This is the accessible text file for GAO report number GAO-05-528 entitled 'Crop Insurance: Actions Needed to Reduce Program's Vulnerability to Fraud, Waste, and Abuse' which was released on October 31, 2005. This text file was formatted by the U.S. Government Accountability Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products' accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. Because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately. Report to the Chairman, Committee on Homeland Security and Governmental Affairs, U.S. Senate: September 2005: Crop Insurance: Actions Needed to Reduce Program's Vulnerability to Fraud, Waste, and Abuse: [Hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-05-528]: GAO Highlights: Highlights of GAO-05-528, a report to the Chairman, Committee on Homeland Security and Governmental Affairs, U.S. Senate: Why GAO Did This Study: Federal crop insurance protects producers against losses from natural disasters. In 2004, the crop insurance program provided $47 billion in coverage, at a cost of $3.6 billion, including an estimated $160 million in losses from fraud and abuse. The U.S. Department of Agriculture’s (USDA) Risk Management Agency (RMA) administers this program with private insurers. The Agricultural Risk Protection Act of 2000 (ARPA) provided new tools to monitor and control abuses, such as having USDA’s Farm Service Agency (FSA) conduct field inspections. GAO assessed, among other things, the (1) effectiveness of USDA’s processes to address program fraud and abuse and (2) extent to which the program’s design makes it vulnerable to abuse. What GAO Found: While RMA employs a range of processes to help prevent and detect fraud, waste, and abuse and has reported more than $300 million in savings over the past 4 years in the crop insurance program, GAO found that RMA does not effectively use all the tools it has available. Specifically: * Inspections during the growing season are not being used to maximum effect. Between 2001 and 2004, FSA conducted only 64 percent of the inspections RMA had requested. Without inspections, producers may falsely claim crop losses. * RMA’s data analysis of the largest farming operations is incomplete. According to GAO’s analysis, in 2003, about 21,000 of the largest farming operations in the program did not report individuals or entities with an ownership interest in these operations. As a result, USDA should be able to recover up to $74 million in claims payments. FSA did not give RMA access to the data needed to identify such individuals or entities. * RMA is not effectively overseeing insurance companies’ quality assurance programs. GAO’s review of 120 cases showed that companies completed only 75 percent of the required reviews and those that were conducted were largely paper exercises. * RMA has infrequently used its new sanction authority to address program abuses. RMA has not issued regulations to implement its new sanction authority under ARPA. RMA imposed only 114 sanctions from 2001 through 2004. Annually, RMA identifies about 3,000 questionable claims, not all of which are necessarily sanctionable. Eight recent crop insurance fraud cases, investigated by USDA’s Office of Inspector General and resulting in criminal prosecutions between June 2003 and April 2005, reflect these issues. Totaling $3 million in insurance claims, these cases show how producers, sometimes in collusion with insurance agents and others, falsely claim prevented planting, weather damage, and low production. In some cases, producers hid or moved production from one field to another. Several of these cases also demonstrate the importance of having FSA and RMA work together to identify and share information on questionable farming practices/activities. RMA’s regulations, as well as statutory requirements, create program design problems that hinder RMA officials’ efforts to reduce program abuse. For example, RMA’s regulations allow producers to insure fields individually rather than all fields combined. This option enables producers to “switch” reporting of yield among fields to either make false claims or build up a higher yield history on a field to increase its eligibility for higher insurance guarantees. High premium subsidies, established by statute, may also limit RMA’s ability to control program abuse because the subsidies shield producers from the full effect of paying higher premiums associated with frequent or larger claims. What GAO Recommends: To reduce program fraud, Congress should consider reducing premium subsidies to producers who repeatedly file questionable claims. In addition, USDA should (1) improve the effectiveness of growing season inspections, (2) recover payments from operations that failed to disclose producers’ ownership interests, (3) strengthen oversight of insurers’ use of quality controls, and (4) issue regulations for its expanded sanction authority. USDA agreed with most of GAO’s recommendations. However, it stated that it does not have the resources to conduct all growing season inspections. www.gao.gov/cgi-bin/getrpt?GAO-05-528. To view the full product, including the scope and methodology, click on the link above. For more information, contact Robert A. Robinson at (202) 512-3841 or robinsonr@gao.gov. [End of section] Contents: Letter: Results in Brief: Background: RMA Has Strengthened Procedures for Preventing Questionable Claims, but the Program Remains Vulnerable to Abuse: RMA's Regulations and Statutory Requirements Hinder RMA Officials' Efforts to Reduce Abuse in the Crop Insurance Program: RMA's Failure to Follow Its Guidelines Has Resulted in Program Losses for Some New and Expanded Crop Insurance Products: Conclusions: Matter for Congressional Consideration: Recommendations for Executive Action: Agency Comments and Our Evaluation: Appendixes: Appendix I: Objectives, Scope, and Methodology: Appendix II: Selected Information on the Federal Crop Insurance Program, 1981-2004: Appendix III: Results of Implementation of the Agricultural Risk Protection Act of 2000: Survey of FSA County Directors of USDA: Appendix IV: Results of Improving Compliance and Integrity in the Federal Crop Insurance Program: Survey of Crop Insurance Agents: Appendix V: Crop Insurance Fraud Cases Criminally Prosecuted, June 2003 to April 2005: Appendix VI: Comparison of Loss Ratios for Crop Development and Expansion Products: Appendix VII: Comments from the U.S. Department of Agriculture: GAO Comments: Appendix VIII: GAO Contact and Staff Acknowledgments: Related GAO Products: Tables: Table 1: Premium Subsidies Before and After ARPA: Table 2: Crop Insurance Policyholders Failing to Disclose Ownership Interest, by Entity Type, Crop Year 2003: Table 3: RMA Sanctions Requested and Imposed, Crop Years 2001 to 2005: Table 4: Number of USDA OIG Crop Insurance Investigations, Number of Referrals to the Department of Justice, and Case Disposition, Fiscal Years 1996 to 2005: Table 5: Production Reported by an Irrigated Cotton Producer Indicating Yield Switching: Table 6: Crop Insurance Fraud Cases Investigated by the USDA/OIG and Resulting in Criminal Prosecution, June 2003 to April 2005: Figures: Figure 1: Location of Producers Identified by RMA for Field Inspections, 2003: Figure 2: FSA Inspectors' Primary Reasons for Not Conducting Field Inspections of Producers with Notable Policy Irregularities: Abbreviations: ARPA: Agricultural Risk Protection Act of 2000: FCIC: Federal Crop Insurance Corporation: FSA: Farm Service Agency: FSI: Office of Forensic Audits and Special Investigations: NASS: National Agricultural Statistics Service: OIG: Office of Inspector General: OMB: Office of Management and Budget: RMA: Risk Management Agency: SRA: Standard Reinsurance Agreement: USDA: U.S. Department of Agriculture: Letter September 30, 2005: The Honorable Susan M. Collins: Chairman, Committee on Homeland Security and Governmental Affairs: United States Senate: Dear Chairman Collins: Federal crop insurance is part of the overall safety net of programs for American farmers. It provides protection for participating farmers against the financial losses caused by droughts, floods, or other natural disasters. Farmers' participation is voluntary, but the federal government encourages it by subsidizing the insurance premiums. In 2004, the crop insurance program provided $47 billion in insurance coverage for over 200 million acres of farmland at a cost of $3.6 billion to the federal government, including an estimated $160 million resulting from fraud, waste, and abuse. The U.S. Department of Agriculture's (USDA) Risk Management Agency (RMA), which supervises Federal Crop Insurance Corporation (FCIC) operations, has overall responsibility for administering the crop insurance program. RMA oversees the development of new insurance products and the expansion of existing insurance products to new areas to help farmers reduce the chance of financial loss. RMA is also responsible for ensuring that the program is carried out efficiently and effectively and for protecting against fraud, waste, and abuse. In this regard, RMA uses a broad range of tools, including compliance reviews, data mining, and on-site field