worksheet and the new review procedures for active cases, and FNS is including 2.75 hours of additional time for State agencies to complete the FNS 380 worksheet for each case review to properly review and document according to the new procedures for the active case review. The revised total ongoing burden associated with this rule for this component is 385,844.12 reporting and recordkeeping burden hours and the first year's total reporting and recordkeeping burden due to rulemaking including the startup hours is 196,915.17 hours, bringing the total burden in the first year of implementation of the rule to 582,759.29 reporting and recordkeeping hours for this component.

Component (2) FNS 380–1 [OMB Control Number: 0584–0299], is SNAP's QC Review Schedule which collects QC and household characteristics data. The information needed to complete this form is obtained from the SNAP case record and State quality control findings. The information is used to monitor and reduce errors, develop policy strategies, and analyze household characteristic data. FNS estimates this rule will require 53 State agencies a revised total of 63,853.892 ongoing reporting and recordkeeping burden hours annually for this component of this collection. This is an increase of 14,735.33 burden hours. An estimated total of 742 additional startup hours are necessary for the first year's implementation of components within this collection for this rule. This startup includes 2 hours for 371 State QC reviewers to be trained on how to use SNAPQCS's automated FNS 380–1 worksheet. The revised total reporting and recordkeeping ongoing burden hours for the first year 64,595.89 hours.

Component (3) In the 275 regulations [OMB Control Number: 0584–0303], each State agency is required to develop a QC sampling plan that demonstrates the integrity of its case selection procedures. The QC system is designed to measure each State agency's payment error rate based on a statistically valid sample of SNAP cases. A State agency's payment error rate represents the proportion of cases that were reported through a QC review as being ineligible, overissued and underissued as well as the proportion of SNAP allotments that were either overissued or underissued to SNAP households. The FNS 311 Handbook is used by State agencies as a reference tool for creating their sampling plans. The current ongoing reporting and recordkeeping burden for this collection is 2,829 hours. FNS estimates this rule will require 53 State agencies a revised total of 2,829 ongoing reporting and recordkeeping hours and an additional 2,120 startup reporting burden hours for this first year of implementation of components within this rule. These startup hours include 40 hours for 53 State agency statisticians to update their State's sampling procedures to comply with the new sampling requirements of this rule. The first year of implementation for this component of the rule will require a total of 4,949 reporting and recordkeeping burden hours.

In total, FNS estimates this rule will decrease the overall ongoing burden associated with these three collections by 5,416 reporting burden hours. The overall total burden for these three components associated with the rule will require 53 State agencies, a total of 422,951 ongoing reporting and recordkeeping burden hours and 59,146 households 29,573 ongoing reporting burden hours annually. FNS also estimates this rule will require State agencies an additional 199,777.17 startup burden hours to implement the changes in this rule. Therefore, a grand total of 652,302 reporting and recordkeeping burden hours are estimated for this first year of implementation for this rule. This rule affects the three components of the QC process mentioned above, the use of (1) FNS forms 380, (2) FNS 380–1 and (3) the creation of the State QC Sampling plan attestation. The average burden per response and the annual startup and ongoing burden hours are explained below and summarized in the charts that follow.

Ongoing Reporting and Recordkeeping Burden

FNS 380: A SNAP worksheet used to determine eligibility and allotment amounts for households selected for review in the QC sample of active SNAP cases.

Reporting Respondents for this Proposed Rule: 59,199 (59,146 Individuals/Households and 53 State, Local and Tribal Government).

Estimated reporting responses for this Rule: 473,172 responses (59,146 for Individuals/Households and 473,172 for State, Local and Tribal Government).

Estimated hours per reporting response: 6.58 hours (0.5 hours for Individuals/Households and 6.08 hours for State, Local and Tribal Government).

Estimated Responses per Respondent to report for this Rule: 7812.81 (7,811.81 responses per State agency and 1 response per Household).

Estimated Reporting hours for this Rule: 384,449 hours (354,876 hours for State agencies and 29,573 burden hours for Households).

Estimated Records to keep per respondent for this Rule: 59,146 records for State agencies and there is no recordkeeping burden imposed on Individuals/Households.

Estimated hours per recordkeeping response: 0.0236 hour.

Estimated Total Recordkeeping hours for this Rule: 1,396 hours.

Estimated Total Ongoing Annual Reporting and Recordkeeping Burden for this rule: 385,845 hours.

Rule impacting ongoing reporting and recordkeeping burden: −20,151 hours.

FNS 380–1: SNAP's QC Review Schedule which collects QC and household characteristics data.

Respondents for this Proposed Rule: 53 State, Local, and Tribal Government.

Estimated Responses for this Proposed Rule: 59,146 responses.

Estimated Responses per Respondent to report for this Proposed Rule: 1,115.96 responses.

Estimated hours to report each response: 1.056 hours.

Estimated Total Reporting burden for this Rule: 62,458 hours.

Estimated Records to keep for this Proposed Rule: 59,146 responses.

Estimated Number of Records to keep per respondent for this Rule: 1,115.96 records per respondent to keep.

Estimated hours per recordkeeping response: 0.0236 hour.

Estimated Total Recordkeeping burden for this Rule: 1,395.84 hours.

Estimated Annual reporting and recordkeeping Burden on Respondents for this Proposed Rule: 63,853.89 hours.

Rule impacting ongoing reporting and recordkeeping burden: +14,735.33 hours.

FNS 275 Rules: Sampling Plan, 3rd Party Contractors, Arbitration, Good Cause, New Investment

Estimated Respondents for this Proposed Rule: 53 State, Local, and Tribal Government.

Estimated Number of Responses per Respondent to report for this Proposed Rule: 129 responses.

Estimated hours to report for each response: 21.91 hours.

Estimated Total Reporting Burden for this rule: 2826 hours.

Estimated Records to keep for this Rule: 246 records.

Estimated Records to keep per respondent for this Rule: 4.64 records per respondent.

Estimated hours to keep records for each response: 0.118 hours.

Estimated Total Recordkeeping burden for this Rule: 2.7612 hours.

Estimated annual reporting and recordkeeping Burden on Respondents for this Proposed Rule: 545 hours.

Rule impacting ongoing reporting and recordkeeping burden: 0 hours.

Estimated Grand Total Reporting and Recordkeeping Ongoing burden for this rule: 2,828.75 hours.

Estimated Total Rule impacting ongoing reporting and recordkeeping burden: 0 hours.

Startup Reporting and Recordkeeping Burden

FNS 380:

Reporting Respondents for this Proposed Rule: 687 State, Local and Tribal Government (53 State agencies and 634 State QC reviewers).

Estimated Number of startup reporting responses for this Rule: 237,218 responses.

Estimated hours per reporting response:.83 hours.

Estimated Number of annual Reporting hours for this Rule: 196,915 hours.

Estimated Total Recordkeeping startup hours for this Rule: 0 hours.

Estimated Total startup year Burden for this Rule: 196,915 hours.

FNS 380–1:

Reporting Respondents for this Proposed Rule: 371 State QC Reviewers.

Estimated Number of startup reporting responses per respondent for this Rule: 1 response.

Estimated hours per reporting response: 2 hours.

Estimated Number of annual Reporting hours for this Rule: 742 hours for State reviewers.

Estimated Total startup year Reporting and Recordkeeping Burden for this Rule: 742 hours.

FNS 275:

Reporting Respondents for this Proposed Rule: 53 State, Local and Tribal Government.

Estimated Number of startup reporting responses for this Rule: 53 responses.

