STATEMENT BY RMA ADMINISTRATOR ELDON GOULD
U.S. Department of Agriculture
Before the House Agriculture Subcommittee on
General Farm Commodities and Risk Management - CONTINUED
Jun 7, 2007
Detection via Data Mining
RMA is making significant progress in preventing fraud, waste and abuse through the expanded use of data mining. As part of ARPA, data warehousing and data mining techniques were explicitly identified as tools to be used by RMA to strengthen the crop insurance program’s oversight efforts. RMA contracts with the Center for Agribusiness Excellence (CAE) at Tarleton State University to develop these technologies. Since employing these technologies in 2001, RMA has achieved substantial program savings through
proactive efforts to identify program vulnerabilities and abuse.
RMA continues to use data mining to identify anomalous producer, adjuster, and agent program results and, with the assistance of Farm Service Agency (FSA) offices, conducts growing-season spot checks to ensure that new claims for losses are legitimate. The annual spot check list combines the strengths of data mining technologies and the farm-level knowledge of FSA, to identify and monitor those producers whose crop insurance losses are not consistent with those of their neighbors. This effort alone has achieved reductions from prior year indemnities for the producers selected of more than $430 million dollars since the 2002 crop year. Specifically, indemnities on spot-checked policies were reduced approximately $112 million in 2002, $82 million for 2003,
$71 million in 2004, $138 million in 2005 and $27 million in 2006.
More importantly, these reductions are achieved without RMA or FSA having to issue administrative
sanctions or engage in lengthy and costly criminal investigations to curb program abuse. These reductions represent
more than a $20 return for every dollar spent by RMA on data mining since its inception. Our analysis shows that
this change in claims behavior for most producers persists for several years, resulting in overall program
compliance benefits that are even higher over a longer-term period.
Data mining findings also demonstrated that the considerable majority of producers participating in the crop insurance program used the risk management tools we offer exactly as they were intended. CAE, using an analysis technique known as a decision tree, classified the entire crop insurance book of business into a range of behavior, from those producers who almost never had losses to those who had frequent and severe losses. Through this method, CAE was able to demonstrate that most producers used the risk management tools as intended and only a small percentage, about 0.2 percent,
of producers exhibited behavior that warranted future review.
In addition, CAE conducts internal data mining research for RMA to assist compliance and underwriting efforts and any other research deemed necessary by the agency to improve the effectiveness and efficiency of the crop insurance program. CAE currently produces approximately 160 such research products per year for RMA, including products such as crop simulation models, planting
date studies and methods for correctly identifying high-risk land.
RMA also uses data mining to verify compliance with established rules and regulations. For example,
data mining identified policies where a comparison of past claims and production data identified certain companies or their agents who had failed to use claim production data to establish future approved yields, as required by regulation. RMA provides this information to the companies to assist them in correcting producer data when such errors occur.
Outside audit bodies such as the USDA’s OIG and the GAO have also recognized our success with the use of data warehousing and data mining technologies. OIG recommended that USDA employ data mining in other farm programs. Further, both OIG and GAO have been customers, using
CAE on occasion to assist them with audits of farm programs.
The benefits from using data warehousing and data mining technologies have increased every year since its inception. RMA expects the benefits generated from using these technologies to continue and plans to expand its use of data mining
technologies to other applicable areas of the program in the near future.
The President’s Fiscal Year (FY) 2008 Budget includes a proposal that would expand the uses of mandatory ARPA
research and development funding for data mining as well as for the Comprehensive Information Management System (CIMS). Specifically, the FY 2008 Budget would authorize the use of $5.4 million for replacement of equipment and $3.6 million to continue regular operations of data mining.
RMA continues to make progress in the Administrative Sanctions arena. In 2005, RMA imposed 24 sanctions, such
as suspensions, debarments, and disqualifications on producers, agents and loss adjusters found to have violated
approved policies and procedures. For 2006, RMA imposed 41 sanctions and had 53 additional sanctions pending at
the end of the year. RMA also routinely publishes the Department of Justice press releases
prosecutions of crop insurance program abuse on our Web site as a
reminder to program participants that maintaining integrity is critical.
We are improving the timing and quality of our sanctions requests as well. RMA continues to work with USDA’s Office of General Counsel (OGC) to limit the number of cases declined due to insufficient evidence. This improvement is attributable to Compliance personnel becoming more proficient at identifying evidence
and establishing cases that will pass legal sufficiency requirements.
Finally, modifications to the Administrative Sanctions regulations that were identified by GAO as requiring publication are in clearance. These regulations will
formalize all the sanctions authority Congress provided RMA in ARPA.
In 2005, GAO audited RMA’s overall compliance activities, and recommended areas for improving our compliance efforts. GAO made several recommendations that RMA accepted and is working to implement. However, data mining remains central to our
compliance efforts because it is cost efficient and cost effective.
Within current resources, compliance managers also continue to concentrate on the mission-critical tasks of evaluating and improving new processes to prevent and deter fraud, waste and abuse in the crop insurance program. We have dedicated significant resources to building and adapting a reporting and tracking system
to complement and integrate the oversight mandates established by ARPA.
Information Technology (IT) System Improvements
A critical area in program integrity improvement is enhancing the capability of RMA’s IT system. The number and types of crop insurance programs is ever expanding and growing more complex. ARPA also instituted new data reconciliation, data mining and other anti-fraud, waste and abuse activities that require the data to be used in a variety of new ways. The current IT system was not designed to handle these types of data operations. Consequently, the data must be stored in multiple databases, which increases data storage
costs and processing times, and increases the risk of data errors.
The President’s FY 2008 Budget includes two proposals that will facilitate funding of our IT needs.