inspections. RMA administers the program in partnership with private insurance companies that share a percentage of the risk of loss or opportunity for gain associated with each insurance policy written. RMA acts as a reinsurer-- reinsurance is sometimes referred to as insurance for insurance companies--for a portion of all policies the federal crop insurance program covers. In addition, RMA pays companies a percentage of the premium on policies sold to cover the administrative costs of selling and servicing these policies. In turn, insurance companies use this money to pay commissions to their agents who sell the policies and fees to adjusters when claims are filed. Insurance companies are responsible for reporting to RMA on policy activity, such as applications for insurance, reports of acres planted, and notices of loss. Insurance companies, as part of their contractual agreement with RMA, also have an important role to play in ensuring that the policies they issue are administered fairly and accurately. For example, insurance companies must conduct quality assurance reviews, such as field inspections for policies with a claim equal to or greater than $100,000, to examine whether the claims they have paid are in compliance with policy provisions. RMA conducts a regular nationwide review of insurance companies' compliance with the crop insurance program's procedures to ensure that companies' quality assurance programs are in place. RMA receives policy information from the insurance companies through its computerized acceptance system. Using this system, RMA checks all policies for completeness and accuracy. In 2004, RMA provided crop insurance on 1.2 million policies and paid claims on 330,000 of these policies through 17 insurance companies. The Federal Crop Insurance Act, as amended, requires RMA to set crop insurance premiums at actuarially sufficient rates, defined as a long-run loss ratio target of no more than 1.075. A loss ratio is calculated as claims paid divided by total premiums collected. A loss ratio greater than 1.00 indicates that the program paid more in claims than was collected in premiums. Generally, producers can purchase crop insurance to insure up to 85 percent of their normal harvest (yield). This yield is calculated by looking at a producer's actual production history. To obtain insurance and receive claims payments, producers must comply with the crop insurance program's provisions. Specifically, they must accurately report to their insurance company the number of acres planted; meet deadlines specified in the policy (e.g., for planting and harvesting crops); pay premiums when due (generally at the end of the growing season); and report any crop losses immediately. Producers are also obligated to exercise good farming practices to minimize the potential for losses and to report their Social Security numbers and the Social Security numbers of all persons with an ownership interest of 10 percent or more in the farming operation (e.g., a corporation) holding the policy. Over the years, concerns have arisen that some producers may have abused the crop insurance program by allowing crops to fail through neglect or deliberate actions in order to collect insurance and that some insurance companies have not exercised due diligence in investigating losses and paying claims.[Footnote 1] In part to improve compliance with, and the integrity of, the crop insurance program, Congress enacted the Agricultural Risk Protection Act of 2000 (known as ARPA). This act provided RMA and the USDA's Farm Service Agency (FSA) with new tools for monitoring and controlling program abuses. (FSA, which has an extensive field office structure, is generally responsible for helping producers enroll in agricultural support programs, overseeing these programs, and issuing program payments.) Specifically, ARPA required the Secretary of Agriculture to develop and implement a coordinated plan for FSA to assist RMA in the ongoing monitoring of the crop insurance program, including conducting fact-finding into allegations of program fraud, waste, or abuse; reporting the results of any such fact-finding to RMA; and assisting RMA and approved insurance companies in auditing a statistically appropriate number of claims made under any policy. Furthermore, ARPA required the Secretary of Agriculture to use information technologies, such as data mining and data warehousing, to administer and enforce the crop insurance program. Data mining is the analysis of data to establish relationships and identify patterns, while data warehousing is storing gathered data so that it can be easily analyzed, extracted, synthesized or otherwise used. RMA conducts data mining to target compliance reviews and investigations on suspect claims. Under USDA guidance, developed pursuant to a requirement in ARPA, RMA is to annually provide FSA and the insurance providers with a list of producers exhibiting high loss ratios, high frequency and severity of losses, or who are suspected of poor farming practices. RMA provides this list--called the spot-check list--every April to the appropriate FSA state offices for distribution to FSA county offices. The FSA county office is to conduct reviews on the larger of the first 10 producers or the top 5 percent of the producers on the list. Staff in FSA county offices review these cases for potential fraud, waste, and abuse by inspecting fields insured by the listed producers. They then refer the results of these inspections to RMA, which provides the results to the insurance companies holding the policies for the producers for further review or investigation, if appropriate. Finally, ARPA gave RMA additional authority to impose sanctions for program abuses. In addition, under the Improper Payments Information Act of 2002, RMA has to provide an estimate of error rates associated with program payments and report on action to reduce improper payments.[Footnote 2] RMA estimates improper payments to be about 5 percent of the claims paid annually. RMA acknowledges that this estimate is not based on a tested methodology and revised its sampling methodology, beginning in 2004, to provide a more accurate estimate. The Office of Management and Budget (OMB) has accepted RMA's proposed sampling methodology, which is to include a 3-year review cycle of insurance companies, to determine the federal crop insurance program's error rate, and satisfy the statutory requirements of the Improper Payments Information Act. RMA's 3-year review cycle will assess insurance companies' adherence to their contract with RMA, quality control guidelines, and RMA-approved policies and procedures. You asked us to examine RMA's procedures for assuring integrity in the crop insurance program. As agreed with your office, we (1) assessed the effectiveness of USDA's procedures and processes to prevent and detect fraud, waste, and abuse in selling and servicing crop insurance policies; (2) determined the extent to which program design issues may make the program more vulnerable to fraud, waste, and abuse; and (3) determined the effectiveness of USDA's procedures to assure program integrity in developing new crop insurance products. Also, as you requested, we provided examples of recent crop insurance fraud prosecutions to show the types of actions that producers, agents, and loss adjusters have used to circumvent RMA's procedures. To address these issues, we reviewed relevant statutory provisions and RMA's regulations and guidelines for managing the crop insurance program and spoke with RMA and FSA officials in headquarters and field offices. We also reviewed relevant reports, including RMA's most recent annual report to Congress in 2002. To assess the effectiveness of USDA's procedures and processes to prevent and detect fraud, waste, and abuse in selling and servicing crop insurance policies, we examined a nonrandom sample of 120 insurance claims from the 2,794 claims that RMA identified as having notable policy irregularities and warranting a field inspection in both 2003 and 2004. Of these 120 claims, 100 were the largest claims paid, and 20 were selected to ensure that all data mining selection criteria were represented, including producers who frequently receive payments because they claim that adverse weather conditions prevented them from planting their crop, as well as policy irregularities that suggested collusion among agents, adjusters, and producers. We also conducted two surveys. In the first, we surveyed all 829 FSA county officials responsible for conducting field inspections in 2003 to assess the effectiveness of USDA's procedures and processes to prevent and detect fraud, waste, and abuse in selling and servicing crop insurance policies. In the second, we surveyed a stratified, random sample of 935 of the approximately 13,000 crop insurance sales agents to solicit their views on control weaknesses and suggestions for improving oversight of the crop insurance program. This sample methodology allows us to project the survey results to all crop insurance agents. We received responses from 92 percent of the 829 FSA officials in the first survey and 76 percent of the 935 insurance agents in