Estimated hours per reporting response: 40 hours.

Estimated Number of Responses per Respondent to report for this Rule: 1 response.

Estimated Number of annual Reporting hours for this Rule: 2,120 hours.

Estimated Total Recordkeeping startup hours for this Rule: 0 hours.

Estimated Total startup Burden for this component of this Rule: 2,120 hours.

Estimated Grand Total Reporting and Recordkeeping Startup burden for this rule: 452,524.40 hours.

Estimated Grand Total Reporting and Recordkeeping burden for first year for this rule: 652,302 hours.

Table A.1—Revised I/H Reporting Burden

Reporting burden for individuals/households FNS 380, OMB 0584–0074
Reg. section	Description of activity	Estimated number of respondents	Estimated responses per respondent	Revised total annual responses	Revised number of burden hours per response	Revised estimated total burden hours	Previous submission total hours	Difference due to program changes	Difference due to adjustments
275.12 (c)(1)	Personal Interviews—Individuals or Households	59,146.00	1	59,146.00	0.5	29,573	22,748		+6,825
Individuals & Households Grand Total Reporting Burden Hours		59,146.00		59,146.00		29,573			+6,825

Table A.2—Revised State agency (SA) Reporting and Recordkeeping Burden 0584–0074

Reg. section	Description of activity	Est number of respondents	Estimated responses per respondent	Revised total annual responses	Revised number of burden hours per response	Revised estimated total burden hours	Previous submission total hours	Difference due to program changes	Difference due to adjustments
Reporting Burden for State Agencies FNS 380, OMB 0584–0074
275.12 (b)	Household Case Record Review	53	1,115.96	59,146	2	118,292	136,490.37	−18,199
275.12 (c)	Field investigation	53	1,115.96	59,146	2	118,292	159,238.77	−40,947
275.12 (c)(1)	Personal interviews	53	1,115.96	59,146	0.5	29,573	22,748.40	6,825
275.12 (d)(1)	Variance identification	53	1,115.96	59,146	0.5	29,573	40,947.11	−11,374
275.12 (e)	Error analysis	53	1,115.96	59,146	0.5	29,573	22,748.40	6,825
275.12 (f)	Reporting of review findings	53	1,115.96	59,146	0.25	14,786	n/a	14,786
275.12 (d)(3)	Reporting all variances to Local offices	53	1,115.96	59,146	0.25	14,786	n/a	14,786
Sub Total Reporting Burden		53	7,811.72	414,021	6.08	354,875	382,173	−27,298	0
Recordkeeping Burden for State Agencies FNS 380, OMB 0584–0074
FNS 380 Recordkeeping
275.4	Record Retention	53	1,115.96	59,146	0.0236	1,396	1,073.73	322.113568
Grand Total Reporting & Recordkeeping Burden		53		473,167	0.75295	356,271	383,247	−26,976	0

Table A.3—State Agency (SA) Reporting Startup Burden 0584–0074

Reg. section	Description of activity	Estimated number of respondents	Estimated responses per respondent	Startup annual responses	Startup number of burden hours per response	Startup estimated total burden hours
Reporting Burden Hours for State Agencies FNS 380, OMB 0584–0074 STARTUP Hours First Year Only
275.12 (d)(1)	Variance identification	53	1,115.96	59,146	0.75	44,359.41
275.12 (e)	Error analysis	53	1,115.96	59,146	0.75	44,359.41
275.12 (f)	Reporting of review findings	53	1,115.96	59,146	0.5	29,572.94
275.12 (d)(3)	Reporting findings to Local offices	53	1,115.96	59,146	0.75	44,359.41
275.12	Training State Agency QC reviewers on New Actives Process only	263	1	263	40	10,520.00
275.12 and 275.21 (b)(1)	Train reviewers on New Actives Process AND train new SNAPQCS users (reviewers) how to use SNAPQCS for 380	371	1	371	64	23,744.00
Grand Total STARTUP Hours Reporting Burden Only		687	345.294789	237,218	106.75	196,915.17

Table A.4—State Agency (SA) Reporting Revised Burden 0584–0299

Reg. section	Description of activity	Estimated number of respondents	Estimated responses per respondent	Revised total annual responses	Revised number of burden hours per response	Revised estimated total annual burden hours	Previous submission total hours	Difference due to program changes	Difference due to adjustments
FNS 380–1 Reporting for State Agencies OMB Control Number 0584–0299
275.12(f)	Reporting of Review Findings	53	1,115.96	59,146	1.056	62,458.049	48,044.83	14,413.22
Grand Total Reporting Burden Hours Only		53	1,115.96	59146	1.056	62,458.049		14,413.22
Reg. section	Description of activity	Estimated number of respondents	Estimated responses per respondent	Annual responses	Ongoing number of burden hours per response	Number of annual burden hours per response	Previously submission total burden	Differences due to program changes	Differences due to adjustments
FNS 380–1 Recordkeeping Ongoing
275.4	Record Retention	53	1,115.96	59,146	0.0236	1395.842768	1,073.7292	322.1136
Grand Total Affected Public		53	2,231.92	118,292	1.08	63,853.892	49,118.56	14,735.33

Table A.5—State Agency (SA) Startup Burden 0584–0299

Reg. section	Description of activity	Estimated number of respondents	Estimated responses per respondent	Startup annual responses	Startup number of burden hours per response	Startup number of annual burden hours per response
FNS 380–1 Reporting for State Agencies OMB Control Number 0584–0299 STARTUP First Year Only
275.21 (b)(1)	Train new State agency reviewers how to use SNAPQCS for 380–1	371	1	371	2	742
Grand Total STARTUP Hours Reporting Burden Only		371	1	371	2	742

Table A.6—State Agency (SA) Revised Burden 0584–0303

Reg. section	Description of activity	Estimated number of respondents	Estimated responses per respondent	Revised total annual responses	Revised number of burden hours per response	Revised estimated total burden hours	Previous submission total hours	Difference due to program changes	Difference due to adjustments
275 Regs Reporting OMB 0584–0303
275.11(a)(1)–(a)(2)	Sampling Plan	53	1	53	20	1,060	1,060	0	0
275.2(c)(1)(i)	Use of 3rd Party Contractors—Notification of intent to hire	3	1	3	0.25	0.75	0.75	0	0
275.2(c)(1)(ii)	Use of 3rd Party Contractors—Submission of signed contract and tasks	3	1	3	0.5	1.5	1.5	0	0
275.2(c)(1)(iii)	Use of 3rd Party Contractors—Submission of completed deliverables	3	1	3	0.5	1.5	1.5	0	0
275.2(c)(1)(iv)	Use of 3rd Party Contractors—Notification of training sessions	3	1	3	0.08	0.24	0.24	0	0
275.2(c)(4)	Arbitration Process	12	3	36	34	1,224	1,224	0	0
273.23(f)	Good Cause Process	1	1	1	160	160	160	0	0
275.23(h)	New Investment Plan Template Form FNS 74 A	9	1	9	32	288	288	0	0
275.23(h)(4)	New Investment Progress Report Template Form FNS 74 B	9	2	18	5	90	90	0	0
Sub-Total Reporting Burden		53	2.433962264	129	21.90689922	2,825.99	2,825.99	0	0
275.4	Sampling Plan Record Retention	53	1	53	0.0236	1.2508	1.2508	0
275.4	Arbitration Process Record Retention	12	3	36	0.0236	0.8496	0.8496	0
275.4	Good Cause Process Record Retention	1	1	1	0.0236	0.0236	0.0236	0
275.4	New Investment Plan Template Form FNS 74 A Record Retention	9	1	9	0.0236	0.2124	0.2124	0
275.4	New Investment Progress Report Template Form FNS 74 B Record Retention	9	2	18	0.0236	0.4248	0.4248	0
Recordkeeping Total		53	2.20754717	117	0.0236	2.7612	2.7612	0