The first is similar to last year’s request, which required insurance providers to share in the cost to develop and maintain a new IT system. Insurance providers would be assessed a fee based on one-half cent per dollar of premium sold. The fee is estimated to generate an amount not to exceed $15 million annually. After the new IT system has been developed, the assessment would be shifted to fund maintenance and would be expected to
reduce the annual appropriation of the salaries and expenses account of RMA.
The second, as noted earlier, would expand the uses of mandatory ARPA research and development funding for data mining and data warehousing
activities required by ARPA, and the testing and development of CIMS.
Conflict of Interest Supplementary Guidance
RMA recognizes that certain types of interactions between agents, loss adjusters and policyholders pose
serious conflict of interest challenges to the integrity of the crop insurance program. RMA investigations and independent audits by OIG and GAO have identified instances where crop insurance claims have been influenced by such conflicts.
The 2005 SRA contained new and enhanced provisions that strengthened RMA’s ability to prevent and detect those conflicts of interest that might adversely affect program integrity. Specifically, the SRA strengthened provisions that 1) prohibited certain conduct by agents during the loss adjustment process, and 2) required increased conflict of interest disclosure by agents, loss adjusters and insurance company employees.
To assist the companies in implementing new SRA provisions dealing with prohibited activities of agents
during loss adjustment, RMA worked closely with companies and agents to develop a comprehensive guidance document
that reflected tough but workable standards. RMA issued the resulting
Manager’s Bulletin in October 2005. The
reaction of the crop insurance industry, agent associations and oversight bodies has generally been very positive to these standards.
After addressing this first area of concern, RMA has now turned to the problem of developing guidance on conflict of interest disclosure. The SRA requires that all company employees and affiliates disclose any potential conflicts of interest to the companies and, in turn, to RMA. Such disclosure is used to determine what conduct may be prohibited and what reviews must be done by the company. RMA has listened to the comments of the industry regarding conflict of interest disclosure to ensure that guidance will contain a workable standard that will be consistent across all companies and will
provide important information for RMA’s data mining efforts.
After seeking company input at the recent National Crop Insurance Service’s Program Integrity Conference, RMA is now finalizing a Manager’s Bulletin that contains further guidance to assist insurance providers in implementing changes to the SRA regarding conflict of interest disclosure. The Bulletin will establish standards for reporting conflicts of interest by insurance company employees, agents, and loss adjusters. This effort will promote program integrity and ensure adequate internal controls based on the identification of certain conflict of interest problems in past audits and
investigations of fraud, waste, and abuse in the program.
Simplification of the Federal Crop Insurance Program
Simplification of the program is a priority of both RMA and the FCIC Board. As new programs have been added, more complexities have arisen.
As stated above, RMA is developing a combination policy, which combines the existing Actual Production History, Crop Revenue Coverage, Income Protection, Indexed Income Protection and Revenue Assurance plans of insurance into one consolidated insurance plan (Combo). We have been working on this for some time now, and the draft final rule is being completed and is intended to be effective for the 2009 crop year, with publication slated for late 2007. We believe this change will provide producers a broader array of insurance options, in a more
straightforward process, and improve product delivery and operations.
RMA is also working closely with FSA to simplify our joint reporting requirements. Where feasible, the two agencies are coordinating certain, similar program requirements seeking commonality and consistency to ease the reporting burden on the producer and on the agencies. Our objective is to vastly improve the reporting accuracy of producer information and share the data between the companies and FSA, ensuring
greater program integrity for several different USDA farm programs.
RMA is actively working on the second phase of a project to implement section 10706 of the 2002 Farm Bill, known as CIMS, which will simplify and improve the programs administered by RMA and FSA. This project will provide an information system that allows RMA, FSA other USDA entities and companies to process, share and report on approved common information. The second phase of the project focuses on the sharing and analysis of existing RMA and FSA producer and acreage data. Recommendations have been provided to both RMA and FSA for subject matter experts to
review elements for producers, land locations, crops and acreage reporting.
The common component of CIMS has been operational internally since July 2006. It is loaded weekly with over 141 million producer and acreage records from RMA and FSA for 2005, 2006 and 2007. This data is processed and is electronically available to approved RMA and FSA users to provide participation summary reports, information on individual producers and discrepancies in reported acreage. Once RMA’s and FSA’s System of Records have been updated for CIMS, the companies will have electronic access to their insured producers’ information only.
All data is secure and subject to controls to prevent unauthorized access.
In March 2006, a ‘Notification Area’ was added to the CIMS web interface to allow FSA County
Offices and companies to communicate on data issues identified by CIMS.
Administration of the crop insurance program requires all interested parties working together to identify viable insurance products and solutions that meet the needs of the agricultural community. Moreover, if the program is to continue to be successful, the resources to provide the checks and balances necessary to guard against the risks of fraud, waste and abuse need more focus and priority.
RMA continues to improve and update the terms and conditions of existing crop insurance policies to enhance coverage and efficacy of the policies, as well as to clarify and define insurance protection and the duties and responsibilities of the policyholder and companies to improve the understanding, use and integrity of the program.
When I accepted this position, Secretary Johanns charged me with administering the crop insurance program in a timely, responsible, and farmer-friendly manner. I will continue to work with the insurance companies, agents’ groups, producer groups and, of course, the Congress, to meet our common goals of providing effective insurance products, processing timely and accurate claims when losses occur and identifying and eliminating fraud, waste and abuse in the program to the greatest extent possible. Thank you all for the support provided by the Committee to help improve program integrity within the Federal crop insurance program. We have much to be proud
of and much to look forward to in continuing to work together.
Again, thank you for the opportunity to participate in this important hearing. I look forward to responding to questions on these issues.