the second survey. To determine the extent to which program design issues may contribute to fraud, waste, and abuse in the crop insurance program, we conducted a qualitative assessment of economic studies. We also discussed these issues with USDA officials in headquarters and field offices. To determine the effectiveness of RMA's procedures for assuring program integrity in developing and expanding crop insurance products, we evaluated the agency's policies, procedures, and other pertinent documents to identify the controls in place to assure program integrity. We selected a nonrandom sample of 16 developmental and expansion programs between 1998 and 2002 to determine whether they complied with RMA's policies and procedures; the sample included policies with low and high claims experience to determine if loss experience was affected by compliance with procedures. To show the types of actions that producers, agents, and loss adjusters have used to circumvent RMA's procedures, our Office of Forensic Audits and Special Investigations (FSI) reviewed eight cases of crop insurance fraud prosecuted between June 2003 and April 2005. To research these cases, it reviewed USDA's Inspector General's case files, spoke with representatives from the U.S. Department of Justice, and reviewed relevant reports, court papers, and other documentation. FSI conducted its investigation from February through June 2005 in accordance with quality standards for investigations as set forth by the President's Council on Integrity and Efficiency. We conducted our review from July 2004 through August 2005, according to generally accepted government auditing standards, which included an assessment of data reliability and internal controls. Appendix I contains more detailed information on our scope and methodology. Results in Brief: Employing a broad range of processes to prevent and detect fraud, waste, and abuse in the crop insurance program, RMA has reported that questionable claims payments fell more than $300 million over the past 4 years. However, our review showed that RMA is not effectively using all of the tools it has available and that producers and others continue to take advantage of the program. In addition, program design issues, including insuring individual fields, risk-sharing provisions, and prevented planting can impede RMA's effort to ensure program integrity. Specifically: * Inspections during the growing season are not being used to maximum effect. Although FSA is assisting RMA, as required under ARPA, by conducting field inspections, FSA is not doing so in accordance with USDA guidance. Between 2001 and 2004, producers filed claims on about 380,000 policies annually, and RMA's data mining identified about 1 percent of these claims as questionable and needing FSA inspection. Under USDA guidance, FSA should have conducted all of the requested inspections. However, FSA conducted only 64 percent of the inspections RMA requested; FSA inspectors said that they did not conduct all requested inspections primarily because they did not have sufficient time. Between 2001 and 2004, FSA offices in nine states did not conduct any of the field inspections RMA requested in one or more of the years. Until we brought this matter to their attention in September 2004, FSA headquarters officials were unaware that FSA offices in these nine states had not conducted field inspections for one or more of the years. FSA may not be as effective as possible in conducting field inspections because RMA does not provide FSA with information on the nature of the suspected abusive behavior or the results of follow-up investigations. About 80 percent of the FSA inspectors we surveyed believe that receiving more information from RMA would help them be more effective in detecting fraud, waste, and abuse. In addition, FSA state officials told us that inspectors are reluctant to conduct field inspections because they believe RMA and insurance companies do not use the information to deny claims for producers who do not employ good farming practices. Finally, these inspections do not always occur in a timely fashion, which would help detect abuse during the growing season. Because of these problems, the insurance companies and RMA cannot always determine the validity of a claim. * RMA's data analysis of the largest farming operations is incomplete. As required by ARPA, RMA is using data mining to administer and enforce the crop insurance program and to analyze patterns that suggest fraudulent activity, such as unusually high or frequent claims. However, RMA's analysis excludes comparisons of the largest farming operations--including those organized as partnerships and joint ventures. RMA cannot make these comparisons because producers do not always report the individuals or entities having a beneficial interest of 10 percent or more in the farming operation holding the policy, as required. RMA's database does not identify a producer's ownership interest in other farming operations, and it has not been given access to similar data that FSA maintains. According to our review of FSA's database for 2003, 21,310 entities, or 31 percent of the entities we analyzed, did not report to RMA one or more individuals or entities who had a beneficial interest in their farming operation, as RMA regulations require. RMA should be able to recover up to $74 million in claims paid to these 21,310 entities--the amount of claims paid in proportion to the interest of the member who was not reported. Additionally, through data mining, we identified 115 of these 21,310 entities with questionable insurance claims totaling $9.2 million. Finally, we identified nine farming entities that had one or more owners who had previously been ruled ineligible to participate in the federal crop insurance program because, for example, they had not paid their insurance premium. We have referred this information to RMA for further investigation. As our analysis indicates, without access to FSA's database, RMA is missing opportunities to compare claims experience among these large operations and to identify potentially fraudulent behavior. * RMA is not effectively overseeing insurance companies' quality assurance programs. Eighty of the 120 insurance claim files we reviewed should have received a quality assurance review because, for example, they claimed more than $100,000 in crop losses. However, we found the insurance companies conducted reviews on only 59 of these claims. Furthermore, the reviews were largely paper exercises, such as computational verifications, rather than comprehensive claim analysis. RMA did not ensure that companies conducted all reviews called for under its guidance and did not examine the quality of the companies' reviews. RMA officials acknowledged that their agency's guidance for conducting quality assurance reviews needs revision to improve the compliance program. They noted RMA is working with a contractor to revise its guidance. * RMA has infrequently used its new sanction authority to address program abuses. RMA has not fully used its new authority under ARPA to sanction producers, insurance agents, and claims adjusters who willingly and intentionally provide false or inaccurate information or fail to comply with other FCIC program requirements. RMA has identified about 3,000 producers with suspicious claim payments--notable policy irregularities compared with other producers growing the same crop in the same county--each year since the enactment of ARPA. While not all of these policy irregularities were necessarily sanctionable, RMA has imposed only 114 sanctions from 2001 through 2004. RMA's referrals to USDA's Inspector General declined from a high of 37 in 2000 to 14 in 2004. According to RMA officials, RMA has requested and imposed few sanctions because it has not issued regulations to implement its expanded authority under ARPA. Without regulations, RMA has not established what constitutes an "FCIC requirement" and how it will determine that a violation has occurred or what procedural process it will follow before imposing sanctions. Insurance agents we surveyed and company officials we contacted believe that RMA needs to more aggressively seek to penalize those producers, agents, and adjusters who abuse the program. RMA officials told us that they will give priority to issuing regulations implementing the sanctions authorized under ARPA. We also found that RMA's insurance information system contains inaccurate data and does not always identify inaccurate claims payments and that RMA did not always account for changes in farming practices in a timely manner. While RMA can improve its day-to-day oversight of the federal crop insurance program in a number of ways, the program's design, as laid out in RMA's regulations or as required by statute, hinders RMA officials' efforts to administer certain program provisions to prevent fraud, waste, and abuse. Specifically: * RMA's regulations allow producers the option of insuring their fields individually rather than combined as one unit. Producers may want to insure fields separately out of concern that they would experience losses in a certain field because of localized