Table A.7—State Agency (SA) Startup Burden 0584–0303

Reg. section	Description of activity	Estimated number of respondents	Estimated responses per respondent	Revised total annual responses	Revised number of burden hours per response	Revised estimated total burden hours
275 Regs Reporting STARTUP OMB Control Number 0584–0303
275.11	Implement new sampling plan	53	1	53	40	2,120
Grand Total Reporting Burden		53		53		2,120

Table A.8—Summary of Grand Total Annual Reporting & Recordkeeping Burden Estimates

Form or citation	Description of activity (ongoing or start up)	Estimated number of respondents	Annual responses	Number of burden hours per response	Estimated total burden hours
380 SA	Ongoing	53	473,172	0.75295	353,272.00
380 I/H	Ongoing	59,146	59,146	0.5	29,573.00
380–1	Ongoing	53	118,292	1.0796	63,854.00
275	Ongoing	53	246	20	2,828
380 SA	Startup	53	236,584	2.75	162,651
380 SA Staff train on Review process	Startup	263	263	40	10,520.00
380 SA Staff train on Review process and SNAPQCS	Startup	371	371	64	23,744.00
380–1 SA Staff for SNAPQCS	Startup	371	371	2	742
275 Regulations	Startup	53	53	40	2,120
Grand Total Reporting and Recordkeeping Burden for Rule		59,886	888,498	98	649,304.00

Table A.8—Summary of Current and Rule Associated Burden

OMB control number 0074		Rule related revised ongoing	Rule related new startup
Estimated Total No. Respondents	45,550	13,649	687
Estimated Average No. Responses per Respondent	6.991833	0.97	345.294
Estimated Total Annual Responses	318,478.00	154,689.00	237,218
Estimated Average Hours per Response	0.00		0.83
Estimated Total Annual Burden Hours Requested	582,759	385,844	196,915.17
Current OMB Inventory	405,997	405,997
Difference Due to Rulemaking	176,762	(−20,152)	196,915.17
OMB control number 0299		Rule related revised ongoing	Rule related new startup
Estimated Total No. Respondents	53.00	53	371
Estimated Average No. Responses per Respondent	2,238.925	2,231.920	1
Estimated Total Annual Responses	118,663.000	118,291.760	371
Estimated Average Hours per Response	0.54654	0.54607	2
Estimated Total Annual Burden Hours Requested	64,854.000	64595.89205	742
Current OMB Inventory	49,119	49,119
Difference Due to Rulemaking	15,477	14,735	742
OMB control number 0303		Rule related revised ongoing	Rule related new startup
Estimated Total No. Respondents	53.00	53	53
Estimated Average No. Responses per Respondent	4.6420	4.642	1
Estimated Total Annual Responses	246	426	1
Estimated Average Hours per Response	11.4989	11.4989	40
Estimated Total Annual Burden Hours Requested	4,949	4,949	2,120
Current OMB Inventory	2,829	2,825	0
Difference Due to Rulemaking	0	10	2120

E-Government Act Compliance

The Department is committed to complying with the E-Government Act, 2002 to promote the use of the internet and other information technologies to provide increased opportunities for citizen access to Government information and services, and for other purposes.

List of Subjects

7 CFR Part 271

• Grant programs—social programs
• Reporting and recordkeeping requirements

7 CFR Part 275

• Grant programs—social programs
• Reporting and recordkeeping requirements

Accordingly, 7 CFR parts 271 and 275 are proposed to be amended as follows:

PART 271—GENERAL INFORMATION AND DEFINITIONS

1. The authority citation for part 271 continues to read as follows:

Authority: 7 U.S.C. 2011–2036.

2. In § 271.2:

a. Revise the definition of “active case”;

b. Remove the definition “active case error rate”;

c. Add in alphabetical order a definition for “Case and Procedural Error Rate (CAPER)”;

d. Revise the definitions of “error” and “negative case”;

e. Remove the definition of “negative case error rate”;

f. Revise the last sentence in the definition of “payment error rate”;

g. Revise the definitions “review date” and “sample month”;

h. Remove “INS” and add in its place “USCIS” in the definition of “Systematic Alien Verification for Entitlements (SAVE)”; and

i. Remove the definition “Underissuance error rate. (See Underpayment error rate.)”,

The revisions and additions read as follows:

§ 271.2

Definitions.

* * * * *

Active case means a case where households experienced an eligibility action during the sample month which resulted in an issuance of benefits. For purposes of this definition, an eligibility action refers to initial certification, recertification, or submission of a required monthly, quarterly, or periodic report in the sample month and a benefit allotment is issued in the following month.

* * * * *

Case and Procedural Error Rate (CAPER) means an estimate of the proportion of denied, suspended, or terminated cases where the household was incorrectly denied, suspended, or terminated or where procedural deficiencies exist. This estimate will be expressed as a percentage of completed negative quality control reviews.

* * * * *

Error for active cases results when a determination is made by a quality control reviewer that a household that experienced an eligibility action—as described in the definition of “active case” in this section—was ineligible, received an incorrect allotment, or was determined “incomplete” by the QC reviewer. Thus, errors in active cases involve dollar loss to either the participant (underissuance) or the government (overissuance). For negative cases, an “error” means that the reviewer determines that the decision or process to deny, suspend, or terminate a household was incorrect.

* * * * *

Negative case means a case where there was an action to deny, suspend, or terminate a household during the sample month.

* * * * *

Payment error rate * * * Each component error rate is the value of benefits either overissued or underissued expressed as a percentage of all allotments issued to completed active sample cases.

* * * * *

Review date for quality control active cases means the date an eligibility action was taken to authorize the allotment. The “review date” for quality control negative cases, depending on the characteristics of individual State systems, could be the date on which the eligibility worker makes the decision to suspend, deny, or terminate the case, the date on which the decision is entered into the eligibility system, or the date of the notice to the client. For no case is the “review date” the day the quality control review is conducted. State agencies must consistently apply the same definition for review date to all sampled cases of the same classification.

* * * * *

Sample month means the month of the sample frame from which a case is selected ( e.g., the January sample month, for active cases, shall be comprised of a selection of cases where the household was certified in January, recertified in January, or required to have submitted a monthly, quarterly, or periodic report in January resulting in an issuance of benefits in February. The January sample month for negative cases would be comprised of all cases that were denied, terminated, or suspended in January).

* * * * *

PART 275—PERFORMANCE REPORTING SYSTEM

3. The authority citation for part 275 continues to read as follows:

Authority: 7 U.S.C. 2011–2036.

4. In § 275.2, revise paragraph (b) and add a sentence to the end of paragraph (d) to read as follows:

§ 275.2

State agency responsibilities.

* * * * *

(b) Staffing standards. The State agency shall employ sufficient staff to perform all aspects of the Performance Reporting System as required in this part.

(1) The State agency shall ensure that the staff used to conduct QC reviews operate independently from those responsible for overseeing the eligibility determination process to ensure objective and accurate assessments of the Performance Reporting System.

