Second Error Rate Pilot Report
Download this document in Word (1,076 KB) or PDF (1.28 MB)
III. MEASURING IMPROPER PAYMENTS IN FIVE STATES
The contractor visited five States—Florida, Kansas, New Jersey, Oregon, and West Virginia—in the second pilot to gather information about the States’ policies and practices regarding improper payments. This chapter presents a summary of the processes, standards and procedures, improper payment strategies, and automation that characterize the five States in the second pilot and contrasts similarities and differences with the four States involved in the first pilot.
Similar to the four States in the first pilot, the five States in the second pilot use many similar processes to identify, investigate, and recover improper payments.
All nine States in both pilots identify improper payments in four primary ways:
- Notification of the child care agency of suspected fraud by providers or the public;
- Exception reports and matches, produced on a regular basis from the State’s automated child care tracking system, identify areas of potential improper payments;
- Regular audits or reviews identify improper payments; and
- Ad hoc studies or reviews identify improper payments.
An investigation typically begins after the State identifies a payment problem. The staff persons responsible for calculating the overpayment differ among the five States in the second pilot. In four States, local staffs calculate the overpayment. In New Jersey, Florida, and West Virginia, Child Care Resource and Referral (CCR&R) agencies calculate the overpayment. The local Social Rehabilitation Services (SRS) offices complete the overpayment calculations in Kansas. In Oregon local staff refer the overpayment information to a State level central unit. Oregon has a central State level Overpayment Writing Unit in the Department of Human Resources that calculates the overpayments.
All nine States in both pilots follow similar processes to recover funds for overpayments. The process proceeds in one of the following ways, listed in order from least to most intensive action:
- The State sends the person receiving the overpayment a notice describing the overpayment and the person's responsibility to resolve the debt;
- The State establishes a repayment schedule, if possible. In some States if the client is an active recipient, a reduction of current benefits may occur;
- The State initiates judgments, liens, and garnishments if the participant does not voluntarily repay; and
- The State employs tax or lottery intercepts. This is a process where the State will intercept any tax refund or lottery winnings that would normally be paid to an individual and apply those monies toward the overpayment amount.
B. Standards, Processes, and Procedures
As was seen in the first pilot, the five States in the second pilot indicate a trend toward establishing more formalized standards, processes, and procedures regarding reducing improper payments. The five second pilot States made the following procedural changes:
- While Florida’s eligibility policy stipulates that a case has, at a minimum, an annual recertification, the State now randomly selects 50 percent of all cases for recertification at six month intervals. This interim recertification also requires a face-to-face interview.
- Kansas has established a Quality Assurance Unit. The unit conducted two child care quality assurance reviews during 2006. Kansas contracted with a firm to develop software that tracks all case review findings and provides aggregate review data. Kansas utilized the software in July 2006 for the mandatory Food Stamp reviews and plans to include the child care reviews in the database in 2007.
- New Jersey established a Quality Assurance in Child Care committee to develop policies and procedures to decrease the potential for improper payments. The committee includes both program and systems staff members. A recently created Program Integrity Unit uses data matching and shared data warehouse technology to identify improper payments.
- Oregon completes a monthly desk audit of 200 randomly selected billing forms. A part of the State’s Food Stamp payment accuracy strategy is a monthly review of approximately 5000 cases. Since roughly 94 percent of these cases are linked to child care cases, the review now includes child care eligibility.
- West Virginia technical assistance staffs conduct random quarterly child care case audits. Prior to the pilot, the audits only targeted the automated system (FACTS) case record. The child care case audits now include a review of the physical case record. CCR&R agencies must develop and implement Quality Assurance Plans. Those plans now require payment accuracy initiatives if a CCR&R agency has an error rate that exceeds the State average.
This section provides a description of automated processes and systems the five States in the current pilot use to reduce improper payments.
All of the States that participated in both pilots have a statewide automated system. The automated systems in the five States from the second pilot differ considerably in function, interfaces, and age. Even though New Jersey and Oregon are locally-administered States, those States mandate that providers use the statewide system to determine eligibility for the child care voucher application.