weather conditions, such as hail or flooding. Insuring fields separately provides greater assurance that such losses will be covered. However, insuring fields separately enables producers to "switch" production among fields-- reporting production of a crop from one field that was actually produced on another field--either to make false insurance claims based on low production or to build up a higher yield history on a particular field in order to increase its eligibility for higher future insurance guarantees. Of the 2,371 producers included on RMA's list of producers with irregular claims in 2003, 12 percent were suspected of switching production among their fields. * To induce insurance companies to deliver crop insurance to all eligible producers, RMA's regulations allow the companies to place producers with frequent or high claims in an insurance fund that shifts almost all of the risk associated with these claims to the federal government. Accordingly, the companies have less incentive to rigorously challenge questionable claims. * RMA is statutorily required to offer producers "prevented planting" coverage. With this coverage, producers can file claims if they are unable to plant the crop because of an insured cause of loss, such as too much rain causing wet fields. However, as RMA and company officials told us, it is often difficult to determine whether the producer had the opportunity to plant a crop, hampering their ability to hold down fraudulent claims. * The statutorily established premium subsidies can be as high as 67 percent and, therefore, may also inhibit RMA's ability to control program abuse. High premium subsidies shield producers from the full effect of paying higher premiums associated with frequent or larger claims because the subsidies significantly reduce producers' premiums. Over one-half of the crop insurance agents responding to our survey believed that crop insurance should cost more for producers with a pattern of claims that are higher or more frequent in comparison with other producers for the same crop in the same location to discourage fraud, waste, and abuse in the program. Eight recent crop insurance fraud cases, investigated by USDA's Office of Inspector General (OIG) and resulting in criminal prosecutions between June 2003 and April 2005, reflect issues we identified. These eight cases, totaling $3.1 million in insurance claims, show how producers, sometimes in collusion with insurance agents and others, falsely claim prevented planting, weather damage, and low production. Some of the cases show producers hiding or moving production from one field to another. Several of these cases also demonstrate the importance of having FSA and RMA work together to identify and share information on questionable farming practices/activities. When developing and expanding new crop insurance products, RMA did not always follow its guidelines, which, according to USDA, were designed, among other things, to minimize exposure to loss. As a result, claims and loss ratios have been substantially higher for some new crop products. For example, for the fall-planted watermelon program launched in southern Texas, RMA approved the product without completing all data collection and reviews of data called for under its guidelines. RMA paid more than $20 million for fall-planted watermelon claims in 1999. RMA cited a shortage of experienced staff as a major factor contributing to its failure to follow its guidelines. In addition, we found that RMA did not always annually evaluate these new products, as laid out in its guidelines. In most cases, this lack of oversight did not appear to result in significantly greater losses. However, in the sweet potato program, a timely evaluation of the loss experience might have averted the payment of several million dollars in claims. According to RMA officials, they do not typically conduct annual reviews, as the guidelines state they are to do, because they believe they need several years of loss experience to adequately evaluate a new product. To better protect the crop insurance program from fraud, waste, and abuse, Congress should consider allowing RMA to reduce premium subsidies for producers who consistently have claims that are irregular in comparison with other producers growing the same crop in the same location. We are also making a number of recommendations to the Secretary of Agriculture to improve RMA's and FSA's implementation of ARPA and oversight of the crop insurance program. Among other things, we are recommending that the Secretary of Agriculture direct the Administrators of RMA and FSA to develop an action plan to improve the effectiveness of the inspections conducted during the growing season. We are also recommending that the Secretary of Agriculture direct FSA to share producer-derived information with RMA to administer and enforce requirements of the crop insurance program. Lastly, we are recommending that the Secretary of Agriculture promulgate regulations implementing the expanded authority under ARPA to impose sanctions and direct RMA to eliminate optional unit coverage for producers who use this coverage to frequently file questionable claims and receive payments. We provided a draft of this report to USDA for its review and comment. USDA agreed to act on most of our recommendations, but it disagreed with two of them. For example, USDA agreed to take steps to improve the effectiveness of its growing season inspections and to strengthen oversight of crop insurance providers' implementation of quality control reviews. It also agreed that promulgating regulations to implement the expanded authority under ARPA to impose sanctions would enhance RMA's sanctions efforts, although it did not believe that the lack of regulations has precluded it from using ARPA's authority to impose sanctions. USDA disagreed with our recommendation that FSA field offices conduct all inspections called for under agency guidance because it believes FSA does not have sufficient resources to complete all of these inspections. USDA may want to study the costs and benefits of conducting these inspections. USDA also disagreed with our recommendation to eliminate optional unit coverage for producers who received payments for questionable claims because, among other things, it did not believe eliminating such coverage would be prudent or cost- effective. However, we continue to believe that it is reasonable for USDA to use all tools at its disposal and that our recommendations will reduce the federal crop insurance program's vulnerability to fraud and abuse. Our detailed response to USDA's comments appears at the end of this letter and following USDA's written comments in appendix VII. Background: Farming is an inherently risky enterprise. In conducting their operations, producers are exposed to both production and price risks. Crop insurance is one method producers can use to protect themselves against these risks. Over the years, the federal government has played an active role in helping to mitigate the effects of these risks on farm income by promoting the use of crop insurance. Appendix II contains information on the crop insurance program from 1981 to 2004. Under the program, participating producers are assigned (1) a "normal" crop yield based on their actual production history and (2) a price for their commodity based on estimated market conditions. Producers can then select a percentage of their normal yield to be insured and a percentage of the price they wish to receive when crop losses exceed the selected loss threshold. The following example illustrates how a claim payment is determined. A producer whose normal crop production averages 100 bushels of corn per acre and who chooses to buy insurance at the 75 percent coverage level will be guaranteed 75 percent of 100 bushels, or 75 bushels per acre. Assuming that the producer had chosen the maximum price coverage and that RMA had estimated the market price for corn at $2 per bushel, the producer would have total coverage of $150 per acre. Should something like drought cut the producer's actual harvest to 25 bushels, the producer will be paid for the loss of 50 bushels per acre--the difference between the insured production level of 75 bushels and the actual production of 25 bushels. The insurance would pay the producer's claim at $2 x 50 bushels, or $100. In addition, under the crop insurance program's "prevented planting" provision, insurance companies pay producers who were unable to plant the insured crop because of an insured cause of loss that is general in their surrounding area, such as weather conditions causing wet fields, and that prevents other producers from planting acreages with similar characteristics. These producers are entitled to claim payments that generally range from 50 to 70 percent of the coverage they purchased, depending on the crop. Critical to the success of the crop insurance program is aligning the premium rates with the risk each producer represents. The risk associated with growing a particular crop varies from location to location, from farm to farm, and from producer to producer. If the rates are too high for the risk represented, producers are less likely to purchase insurance, lowering the program's income from premiums. Conversely, if the rates are too low, producers are more likely to purchase crop insurance, but because the rates are too low, the income from premiums will be insufficient to cover the claims. Economists refer to this situation as adverse selection. To align crop insurance premium rates with the risk represented, RMA establishes rates that vary by crop, location (county), farm, and producer. RMA's objective is to set the rates that each producer pays according to the risk associated with the producer's location, crop, and past production. For the major field crops, RMA begins its premium rate-setting process by looking at past crop insurance experience for each county and state. On the basis of that historical experience, RMA sets a premium rate for each crop in each county at the 65 percent coverage level for average production. Using this premium rate, RMA makes adjustments to establish rates for other coverage levels. RMA also adjusts premium rates to assume producers will insure individual fields, called "optional units," rather than all fields combined, called "basic units."