(2) The staff used to conduct QC reviews shall not have prior knowledge of either the household or the decision under review. Where there is prior knowledge, the reviewer must disqualify themselves. To ensure no prior knowledge on the part of QC or ME reviewers, local project area staff shall not be used to conduct QC or ME reviews; exceptions to this requirement concerning local level staff may be granted with prior approval from FNS. However, local personnel shall not, under any circumstances, participate in ME reviews of their own project areas. Prior knowledge is defined as having:

(i) Taken any part in the eligibility determination that has been made in the case;

(ii) Discussed the case with staff who participated in the decision; or

(iii) Personal knowledge of or acquaintance with persons in the case itself.

(3) Nothing in this part shall preclude a State-level staff person to be used as a collateral contact for purposes of the QC review. Such contact must, however, be limited to those same rules governing all other collateral contacts, including privacy-related rules, found in 7 CFR 273.2(f)(4)(ii).

* * * * *

(d) * * * Non-compliance with this requirement, as determined by the Secretary, may result in the suspension or disallowance of Federal reimbursements for costs of the administration of SNAP for the system(s) found to be out of compliance, pursuant to 7 CFR 277.16.

5. In § 275.3:

a. Revise paragraph (d)(1)(i);

b. Revise the paragraph (d)(3) heading'

c. Remove “negative case error rate” and add in its place “case and procedural error rate” in paragraph (d)(3) introductory text;

d. Add a sentence at the end of paragraph (d)(4)(i)(A);

e. Remove the phrase “appropriate FNS regional office addressed to the attention of the FNS Arbitrator” and add in its place the phrase “FNS Arbitrator and copy the appropriate FNS regional office” in paragraph (d)(4)(iv) introductory text; and

f. Remove the word “may” and add in its place the word “must” in the second sentence of paragraph (d)(4)(iv) introductory text.

The revision and addition read as follows:

§ 275.3

Federal monitoring.

* * * * *

(d) * * *

(1) * * *

(i) FNS will select a subsample of a State agency's active cases, as follows:

(A) The Federal review sample for active cases is determined as follows:

Table 1 to Paragraph (d)(1)(i)(A)

Average monthly reviewable caseload (N)	Federal subsample target (n′)
60,000 and over	n′ = 400.
10,001 to 59,999	n′ = .005 N + 100.
10,000 and under	n′ = 150.

(B) In the above formula, n' is the minimum number of Federal review sample cases which must be selected when conducting a validation review, except that FNS may select a lower number of sample cases if:

( 1) The State agency does not report a change in sampling procedures associated with a revision in its required sample size within 10 days of effecting the change; or

( 2) The State agency does not complete the number of case reviews specified in its approved sampling plan.

(C) The reduction in the number of Federal cases selected will be equal to the number of cases that would have been selected had the Federal sampling interval been applied to the State agency's shortfall in its required sample size. This number may not be exact due to random starts and rounding.

(D) In the above formula, N is the State agency's minimum active case sample size as determined in accordance with § 275.11(b)(1).

(E) Once the minimum Federal subsample size n′ is determined, the Federal subsample must be proportionally allocated across the five strata, defined in § 275.11(g), to the State final weights to ensure there is no loss of precision due to differential sampling probabilities. The Federal subsample size for each stratum shall be determined as follows:

( 1) In the formulas in the table above, N is the sampling universe/monthly caseload; F is the sampling frame; n is the State sampling size; n' is the Federal subsample size; W is the State sampling weights; and W' is the Federal sampling weights.

( 2) For stratum i, the Federal subsample size n'_i shall be proportional to the final State weight for that stratum, W_i, in other words

This means that the sampling probability for cases in stratum i is W_i /W and the final weight for the cases selected for the Federal subsample from stratum i, W′_i, is given by:

for every i, which is a constant across the strata. This makes the Federal subsample self-weighting.

* * * * *

(3) Case and procedural error rate. * * *

* * * * *

(4) * * *

(i) * * *

(A) * * * No other types of disagreement are eligible for arbitration.

* * * * *

6. In § 275.11:

a. Revise paragraphs (a) introductory text, (a)(1) through (3), and (b)(1) and (2);

b. Remove paragraph (b)(4);

c. Revise paragraphs (c)(1) and (2), (e) introductory text, (e)(1), and (f)(1);

d. Redesignate paragraph (g) as paragraph (i); and

e. Add new paragraph (g) and paragraph (h).

The revisions and additions read as follows

§ 275.11

Sampling.

(a) Sampling plan. Each State agency shall develop a quality control sampling plan that is compliant with this section and demonstrates the integrity of its sampling procedures.

(1) Content. The sampling plan shall include a complete description of the frame, the method of sample selection, and methods for estimating characteristics of the population and their sampling errors that the State agency will apply when conducting its quality control procedures. The description of the sample frames shall include: source, availability, accuracy, completeness, components, location, form, frequency of updates, deletion of cases not subject to review, and structure. The description of the methods of sample selection shall include procedures for: estimating caseload size, addressing corrections, computation of sampling intervals and random starts (if any), stratification, identifying sample cases, correcting over-or under sampling, and monitoring sample selection and assignment. The State agency shall provide FNS with a schedule for completion of each step in the sampling procedures contained in this section.

(2) Criteria. All sampling plans shall:

(i) Conform to principles of probability sampling; and (ii) Select an overall quality control sample size in accordance with paragraphs (b) through (f) of this section.

(3) Design. Each State agency shall, over the course of the annual review period, implement a sample design each month for both active and negative case samples. For the active case sample selection, the State agency shall define the sampling frame as the monthly list of active cases as defined in 7 CFR 271.2. This list reflects a subset of all the active SNAP cases in a given month. This list will then be stratified for sample selection based on action type and the length of the certification or reporting period. Within each of these strata, each month, the State agency may select a systematic sample or use another method of random selection ( e.g., sorting the cases in a random order and selecting the first `m' number of cases required to meet the monthly sample target) of cases to equal the overall sample size required for the year, divided evenly by twelve.

* * * * *

(b) * * *

(1) Active cases. (i) All active cases shall be selected in accordance with procedures specified in this section, and the review findings shall be included in the calculation of the State agency's payment error rate. A State agency shall select a sample that is divided equally across 12 months. Sample size is specified as follows:

Table 1 to Paragraph (b)(1)(i)

Average monthly reviewable caseload (N)	Minimum annual sample size (n)
60,000 and over	n = 1,326.
12,942 to 59,999	n = 390 + [0.0199(N−12,941)].
Under 12,942	n = 390.

(ii) In the formulas in paragraph (b)(1)(i) of this section, n is the required active case sample size. This is the minimum number of active cases subject to review which must be selected during the annual review period. One-twelfth of this value shall be selected each month, rounded to the next whole number ( e.g., if the yearly sample size is 1,326 then 111 would be sampled monthly). In the same formulas, N is the average monthly participating caseload subject to quality control review ( i.e., cases which are included in the active universe defined in paragraph (e)(1) of this section) during the annual review period.

(2) Negative cases. (i) All negative cases shall be selected in accordance with procedures specified in this section, and the review findings shall be included in the calculation of the State agency's case and procedural error rate.

(ii) The minimum number of negative cases to be selected and reviewed by a State agency during each annual review period shall be determined as follows:

Table 2 to Paragraph (b)(2)(ii)

Average monthly reviewable negative caseload (N)	Minimum annual sample size (n)
5,000 and over	n = 680
684 to 4,999	n = 150 + [ 0.1224(N−683)]
Under 684	n = 150

(iii) In the formulas in paragraph (b)(2)(ii) of this section, n is the required negative sample size. This is the minimum number of negative cases subject to review which must be selected each review period.