The Florida Child Care Enhanced Field System (EFS) is a Windows based system that has an eligibility determination module, generates payments, manages waiting lists, and collects data for Federal reports and the market rate survey. The system does not interface with other benefits programs. EFS produces a selection of management reports to identify improper payments. There is ad hoc report writing capability at the local level. Sharing of the local report formats occurs throughout the State through an e-mail Q & A system. An area of concern for the EFS and the quality of its data is its decentralized structure. Each of the 31 Early Learning Coalitions has an independent section and differing processes. The result is data that is inconsistent, incompatible, and centrally unavailable. The proposed enhanced system Early Learning Information System (ELIS) will address the weaknesses of this decentralized system.
The Kansas KSCARES system, a transferred mainframe system, has been operational since 1994. KSCARES determines eligibility, authorizes and tracks payments for the Child Care Subsidy program and processes TANF Work Program referrals and payments for employment related activities and support services. KSCARES interfaces with a number of other State systems including KAECSES (TANF and Food Stamp eligibility and Child Support Enforcement), the Electronic Benefits Transfer (EBT) system, and the Kansas Department of Health and Environment for access to child care provider data. Planning and development are currently underway to build a new integrated enterprise wide web based system that would incorporate the functionality of KSCARES and several other agency systems. This multi-system interface will strengthen eligibility processes by sharing data across programs.
The New Jersey automated system occupies three different databases. Case information for Work First New Jersey (WFNJ/TANF) resides on the OMEGA (TANF) system. The Care for Kids (NJCK) case information resides in the Child Care Automated Resources Eligibility System (CARES). The voucher payments system and case information on clients receiving Center Based Services (CBC) resides on the Contracts System (CTRX) and Contracts Accounting Tracking System (CATS). None of the systems interface. CARES includes intake, eligibility determination, agreement establishment, subsidy and co-pay calculation, voucher generation, attendance input, check write process, accounting procedures, letters, notices, forms, and reports. New Jersey’s Consolidated Assistance and Support System (CASS) is a re-engineering effort designed to integrate all of the Division of Family Development’s support systems by utilizing an Enterprise Architecture and Framework Approach. This future approach will increase New Jersey’s ability to share data across programs, eliminating the need to do data matching.
The Oregon Client Maintenance System (CMS) is one of thirty-seven components of a 28 year old legacy system. CMS bundles child care eligibility with TANF and Medicaid and interfaces with the Service Authorization System, the State’s payment system. The child care application process includes a match with other DHS systems as well as child support, Employment Department wage and unemployment records and the Social Security Benefit system. CMS calculates the provider payment based on the family’s co-pay, the provider’s usual charge, hours authorized and billed, age of the child(ren), type of care, and provider’s zip code. This capability remedies many of the payment accuracy problems associated with fee schedule and payments.
West Virginia transferred the Family and Child Tracking System (FACTS) from Oklahoma. FACTS is a customized statewide automated case management system for all Child Welfare, Child Care, and Adult Service Programs. In August 2000, the State modified FACTS to include case management and payment functionality for the Child Care Program. This change fully implemented the transition of the contracted CCR&R agencies to an automated system for on-line and real time determination of eligibility for child care services and an automated determination for child care payments. The FACTS System is designed to take information entered in the system by the CCR&R staff and use it to determine eligibility. This eliminates many errors with the exception of those caused by faulty data entry. The system has payment controls of eligible days and children only.
FACTS also provides the capability of accessing the State’s TANF, Food Stamp and Medicaid system (RAPIDS). This allows additional verification of information provided in the eligibility process.
Appendix K provides an extensive summary table for all nine States who participated in the two phases of the error rate pilot. As was the case in the first pilot, the five States in the second pilot were a mixture of State administered (Florida, Kansas and West Virginia) and County administered but State supervised (New Jersey, Oregon). Four States in each pilot defined an improper payment in either State law, administrative policy or both. Florida and Ohio were the only States where fraud was not articulated in State policy. All nine States follow similar processes to identify and recover overpayments funds. There is considerable uniformity in the approaches all nine States take in assessment of risk. Some form of audit, ad hoc review, compliance review or site visit exists in all nine states. Case reviews, either independent or in conjunction with other programs (Food Stamps), are common. Two of the nine States included technical support and training as major payment accuracy risk mitigators. Although all of the States that participated in the pilots have a statewide automated system, the systems differ considerably in function, interfaces, and age. All of the States have incorporated in their payment accuracy strategies various automated tools (edits, benchmark/exception/anomaly reports, data mining).
Findings and States' Responses >>
June, 2007