[Footnote 3] RMA uses an algorithm to make adjustments to establish premium rates for producers whose production levels are higher or lower than the county's average. According to RMA, this latter adjustment is based on the assumption that producers with higher-than-average production levels are less likely to experience losses. Finally, to encourage participation in the crop insurance program, the federal government subsidizes the premiums. Moreover, for producers that do not have a sufficient number of years- -at least 4--of actual production history records, RMA uses the historical average county yield (called a transitional yield), adjusted by a factor based on the number of years for which the producers have provided records. Producers may also substitute the transitional yield for actual yields in disaster years. In general, RMA sets a floor under a producer's annual yield so that a yield in any year cannot fall below 60 percent of the transitional yield for that crop. RMA establishes the terms and conditions that the private insurance companies selling and servicing crop insurance policies are to use through a contract called the standard reinsurance agreement (SRA). The SRA is a cooperative financial assistance agreement between RMA, through the FCIC, and the private crop insurance companies to deliver federal crop insurance under the authority of the Federal Crop Insurance Act. The SRA establishes the minimum training, quality control review procedures, and performance standards required of all insurance providers in delivering any policy insured or reinsured under the Federal Crop Insurance Act, as amended. For example, under the SRA, companies must provide training to their sales agents that includes information on how to recognize common indicators of misrepresentation or abuse, review anomalies identified by FCIC that suggest an unusual claims pattern, and report all cases of suspected misrepresentation, fraud, waste, or abuse. To distinguish among different levels of risk, the SRA establishes three reinsurance funds with commensurate requirements for the amount of risk companies can cede back to FCIC: assigned risk, developmental, and commercial. FCIC created the assigned risk fund for the riskiest policies. Under the SRA, insurance companies may include individual policies in this fund up to limits established for each state. Beginning in 2005, the maximum amount of premium and associated liability for claims payments that can be allocated to the assigned risk fund varies from 25 percent in some states (e.g., Illinois, Indiana, and Iowa) to 75 percent in others (e.g., Mississippi, North Dakota, and Texas). Companies must retain 15 to 25 percent of the policies' premiums and associated liability for claims payments for policies in this fund, depending on the state. RMA is responsible for ensuring that the federal crop insurance program is carried out efficiently and effectively and for protecting against fraud, waste, and abuse in the program. In this regard, RMA uses a broad range of tools, including compliance reviews, company quality assurance reviews, data mining, and FSA field inspections. RMA has a compliance staff of 78 employees in six field locations to review company quality assurance activities and investigate anomalous claims payments. For their part, insurance companies must conduct quality assurance reviews, such as program or field reviews, for policies with a claim RMA has identified as anomalous and policies with a claim equal to or greater than $100,000; these reviews are to determine whether the claims they have paid are in compliance with policy provisions. In 2004, RMA initiated a new operational review program that provides for extensive review of each insurance provider's operation every 3 years. RMA's 3-year review cycle will assess insurance providers' adherence to their contract with RMA, quality control guidelines, and RMA-approved policies and procedures. This review will differ from prior reviews in that RMA will direct the companies to investigate policies that RMA has identified as having anomalous claims and require RMA to assess a statistical sample of additional policies. In the past, the insurance companies reviewed a statistical sample of claims and policies, and RMA examined the results of the companies' reviews. To strengthen oversight at the local level, RMA conducts data mining and uses past loss experience to develop a sample of producers with notable policy irregularities, such as unusually high or frequent losses. Staff in FSA county offices review these cases for potential fraud, waste, and abuse by inspecting the fields of the producers on the list. Figure 1 shows the location of producers RMA identified for field inspections in 2003. Figure 1: Location of Producers Identified by RMA for Field Inspections, 2003: [See PDF for image] [End of figure] Congress enacted ARPA, amending the Federal Crop Insurance Act, in part, to improve compliance with, and the integrity of, the crop insurance program. ARPA expanded RMA's authority to impose sanctions in two ways. First, it provided RMA authority to impose sanctions against producers, agents, loss adjusters, and insurance companies that willfully and intentionally provide false or inaccurate information to FCIC or to an approved insurance provider. (Previously, RMA had authority to impose sanctions only on individuals who willfully and intentionally provided false information.) Second, ARPA provided authority to impose sanctions against producers, agents, loss adjusters, and insurance companies for willfully and intentionally failing to comply with any other FCIC requirement. RMA has the authority to disqualify producers who have committed a violation not only from the insurance program but also from most other farm programs for up to 5 years. RMA can also impose a civil fine for each violation, up to the financial gain the individual obtained as a result of the false or inaccurate information provided or of the noncompliance, or $10,000, whichever is greater. Working with RMA's regional compliance offices, RMA's sanctions office processes requests for sanctions from the field offices and forwards the findings and recommendations to RMA's appeals and litigation office. Following this office's review, USDA's Office of General Counsel provides a legal opinion on the sanction request. After consulting with the Office of General Counsel, if the Administrator of RMA considers the case valid, RMA files a complaint with USDA's Administrative Law Office. At the defendant's request, the Administrative Law Office will hold a hearing, after which the administrative law judge will render a decision. ARPA also increased the percentage share of the premium the government pays for most coverage levels of crop insurance, beginning with the 2001 crop year. Although the percentage of the premium the government pays declines as producers select higher levels of coverage, the government contribution significantly increases for all levels of coverage, particularly for the highest levels of coverage. For example, as shown in table 1, the share of the premium paid by the government rose from 42 to 59 percent of the premium for 65 percent coverage.