(iv) In the formulas in paragraph (b)(2)(ii) of this section, N is the average monthly number of negative cases which are subject to quality control review ( i.e., cases which are part of the negative universe defined in paragraph (e)(2) of this section) during the annual review period.

* * * * *

(c) * * *

(1) Substitutions. Once a case has been identified for inclusion in the sample by a predesigned sampling procedure, substitutions are not acceptable. An active case must be reviewed each time it is selected for the sample. If a case is selected more than once for the negative sample as the result of separate and distinct instances of denial, suspension, or termination, it must be reviewed each time.

(2) Corrections. Under sampling must be corrected during the annual review period. Oversampling may be corrected at the State agency's option. Cases which are dropped to compensate for oversampling shall be reported as not subject to review. Because corrections must not bias the sample results, cases which are dropped to compensate for oversampling must comprise a random subsample of all cases selected (including those completed, not completed, and not subject to review). Cases which are added to the sample to compensate for under sampling must be randomly selected from the entire frame in accordance with the sample size, sample selection, sampling frame, and sample allocation procedures specified in paragraphs (b), (c) (e), and (g) of this section. All sample adjustments must be fully documented and available for review by FNS.

* * * * *

(e) Sample frame. The State agency shall select cases for quality control review from a sample frame. Complete coverage of the sample universes, as defined in paragraph (f) of this section, must be assured so that every case subject to quality control review has an equal or known chance of being selected in the sample. Since the SNAP quality control review process requires an active and negative sample, two corresponding sample frames are also required.

(1) Active cases. The sample frame shall consist of all active cases as defined in 7 CFR 271.2. Cases which did not experience any of these eligibility actions in the sample month shall be removed prior to sampling. State agencies must use a list of certified and recertified cases as well as the household's report due date for the following reporting systems: monthly, quarterly, and simplified.

* * * * *

(f) * * *

(1) Active cases. The universe for active cases shall include all households in which any of the following eligibility actions occurred in the sample month: initial certification, recertification, or a required monthly, quarterly, or periodic report was due during the sample month and a benefit allotment is issued in the following month. The following shall be excluded from the sampling frame:

(i) A household receiving Disaster-SNAP benefits under the authority of the Food and Nutrition Act of 2008, as amended, and the Robert T. Stafford Disaster Relief and Emergency Assistance Act;

(ii) A household which is under investigation for an intentional Program violation, including a household with a pending administrative disqualification hearing;

(iii) A household appealing an adverse action which was the result of an eligibility action taken during the sample month; or

(iv) Other households excluded from the active case universe during the review process are identified in 7 CFR 275.12(g).

* * * * *

(g) Active sample allocation. States shall stratify both the sample universe and the sampling frame specified in paragraphs (e)(1) and (f)(1) of this section according to the five strata described in paragraphs (g)(1)(i) through (v) of this section. If not all 5 strata exist, the State shall allocate the full sample across the existing strata according to the following guidelines:

(1) In table 3 to paragraph (g) introductory text, strata a through e have the following action types and reporting periods:

(i) Stratum a—Action type: initial certification or re-certification; Reporting period: less than 6 months;

(ii) Stratum b—Action type: redetermination based on a monthly, quarterly or periodic report; Reporting period: less than 6 months;

(iii) Stratum c—Action type: initial certification or recertification; Reporting period: 6 months;

(iv) Stratum d—Action type: redetermination based on a monthly, quarterly or periodic report; Reporting period: 6 months;

(v) Stratum e—Action type: all; Reporting period: more than 6 months.

(2) State agencies shall allocate 10 percent of the sample to stratum a and 10 percent of the sample to stratum b. The remaining 80 percent of the sample should be allocated in a manner proportionate to the size of the strata in the sample universe as described in table 3 to paragraph (g) introductory text.

(3) If a State agency does not have stratum a or b or both, it shall allocate the full sample size proportionately to the size of the existing strata in the sample universe.

(4) In the formulas in table 3 to paragraph (g) introductory text, N represents the sample universe ( i.e., the total number of eligible SNAP cases) and N_a (for example) represents the number of SNAP cases in the universe that belong to stratum a; F represents the total number of eligible SNAP cases in the sampling frame ( i.e., the list of eligible SNAP cases having one of the three actions in the month of selection) and F_a (for example) represents the number of SNAP cases in the frame that belong to stratum a; n represents the total sample size and n_a (for example) represents the number of SNAP cases selected from stratum a meeting the stratum requirements.

(5) Within each stratum the State agency shall select the designated number of cases at random or using a systematic method upon a random sort of the cases. If in any strata the number of cases to be sampled exceeds the actual number in the frame, the State shall select all of the cases in that stratum. For example, if the proposed sample size for stratum a is greater than F_a then the State agency shall take all N_a SNAP cases for stratum a, hence n_a = F_a, and allocate (n-n_a -n_b) to the last three strata proportionately to the size of these strata in the sample universe (N).

(h) Weighting. Given that the active stratified sampling design oversamples some strata and under-samples others, weighting is necessary. The weights for the active cases sample are defined as follows:

W_i = (N_i /F_i) × (F_i /n_i) = N_i /n_i  for i=a,b,c,d,e

State agencies are responsible for providing to FNS the counts N_i, F_i, and n_i (for i=a,b,c,d,e) as part of the sampling plan described at 7 CFR 275.11(a)(4).

* * * * *

7. In § 275.12:

a. Revise paragraphs (a) and (b);

b. Add a sentence at the end of paragraph (c) introductory text;

c. Revise paragraphs (c)(1) introductory text and (c)(1)(iii) ;

d. Remove the phrase “the State” and add in its place the phrase “State law” in the third sentence of paragraph (c)(1)(iv);

e. Add the phrase “or in the case file” at the end of the first sentence of paragraph (c)(2);

f. Remove the phrase “as of the review date” and add in its place the phrase “by the reviewer” and remove the phrase “at the most recent certification action)” and add in its place the phrase “in the eligibility action under review)” in paragraph (d) introductory text;

g. Remove the phrase “for the sample month” in paragraph (d)(2) introductory text;

h. Add a sentence after the second sentence and remove the last sentence of paragraph (d)(2)(ii);

i. Remove paragraphs (d)(2)(iii) and (ix);

j. Redesignate paragraphs (d)(2)(iv) through (viii) as paragraphs (d)(2)(iii) through (d)(2)(vii);

k. Remove the phrase “Immigration and Naturalization Service's (INS)” and add in its place the phrase “United States Citizenship and Immigration Services (USCIS)” in newly redesignated paragraph (d)(2)(v) introductory text;

l. Remove “INS” and add in its place “USCIS” in newly redesignated paragraphs (d)(2)(v)(A) introductory text, (d)(2)(v)(A)( 3) (two occurrences), (d)(2)(v)(B) introductory text, and (d)(2)(v)(B)( 2);

m. Remove the phrase “category three and four Policy Memoranda under the Policy Interpretation Response System” and add in its place the phrase, “FNS policy memoranda” in newly redesignated paragraph (d)(2)(vii);

n. Revise paragraph (d)(3);

o. Remove the phrase “in the sample month” and add in its place “as an error” in the first sentence of paragraph (f)(1) and revise the last sentence of paragraph (f)(1);

p. Remove the phrase “in the sample month” in the first sentence of paragraph (f)(2) and revise the last sentence of paragraph (f)(2);

q. Add a sentence to the end of paragraph (g)(1) introductory text;

r. Revise paragraphs (g)(1)(ii) and (g)(2)(i) and (ii);

s. Remove paragraphs (g)(2)(iv), (ix), and (x);

t. Redesignate paragraphs(g)(2)(v) through (viii) as paragraphs (g)(2)(iv) through (vii); and

u. Revise newly redesignated paragraphs (g)(2)(vi) and (vii).

The revisions and additions read as follows:

§ 275.12

Review of active cases.

(a) General. A sample of households that experienced an eligibility action— i.e., certified for SNAP, recertified for SNAP, or required to have submitted a monthly, quarterly, or periodic report in the sample month which resulted in an issuance of benefits in the following month, shall be selected for active case review. These active cases shall be reviewed to determine if the household was eligible and, if eligible, whether the household received the correct allotment. The determination of a household's eligibility shall be based on an examination and verification of all elements of eligibility ( i.e., non-financial eligibility requirements, resources, income, and deductions). The verified circumstances and the resulting benefit level determined by the quality control review shall be compared to the benefits authorized by the State agency. The review of active cases shall include: a household case record review; a field investigation; the identification of any variances; an error analysis; and the reporting of review findings.

(b) Household case record review. The reviewer shall examine the household case record to identify the facts relating to the household's eligibility and basis of issuance. The case record review shall include all information applicable to the eligibility action under review, including the application and, as applicable, the monthly, quarterly, or periodic report and worksheet in effect as of the review date. Documentation contained in the case record should be used as verification if it was verified using documentary evidence at the time of the certification action. If during the case record review the reviewer can determine and verify the household's ineligibility, the review can be terminated at that point, provided that, if the determination is based on information not obtained from the household, the correctness of that information is confirmed as specified in paragraph (c)(2) of this section. The reviewer shall utilize information obtained through the case record review to complete column (2) of the Form FNS–380, and to tentatively plan the content of the field investigation.

(c) * * * In obtaining documentary evidence and collateral contact verification, the State agency is encouraged to utilize technology to assist in the gathering of documentary evidence, however the State agency shall ensure it preserves the privacy and confidentiality of the household regardless of what technology it uses.

(1) Personal interviews. State agencies shall conduct interviews in a manner that respects the rights, privacy, and dignity of the participants. The personal interview shall be a telephone interview unless the household requests a face-to-face interview or indicates they lack access to a telephone. Prior to conducting the personal interview, the reviewer must notify the household that it has been selected, as part of an ongoing review process, for review by quality control, that a personal interview will be conducted in the future, and that the household may request a face-to-face interview in lieu of a telephone interview. For face-to-face interviews, the interview may take place at the participant's home, at an appropriate State agency certification office, by secure video call, or at a mutually agreed upon alternative location. Should a face-to-face interview be warranted, the State agency shall determine the best location for the face-to-face interview, taking into account input from the household, including any hardship conditions or disability needs of the household. If the household meets any of the hardship conditions at 7 CFR 273.2(e)(2), the quality control reviewer shall either conduct the personal interview with the participant's authorized representative, if one has been appointed by the household, or at a place of the participant's choosing. During the personal interview with the participant, the reviewer shall:

* * * * *

(iii) Request and review with the household documentary evidence in the case file, as well as documentary evidence that may be in the household's possession, and secure information about collateral sources of verification; and

* * * * *

(d) * * *

(2) * * *

(ii) * * * For this exclusion to apply, the case record must include documentation specifying the elements of eligibility for which verification was postponed.

(3) Other findings. All QC review findings must be reported to the local office by the State agency. However, State agencies may determine if and how to act upon findings made during the review that are pertinent to the case record but do not result in a variance. For example, the State may establish its own procedures for cases when a household member's age is shown incorrectly in the case record, if their age is unrelated to an element of eligibility.

* * * * *

(f) * * *

(1) * * * In addition, the reviewer shall code and report any other variances related to eligibility which were discovered and verified during the course of the review and the State agency shall notify the local office of all variances in elements of eligibility.

(2) * * * The reviewer shall code and report any other variances in the basis of issuance which were discovered and verified during the course of the review and the State agency shall notify local offices of all variances in elements of the basis of issuance.

(g) * * *

(1) * * * The total allotment issued, shall be coded and reported as an error when the case is reported as not complete.

* * * * *

(ii) If a household refuses to cooperate with the quality control reviewer and the State agency has taken other administrative steps to obtain that cooperation without obtaining it, the household shall be notified of the penalties for refusing to cooperate with respect to termination and reapplication and of the possibility that its case will be referred for investigation for willful misrepresentation. If a household refuses to cooperate after such notice, the reviewer must attempt to complete the case and shall report the household's refusal to the State agency for termination of its participation without regard for the outcome of that attempt. For a determination of refusal to be made, the household must be able to cooperate but clearly demonstrate that it will not take actions that it can take and that are required to complete the quality control review process. In certain circumstances, the household may demonstrate that it is unwilling to cooperate by not taking actions after having been given every reasonable opportunity to do so, even though the household or its members do not state that the household refuses to cooperate. Examples of when a household appears to be unwilling to cooperate with a QC review that have the effect of a refusal to cooperate shall include:

(A) The household does not attend an agreed upon interview with the reviewer and then does not contact the reviewer within 10 calendar days of the date of the scheduled interview to reschedule the interview.

(B) The household does not return a signed release of information statement to the reviewer within 10 calendar days of either agreeing to do so or receiving a request from the reviewer sent Certified Mail-Return Receipt Requested.

(C) The household does not respond to any communication sent from the reviewer, and either:

( 1) The household has not responded to a letter sent to the household's current known address via Certified Mail-Return Receipt requesting a response within 30 calendar days of the date of receipt, or

( 2) The household has not responded to a written request for quality control contact (RFQCC). The RFQCC, sent by the State agency at the reviewer's request, must have advised the household of the following: that quality control is trying to contact them for a review; the requirement of the household to cooperate with the quality control review; the penalties for not responding to the RFQCC and not cooperating with the quality control review; and the contact information the household must use to get in touch with the quality control reviewer or office. The RFQCC must have afforded the household at least 10, but no more than 15 calendar days to respond from the date on the letter. RFQCC requests from the reviewer must be fulfilled by the State agency.

( i) If the household does not respond to the RFQCC by the deadline provided in the request, the reviewer must report the household's failure to respond to the State agency and request the State agency issue a notice of adverse action to suspend the household for one month before a termination becomes effective. The notice of adverse action must include the reason the household's benefits were suspended, an explanation that the consequence for failing to respond to quality control during the suspension period is for their benefits to be terminated for refusal to cooperate with quality control, and the quality control reviewer's or office's contact information.

( ii) If the household does not respond to the RFQCC, but does respond during the period of suspension and cooperates with quality control, the reviewer must notify the State agency to reinstate the household without requiring a new application and issue the allotment for the month of suspension.

(D) In these and other situations, if there is any question as to whether the household has merely failed to cooperate, as opposed to refused to cooperate, the household shall not be reported to the State agency for termination.

* * * * *

(2) * * *

(i) Death of all members of a household if they died after the eligibility action but before the review could be undertaken or completed;

(ii) The household moved out of State after the eligibility action but before the review could be undertaken or completed;

* * * * *

(vi) A case incorrectly listed in the active frame; or

(vii) A household appealing the eligibility action under review.