[Footnote 4] Table 1: Premium Subsidies Before and After ARPA: Percentage of coverage selected by producer: 50; Percentage of premium paid by the government: Before ARPA[A]: 55; Percentage of premium paid by the government: After ARPA: 67. Percentage of coverage selected by producer: 55; Percentage of premium paid by the government: Before ARPA[A]: 46; Percentage of premium paid by the government: After ARPA: 64. Percentage of coverage selected by producer: 60; Percentage of premium paid by the government: Before ARPA[A]: 38; Percentage of premium paid by the government: After ARPA: 64. Percentage of coverage selected by producer: 65; Percentage of premium paid by the government: Before ARPA[A]: 42; Percentage of premium paid by the government: After ARPA: 59. Percentage of coverage selected by producer: 70; Percentage of premium paid by the government: Before ARPA[A]: 32; Percentage of premium paid by the government: After ARPA: 59. Percentage of coverage selected by producer: 75; Percentage of premium paid by the government: Before ARPA[A]: 24; Percentage of premium paid by the government: After ARPA: 55. Percentage of coverage selected by producer: 80; Percentage of premium paid by the government: Before ARPA[A]: 17; Percentage of premium paid by the government: After ARPA: 48. Percentage of coverage selected by producer: 85; Percentage of premium paid by the government: Before ARPA[A]: 13; Percentage of premium paid by the government: After ARPA: 38. Source: RMA. [A] For crop years 1999 and 2000, the actual premium subsidy was higher than shown. Under emergency supplemental acts, producers received an additional 30 percent discount in 1999 and 25 percent discount in 2000. [End of table] RMA Has Strengthened Procedures for Preventing Questionable Claims, but the Program Remains Vulnerable to Abuse: Since ARPA, RMA has taken a number of steps to improve its procedures and processes to prevent and detect fraud, waste, and abuse in selling and servicing crop insurance policies. Most notably, RMA reports that data mining analyses and subsequent communication to producers resulted in a decline of at least $300 million in questionable claims payments from 2001 to 2004. However, we found that RMA is not effectively using all of the tools it has available and that producers and others can continue to take advantage of the program. We identified weaknesses in four key areas: (1) field inspections, (2) analysis that excludes many large farming operations when producers do not report their interest in them, (3) quality assurance reviews, and (4) imposition of sanctions. Weaknesses in these areas continue to leave the program vulnerable to questionable claims, and insurance companies and RMA cannot always determine the validity of a claim to minimize fraud, waste, and abuse. We also found that RMA's insurance information system does not always identify policies that fail to comply with policy provisions and that RMA's implementation approach may not always respond to unanticipated vulnerabilities in a timely manner. Data Mining and Other Actions Have Improved RMA's Ability to Manage the Crop Insurance Program: Each year, RMA develops a list of producers whose operations warrant an on-site inspection (the spot-check list) during the growing season because data mining uncovered patterns in their claims that are consistent with the potential for fraud and abuse. For example, the list includes: * producers, agents, and adjusters linked in irregular behavior that suggests collusion; * producers who for several consecutive years received most of their crop insurance payments from prevented planting indemnity payments; * producers who appear to have claimed the production amounts for multiple fields as only one field's yield, thereby creating an artificial loss on their other field(s); and: * producers who, in comparison with their peers, have excessive harvested losses over many years. Since RMA began using data mining in 2001, it has identified about 3,000 producers annually who warrant an on-site inspection because of anomalous claims patterns. In addition, RMA annually performs about 100 special analyses to identify areas of potential vulnerability and trends in the program. RMA provides the list of producers from its spot-check list to the appropriate FSA state offices for distribution to FSA county offices, as well as to the insurance company selling the policy to the producer. Staff in FSA county offices advise the selected producers that they have been identified for an inspection as a result of data mining and conduct field inspections during the growing season. In conducting these inspections, inspectors are to determine the tillage method used; weed control practices; type and amount of fertilizer applied; weather conditions; and how the inspected crop compares with others in the area. As a result of these inspections and other information, RMA reported total cost savings of $312 million, primarily in the form of estimated payments avoided: $48 million in 2001, $112 million in 2002, $81 million in 2003, and $71 million in 2004. For example, according to RMA, claims payments to producers identified for an inspection decreased nationwide from $234 million in 2001 to $122 million in 2002. According to RMA, some of the producers on the list bought less insurance and a few dropped crop insurance entirely, but most simply changed their behavior regarding loss claims. Field Inspections Specified in RMA's Coordination Plan Are Not Being Used to Maximum Effect: ARPA required USDA to have a plan for FSA to assist RMA and approved insurance providers in auditing a statistically appropriate number of crop insurance claims. Under USDA guidance, developed pursuant to this requirement, RMA is to annually provide a list of producers who exhibit high loss ratios and high frequency and severity of losses or who are suspected of poor farming practices. Upon receipt of this list, the FSA county office is to review the first 10 producers or the top 5 percent of the producers on the list, whichever is larger. If less than 10 producers are on the list, then FSA is to check all of them.[Footnote 5] All the lists that RMA has provided to FSA county offices include 10 or fewer producers, but FSA is not conducting field inspections for all producers on the list. Between 2001 and 2004, producers filed about 380,000 claims annually. RMA's data mining identified about 1 percent of these claims as questionable and needing inspection. Overall, FSA conducted only 64 percent of the inspections RMA requested from 2001 to 2004. Specifically, FSA submitted inspection reports for only 70 percent of the inspections RMA requested in 2001 (1,737 requested), 49 percent in 2002 (3,303 requested), 67 percent in 2003 (3,094 requested), and 73 percent in 2004 (3,832 requested). During this period, FSA offices in nine states failed to conduct any of the field inspections RMA had requested in one or more of the years. Until we brought this matter to their attention in September 2004, FSA headquarters officials were unaware that these nine states had not conducted field inspections for one or more of the years. According to FSA officials in five states we contacted, county directors are reluctant to conduct field inspections because they believe RMA and insurance companies do not use the information to deny claims for producers who do not employ good farming practices. As such, they believe it does not make sense for them to spend time conducting these reviews. However, by not conducting all requested inspections, FSA is missing opportunities to identify producers who file unwarranted claims. For their part, FSA inspectors believe they would be more effective in determining fraud, waste, and abuse if they received information from RMA on the claims patterns RMA's data mining has identified as questionable. For example, of the 3,832 claims RMA identified for field inspections in 2004, approximately two-thirds were selected for anomalous claims patterns associated with fraud, such as switching information on production yields from one insured field to another. About 80 percent of the FSA inspectors we surveyed believed that receiving more information from RMA would help them be more effective in detecting fraud, waste, and abuse when they conduct field inspections. (See app. III for a summary of the results of our survey of FSA inspectors.) Additionally, several FSA inspectors surveyed provided written comments regarding the need for feedback. As one respondent noted, there is little incentive to document field inspection findings because FSA rarely learns what, if any, action was taken. Another respondent commented that he would like feedback from RMA on how useful the inspections have been. He would like to avoid spending time on inspections that may not be useful to RMA. RMA headquarters officials acknowledged that providing feedback to FSA inspectors might help improve the quality of the field inspections. Similarly, company officials told us that information from RMA's data mining would help claims adjusters pay particular attention to determining the total production for the producer's farming operation and differences between fields with and without losses. Although FSA inspectors cited a lack of communication with RMA on specific cases and findings as a major impediment to completing inspections, they also identified other reasons. As figure 2 shows, the most commonly cited reason was "not having enough time." Figure 2: FSA Inspectors' Primary Reasons for Not Conducting Field Inspections of Producers with Notable Policy Irregularities: [See PDF for image] [End of figure] We discussed the reported lack of time with FSA headquarters officials, who advised us that field offices' broad range of responsibilities provide limited time for field inspections in support of RMA. They said our survey results taken as a whole underscore the importance of effective communication and information sharing between RMA and FSA to maximize the effectiveness of field inspectors' work. FSA's field inspections also do not always occur in a timely manner and, therefore, FSA inspectors may miss opportunities to detect abuse during the growing season. RMA generally provides its spot-check list to FSA in April, at the start of the growing season. USDA guidance directs FSA staff to perform at least two field inspections--one within 30 days of the final planting date and one before harvest--on a minimum of one representative tract. FSA selects a representative tract for each crop listed by RMA on the spot-check list.[Footnote 6] However, about 17 percent of FSA inspectors reported that they received RMA's request for a field inspection more than 30 days after the final planting date. In some cases, inspection requests came in as much as 6 months later. Additionally, insurance companies may receive the results of some field inspections too late to determine the validity of the claim. After FSA county offices conduct the field inspections, they report the findings to RMA, which then provides the results to the insurance companies holding the policies for the producers. According to company officials, they are unable to use the results of some field inspections because the information is received months after the claim was paid. For example, in one claim file we reviewed, on November 24, 2003, RMA referred to an insurance company a soybean producer in Ohio who had received claims payments in each of the past 5 years and was suspected of underreporting his production in 2003. FSA's field inspection, conducted in September just prior to harvest, found the crop to be "above average to average" for the county and did not identify any concerns regarding the crop's expected yield. To determine whether the producer underreported production, the insurance company needed to conduct a preharvest appraisal of the producer's fields. While the insurance company conducted a quality assurance review of the claim, it received RMA's reports after the producer harvested the soybeans--too late to conduct preharvest appraisals to validate production.[Footnote 7] FSA may also be missing opportunities to provide RMA with critical information to assess a claim's validity. In reviewing claims, we found that FSA frequently inspects only one tract, but that tract was not always the tract on which a claim was filed. In written comments on our survey, several FSA inspectors reported that they believe conducting a growing season inspection on more than one tract is necessary to ensure the monitoring program is effective. However, these inspectors also noted that conducting inspections on more than one tract of land would place additional demands on their time. RMA's Analysis to Detect Potential Program Fraud and Abuse for Many Large Farming Operations Is Incomplete: RMA's data mining excludes many large farming operations because producers fail to report other individuals' and entities' interests in these operations. However, these entities, such as partnerships and corporations, may include individuals who are also members of one or more other entities. Because it does not know the ownership interests in these farming operations, RMA cannot readily identify potential fraud. For example, producers who are members of more than one farming operation may have the opportunity to move production from one operation to another to file unwarranted claims, without RMA's knowledge that these producers participate in more than one farming operation. These farming operations do not always report other individuals or entities who hold or acquire a beneficial interest of 10 percent or more in the insured operation, as required by RMA regulations. RMA was unaware that these entities had failed to fully disclose ownership interest because it has not been given access to the FSA data file identifying a producer's ownership interest in other farming operations. However, ARPA requires the Secretary of Agriculture to develop and implement a coordinated plan for RMA and FSA to reconcile all relevant information received by either agency from a producer who obtains crop insurance coverage. The Secretary of Agriculture also must require RMA and FSA to reconcile this producer-derived information on at least an annual basis, starting with the 2001 crop year, to identify and address any discrepancies. We were able to obtain the FSA data file and determine whether (1) farming operations report all members who have a substantial beneficial interest in the operation, (2) these farming operations file questionable crop insurance claims, and (3) agents or claims adjusters had financial interests in the claim.[Footnote 8] As shown in table 2, of the 69,184 entities that had crop insurance policies in 2003 and that were in both RMA's and FSA's databases, 21,310, or 30.8 percent, did not report one or more members who held a beneficial interest of 10 percent or more in the farming operation holding the policy. Table 2: Crop Insurance Policyholders Failing to Disclose Ownership Interest, by Entity Type, Crop Year 2003: Entity type of policyholder: Corporation; Number of entities analyzed[A]: 38,463; Number of entities failing to disclose ownership interest[B]: 12,130; Percentage: 31.5%. Entity type of policyholder: General partnership; Number of entities analyzed[A]: 24,780; Number of entities failing to disclose ownership interest[B]: 7,486; Percentage: 30.2%. Entity type of policyholder: Limited partnership; Number of entities analyzed[A]: 4,401; Number of entities failing to disclose ownership interest[B]: 1,479; Percentage: 33.6%. Entity type of policyholder: Sole proprietorship[C]; Number of entities analyzed[A]: 1,540; Number of entities failing to disclose ownership interest[B]: 215; Percentage: 14.0%. Total; Number of entities analyzed[A]: 69,184; Number of entities failing to disclose ownership interest[B]: 21,310[D]; Percentage: 30.8%. Sources: GAO analysis of RMA and FSA data. [A] We excluded trusts and joint ventures from the analysis because RMA and FSA use conflicting definitions. We then identified 69,184 entities in both the FSA and RMA files. FSA's database for ownership in entities contained 345,421 entities for 2003, and the RMA database contained 112,467 entities that could have one or more members holding a beneficial interest of 10 percent or more. [B] Entities and members in the RMA database were compared against the FSA database. If the entity or any member that held a beneficial interest of 10 percent or more as reported in the FSA database did not match the RMA database, the policy was identified as an entity failing to disclose ownership interest. [C] Sole proprietors operate farming entities using an employer tax identification number and may conduct business under an assumed name. [D] Of the 21,310 entities failing to disclose ownership interest, 5,848 entities had members with tax identification numbers that differed by one digit in the RMA and FSA databases. [End of table] RMA should be able to recover a portion of the $224.8 million in claims paid to the 21,310 entities that failed to disclose the ownership interest of one or more members in 2003. According to RMA regulations, if the policyholder fails to disclose the ownership interest in the farming operation as required, the policyholder must repay the amount of the claims payment that is proportionate to the interest of the person who was not disclosed.[Footnote 9] The average ownership interest of the persons not disclosed for the 21,310 entities was 33 percent; as a result, RMA should be able to recover up to $74 million in claims payments. According to our analysis of RMA's and FSA's databases, results were similar for 2004--20,659 entities failed to disclose the ownership interest of one or more members. As a result, RMA should be able to recover up to $70 million in claims payments. In addition, we identified 24 crop insurance agents who sold policies to farming entities in which the agents held a substantial beneficial interest but failed to report their ownership interest to RMA as required.[Footnote 10] These farming entities received $978,912 in claims payments in 2003 and 2004. RMA regulations require that, if a person who is not reported is also ineligible to participate in the crop insurance program, the crop insurance policy is void, and the policyholder must repay the entire claims payment. For example, a person can be ineligible because of delinquent debt, such as unpaid premiums, to RMA or insurance companies. For 2003 and 2004, using FSA's data, we found that nine farming operations contained one or more members participating in the crop insurance program who RMA had determined were ineligible to participate. If RMA had complete information on entity ownership interests, it could strengthen the review of some of the largest claims. For example, in analyzing the 21,310 entities failing to disclose ownership interest in 2003, we found 210 entities with questionable insurance claims totaling $11.1 million on 244 policies. Furthermore, we identified one claims adjuster who adjusted a policy in 2004 with claims payments of $91,094 for a farming operation in which he held a beneficial interest of 33 percent.