§ 275.13

[Amended]

8. In § 275.13:

a. Amend paragraph (c)(2) introductory text by removing the words “Immigration and Nationalization Services (INS)” and adding in their place “USCIS” and removing the words “Systematic Alien Verification for Entitlements (SAVE)” and adding in their place “SAVE”; and

b. Amend paragraph (c)(2)(iii) by removing “INS” and adding in its place “USCIS” in both occurrences.

§ 275.16

[Amended]

9. In § 275.16:

a. Remove the phrase “of 1 percent or more in negative cases” and add in its place the phrase “above the national average CAPER” in paragraph (b)(2);

b. Remove paragraph (b)(4); and

c. Redesignate paragraph (b)(5) as paragraph (b)(4).

10. In § 275.21, revise paragraph (b)(1) to read as follows:

§ 275.21

Quality control review reports.

* * * * *

(b) * * *

(1) The State agency shall use SNAP–QCS, FNS's automated, web-based QC System for State agency users, to input, edit, and upload supporting evidence and information necessary to understand the disposition and findings for all active and negative sampled cases.

* * * * *

11. In § 275.23:

a. Remove the words “active case, payment, and negative case error rate” and add in its place the words “payment and case and procedural error rates” in paragraph (b) introductory text;

b. Remove the words “active case error rate, payment error rate, and negative case error rate” and add in its place the words “payment and case and procedural error rates” in paragraph (b)(1);

c. Revise paragraphs (b)(2)(i)(A) and (B) and (b)(2)(ii) and (iii);

d. Revise the second and third sentences in paragraph (c); and

e. Revise the second sentence in paragraph (d)(1).

The revisions read as follows:

§ 275.23

Determination of State agency program performance.

* * * * *

(b) * * *

(2) * * *

(i) * * *

(A) Y₁ ′ = y₁ + b₁ (X₁ −x₁), where Y₁ ′ is the estimated average value of allotments overissued to eligible and ineligible households in the full Quality Control (QC) sample; y₁ is the average value of allotments overissued to eligible and ineligible households in the re-review sample according to the Federal finding; b₁ is the estimated regression coefficient through a regression of the Federal findings of allotments overissued to eligible and ineligible households on the corresponding State agency findings; x₁ is the re-review average value of allotments overissued to eligible and ineligible households according to State agency findings; and X₁ is the weighted average value of allotments overissued to eligible and ineligible households in the full QC sample according to State agency's findings. Based on the sample design, only X₁ is weighted to account for the probability of selection in the full QC sample.

(B) Y₂ ′ = y₂ + b₂ (X₂ −x₂), where Y₂ ′ is the estimated average value of allotments underissued to households included in the active error rate; y₂ is the average value of allotments underissued to participating households in the re-review sample according to the Federal finding; b₂ is the estimated regression coefficient obtained through a regression of the Federal findings of allotments underissued to participating households on the corresponding State agency findings; x₂ is the re-review average value of allotments underissued to participating households according to State agency findings; and X₂ is the weighted average value of allotments underissued to participating households in the full QC sample according to the State agency's findings. Based on the sample design, only X₂ is weighted to account for the probability of selection in the full QC sample.

* * * * *

(b) * * *

(2) * * *

(ii) If FNS determines that a State agency has sampled incorrectly, estimated improperly, or has deficiencies in its QC data management system, FNS will correct the State agency's payment error rate and case and procedural error rate based upon a correction to that aspect of the State agency's QC system which is deficient. If FNS cannot accurately correct the State agency's deficiency, FNS will assign the State agency a payment error rate or case and procedural error rate based upon the best information available. After consultation with the State agency, the assigned payment error rate can then be used in a liability determination, if applicable per the rules governing liabilities at 7 CFR 275.23(d). State agencies shall have the right to appeal the assignment of an error rate in this situation in accordance with the procedures of part 283 of this chapter. State agencies assigned error rates that do not result in the determination of a liability amount, as discussed in 7 CFR 275.23(d), are not eligible for appeal.

(iii) Should a State agency fail to sample and disposition its required minimum annual sample size for the fiscal year, FNS shall adjust the State agency's regressed error rate using the following equations:

(A) r₁ ″ = r₁ ′ + 5(1− C) S_1, where r₁ ″ is the adjusted regressed overpayment error rate, r₁ ′ is the regressed overpayment error rate computed from the formula in paragraph (b)(2)(i)(C) of this section, C is the State agency's rate of completion of its required sample size expressed as a decimal value, and S₁ is the standard error of the State agency sample overpayment error rate. If a State agency completes all of its required sample size, then r₁ ″ = r₁ ′.

(B) r₂ ″ = r₂ ′ + 5(1− C) S_2, where r₂ ″ is the adjusted regressed underpayment error rate, r₂ ′ is the regressed underpayment error rate computed from the formula in paragraph (b)(2)(i)(C) of this section, C is the State agency's rate of completion of its required sample size expressed as a decimal value, and S₂ is the standard error of the State agency sample underpayment error rate. If a State agency completes all of its required sample size, then r₂ ″ = r₂ ′.

* * * * *

(c) * * * FNS shall determine and announce the national average payment error rate for the fiscal year by June 30 following the end of the fiscal year and shall determine and announce the national average case and procedural error rate for the fiscal year by September 30 following the end of the fiscal year. At those times, FNS shall notify all State agencies of their individual payment and case and procedural error rates, respectively, and payment error rate liabilities, if any.

* * * * *

(d) * * *

(1) * * * The national performance measure is the sum of the products of each State agency's payment error rate multiplied by that State agency's proportion of the total value of allotments issued for the fiscal year using the most recent issuance data available at the time the State agency is notified of its payment error rate. * * *

12. Revise § 275.24 to read as follows:

§ 275.24

Performance measures.

(a) Performance measures. FNS will measure performance for the following categories of performance measures:

(1) Payment accuracy. FNS will assess State agencies annually for individual and overall payment accuracy, including measurements for overpayments and underpayments of SNAP benefits issued based on the results of cases reviewed in the actives sampling frame.

(i) Improvements in payment accuracy. FNS will assess the percentage point decrease in a State agency's combined payment error rates based on the comparison of the State agency's validated payment error rates for the performance measurement year to those of the previous fiscal year.

(ii) [Reserved]

(2) Case and procedural errors. FNS will assess State agencies annually to produce case and procedural error rates based on the results of cases reviewed in the negative sampling frame.

(i) Most improved case and procedural error rate. FNS will assess the percentage point decrease in a State agency's case and procedural error rates, based on the comparison of the State agency's performance measurement year's validated quality control case and procedural error rates for the performance measurement year with to those of the previous fiscal year.

(ii) [Reserved]

(3) Program access index (PAI). FNS will annually assess the degree to which each State agency provides low-income people access to SNAP benefits. The PAI is the ratio of participants to persons with incomes below 125 percent of poverty, as calculated in accordance with paragraph (a)(3)(i) of this section.

(i) Data. For the number of participants (numerator), FNS will use the administrative annual counts of participants minus the average amount of new participants certified under special disaster program rules by the State agency averaged over the calendar year. For the number of people below 125 percent of poverty (denominator), FNS will use the Census Bureau's American Community Survey (ACS) count of people below 125 percent of poverty for the same calendar year. FNS will reduce the count in each State where a Food Distribution Program on Indian Reservations (FDPIR) program is operated by the administrative counts of the number of individuals who participate in this program averaged over the calendar year.

(ii) [Reserved]

(4) Application processing timeliness. FNS will annually assess the timeliness of processed applications for each State agency.

(i) Data. FNS will use quality control data to determine each State agency's rate of application processing timeliness.

(ii) Timely processed applications. A timely processed application is one that provides an eligible applicant the “opportunity to participate” as defined in 7 CFR 274.2, within thirty days for normal processing or 7 days for expedited processing. New applications that are processed outside of this standard are untimely for this measure, except for applications that are properly pended in accordance with 7 CFR 273.2(h)(2) because verification is incomplete and the State agency has taken all the actions described in 7 CFR 273.2(h)(1)(i)(C). Such applications will not be included in this measure. Applications that are denied will not be included in this measure.

(b) [Reserved]

Cynthia Long,

Administrator, Food and Nutrition Service.

Footnotes

1.   https://www.ers.usda.gov/​topics/​food-nutrition-assistance/​food-assistance-data-collaborative-research-programs/​snap-and-wic-administrative-data/​.

2.  In fiscal year 2012, the procedures for reviewing cases in the negative frame changed to include the State's procedural processes in determining a negative case's validity. FNS has referred to the negative error rate since then as the case and procedural error rate, or CAPER, to reflect this change.

3.   https://www.govinfo.gov/​content/​pkg/​FR-2023-04-18/​pdf/​2023-08122.pdf.

4.   https://www.govinfo.gov/​content/​pkg/​FR-2021-09-02/​pdf/​2021-18743.pdf.

5.   https://www.govinfo.gov/​content/​pkg/​FR-2023-04-18/​pdf/​2023-08005.pdf.

6.  S.375—116th Congress (2019–2020): Payment Integrity Information Act of 2019 | Congress.gov |Library of Congress: https://www.congress.gov/​bill/​116th-congress/​senate-bill/​375.

7.   https://www.usda.gov/​oig/​webdocs/​27601-0002-41.pdf.

8.   https://www.fns.usda.gov/​snap/​enhancing-completion-rates-supplemental-nutrition-assistance-program-snap-quality-control-reviews.

9.   https://fns-prod.azureedge.net/​sites/​default/​files/​snap/​FNS_​310_​Handbook.pdf.

10.   https://fns-prod.azureedge.net/​sites/​default/​files/​snap/​QCPolicyMemo05-01.pdf.

11.   https://www.fns.usda.gov/​snap/​integrity-snap-quality-control-system.

12.   https://www.fns.usda.gov/​snap/​fns-handbook-310 https://www.fns.usda.gov/​snap/​fns-handbook-310.

13.  As determined by guidelines promulgated pursuant to 7 CFR 273.2(f)(2).

14.   https://www.federalregister.gov/​documents/​2018/​06/​01/​2018-11849/​request-for-information-supplemental-nutrition-assistance-program-snap-quality-control-integrity-and.

15.  Discussed throughout 7 CFR 275.

16.  FNS temporarily provided State agencies the flexibility to conduct QC interviews by telephone to assist State agencies in case completion during the COVID–19 public health emergency. The flexibilities have been extended through September 2024.

17.  Under Section 16(c)(1)(A)(ii) of the FNA, the tolerance level for excluding small errors is $37 for fiscal year 2014, adjusted by the percentage adjustment of thrifty food plan for each subsequent fiscal year. The most recently announced threshold, for FY 2023, is $54.

18.  As previously noted, the OIG's 2015 report recommended that this two-comparison allotment test be eliminated.

19.  QC cases in the active frame are comprised of households certified prior to or during the sample month and issued a SNAP allotment for the sample month. QC review years follow the Federal fiscal year from October through September the following year.

20.  “Not Subject to Review” is a term that refers to cases that are not subject to quality control review. This means they do not meet the requirements to be reviewed by a QC reviewer. Examples include cases under active investigation by the State's fraud unit and cases where the household did not receive an allotment for the sample month under review.

21.  The FQCR review may result in agreement or disagreement with the State's findings. When there is a disagreement, the State can dispute the Federal finding. If the State office and RO cannot agree on the outcome of a case, the State may appeal to the national arbitrator, a neutral third party whose decisions are final.

22.   https://www.govinfo.gov/​content/​pkg/​FR-2018-06-01/​pdf/​2018-11849.pdf.

23.   https://archives.federalregister.gov/​issue_​slice/​1984/​2/​17/​6278-6313.pdf#page=​33.

24.  In fiscal year 2012, the procedures for reviewing cases in the negative frame, changed to include the State's procedural processes in determining a negative case's validity. FNS has referred to the negative error rate since then as the case and procedural error rate, or CAPER, to reflect this change.

25.  For a definition of collateral contact, see 7 CFR 273.2(f)(4)(ii)— https://www.ecfr.gov/​current/​title-7/​subtitle-B/​chapter-II/​subchapter-C/​part-273#p-273.2(f)(4)(ii).

26.  State systems contain information on participants in a multitude of different programs, each of which can have (as some currently do) statutory and regulatory language prohibiting the disclosure of such information outside of said program. As such, the State systems must be able to limit user access to SNAP information only before sharing the system with FNS for SNAP QC purposes to ensure compliance with Federal law.

27.   https://www.federalregister.gov/​documents/​1997/​06/​02/​97-13946/​food-stamp-program-quality-control-provisions-of-the-mickey-leland-childhood-hunger-relief-act.

28.   https://www.govinfo.gov/​content/​pkg/​FR-2010-06-11/​pdf/​2010-13446.pdf.

29.  Not all adverse actions are against an entire household. Sometimes an adverse action will occur against a specific participant in a household.

30.  Enhancing SNAP Quality Control Completion Rates Final Report, January 2016, https://fns-prod.azureedge.us/​sites/​default/​files/​ops/​SNAPQCCompletion.pdf.

31.  See Section 5.08 of the Government Auditing Standards, GAO 21–368G, April 2021, Government Auditing Standards: 2018 Revision Technical Update April 2021 (Supersedes GAO–18–568G) | U.S. GAO.

32.  See Section 3.19 of the Government Auditing Standards, GAO 21–368G, April 2021, Government Auditing Standards: 2018 Revision Technical Update April 2021 (Supersedes GAO–18–568G) | U.S. GAO.

33.  See Section 16(c)(1)(A)(ii) of the FNA or 7 CFR 275.12(f)(2).

34.  OIG Audit Report 27601–0002–41.

35.  GAO 16–708–T, https://www.gao.gov/​products/​GAO-16-708T.

36.   https://www.govinfo.gov/​content/​pkg/​FR-2018-06-01/​pdf/​2018-11849.pdf.

37.   https://www.govinfo.gov/​content/​pkg/​FR-2018-06-01/​pdf/​2018-11849.pdf.

38.  31 U.S.C. 3351.

39.   https://www.whitehouse.gov/​wp-content/​uploads/​2018/​06/​M-18-20.pdf; Page 8.

40.   https://www.govinfo.gov/​content/​pkg/​FR-2010-06-11/​pdf/​2010-13446.pdf.

41.   https://www.govinfo.gov/​content/​pkg/​FR-2018-06-01/​pdf/​2018-11849.pdf.

42.   https://www.govinfo.gov/​content/​pkg/​FR-2018-06-01/​pdf/​2018-11849.pdf.

43.  FNS 380—OMB Form 0584–0074; FNS 245–OMB Form 0584–0034; FNS 380–1—OMB Form 0584–0299.

44.   https://www.govinfo.gov/​content/​pkg/​FR-2018-06-01/​pdf/​2018-11849.pdf.

45.  The changes in sampling procedures only affect active cases, not negative cases.