[Footnote 11] RMA guidance prohibits conflict-of-interest activities. Among other things, insurance providers are not to permit adjusters to adjust a claim of a party in which the adjuster has a material or financial interest.[Footnote 12] Without FSA's entity data, RMA is missing opportunities to identify potentially fraudulent behavior in these operations. Furthermore, an internal RMA study found that entities that purchase crop insurance for only 1 or 2 years have higher claims experience than entities that participate continuously over a number of years.[Footnote 13] According to FSA officials in two states we contacted, some entities are apparently created temporarily to avoid tracking by RMA, making it difficult for RMA to identify questionable claims patterns over time. They told us that after these entities participate in the crop insurance program for a few years, they are dissolved, and the farming operations are reestablished under new entity names. RMA Cannot Effectively Assess Insurance Companies' Performance Because of Weaknesses in Quality Assurance Reviews: RMA also looks to insurance companies that are selling and servicing crop insurance to help them ensure program compliance and minimize losses. RMA guidance states that insurance providers will provide oversight to properly underwrite the federal crop insurance program, including implementing a quality control program, conducting quality control reviews, and submitting an annual report to FCIC. However, RMA is not effectively overseeing insurance companies' quality assurance programs and, for the claims we reviewed, it does not appear that most companies are rigorously carrying out their quality assurance functions. For example, 80 of the 120 insurance claim files we reviewed claimed more than $100,000 in crop losses or met some other significant criteria; RMA's guidance states that the insurance provider must conduct a quality assurance review for such claims. However, the insurance companies conducted reviews on only 59 of these claims, and the reviews were largely paper exercises, such as computational verifications, rather than comprehensive analysis of the claim. In 2002, USDA's OIG reported that RMA's efforts to develop a quality control review system had been rendered ineffective by the absence of a policy establishing what the system should measure and what standards of accountability should apply.[Footnote 14] The Inspector General noted that RMA had not (1) determined whether it should measure each insurance company's performance, (2) established an acceptable standard error rate to hold companies accountable for excessive errors, and (3) defined an error so that error rates or improper payment measurements were meaningful. As a result, the Inspector General stated, RMA is no closer to having a fully developed and reliable quality control review system to evaluate the delivery of the federal crop insurance program than it was in 1993, when the Inspector General recommended that RMA develop and implement such a system. Similarly, in 1999, we recommended that RMA improve its methodology for estimating error rates for claims payments.[Footnote 15] We reported that such information is essential for evaluating the crop insurance program's effectiveness over time and for providing controls over claims payments. RMA's own most recent internal review reached a similar conclusion. In September 2002, RMA's Deputy Administrator for Compliance reported that RMA needed to significantly revise its guidance to accomplish meaningful quality assurance reviews with measurable results. This conclusion was based on a review of 17 insurance providers' compliance with FCIC's quality assurance requirements. For example, the Deputy Administrator noted, RMA's guidance did not define the type and amount of documentation needed to meet review requirements and to support the insurance companies' review results and conclusions. Furthermore, because the insurance companies relied heavily on the use of check sheets to document and report the results of their reviews, rather than inspecting fields, RMA could not confirm that quality assurance reviews were performed as required. According to RMA officials, RMA is working with a contractor to incorporate the report's recommendations and revise its guidance. As of August 2005, RMA had not issued revised guidance on the companies' conduct of quality assurance reviews. RMA May Be Missing Opportunities to Impose Sanctions Because It Has Not Developed Regulations Implementing its Expanded Authority to Impose Sanctions under ARPA: While ARPA expanded RMA's authority to impose sanctions on producers, agents, and adjusters who abuse the crop insurance program, RMA has only used this authority on a limited basis. RMA has imposed sanctions on individuals who have provided false or inaccurate information, but it has not used its new authority to impose sanctions on individuals who willfully and intentionally fail to comply with FCIC requirements. Under ARPA, RMA has authority to impose sanctions on agents, loss adjusters, approved insurance providers, and others who willfully and intentionally (1) provide false or inaccurate information or (2) fail to comply with other FCIC requirements. Earlier legislation allowed RMA to impose sanctions only on individuals who willfully and intentionally provided false information. ARPA provides RMA with the authority to disqualify producers who have committed a material violation from receiving benefits under the insurance program and from most other farm programs for up to 5 years. Previously, RMA had authority to disqualify producers from purchasing catastrophic risk protection or receiving noninsured assistance for up to 2 years and from receiving any other benefit under the crop insurance program for up to 10 years. The new legislation also provides RMA with greater flexibility to impose civil fines. ARPA expanded RMA's authority to impose sanctions in order to improve compliance with, and the integrity of, the crop insurance program. However, as table 3 shows, except for 2004, RMA imposed few sanctions even though it has identified about 3,000 suspicious claim payments each year since 2001. From 2001 to 2004, RMA imposed 114 sanctions. Table 3: RMA Sanctions Requested and Imposed, Crop Years 2001 to 2005: Action: Requests for sanctions; 1996: [B]; 1997: 22; 1998: 27; 1999: 27; 2000: 16; 2001: 15; 2002: 83; 2003: 56; 2004: 81; 2005[A]: 21. Action: Sanctions imposed; 1996: 8; 1997: 16; 1998: 28; 1999: 8; 2000: 10; 2001: 9; 2002: 19; 2003: 19; 2004: 67; 2005[A]: 14. Source: RMA. Note: Sanctions requested and imposed include civil fines, disqualifications, debarments, and suspensions. A civil fine may be imposed against a producer, agent, loss adjuster, an approved insurance company, or other person that willfully and intentionally provides any false or inaccurate information to RMA or to an approved insurance provider with respect to a policy or plan of insurance or willfully and intentionally fails to comply with an RMA requirement. The fine may be imposed for each violation in an amount not to exceed the greater of $10,000 or the amount of financial gain obtained as a result of the false or inaccurate information or the noncompliance. In the case of a violation committed by an agent, loss adjuster, an approved insurance company, or other person (other than a producer), the violator may be disqualified for up to 5 years from participating in the USDA crop insurance program. In the case of a violation committed by a producer, the producer may be disqualified for up to 5 years from receiving any monetary or nonmonetary benefit under both the crop insurance program and other farm programs, such as price supports. [A] Data as of July 2005. [B] Data not available. [End of table] According to RMA officials, RMA's ability to impose sanctions is limited because it has not developed regulations to implement its new authority under ARPA to impose sanctions on individuals who willfully and intentionally fail to comply with an FCIC requirement. RMA's sanctions office submitted draft regulations to USDA's Office of General Counsel in 2001 and again in 2003. However, the Office of General Counsel has not approved the draft regulations. RMA headquarters officials we spoke with in April 2005 told us that the number of sanctions has not substantially increased because regulations have not been promulgated to establish what constitutes an FCIC requirement and how USDA will determine that a material violation has occurred or what process would be followed before imposing sanctions. RMA officials told us that they will give priority to issuing regulations implementing the sanctions authorized under ARPA. Furthermore, since ARPA, the number of RMA referrals to USDA's OIG has declined from a high of 37 in 2000 to 14 in 2004. Crop insurance investigations opened by OIG have declined from 40 in 2000 to 12 in 2004, as shown in table 4. The table also shows the number of convictions, on average, is less than 10 per year. Table 4: Number of USDA OIG Crop Insurance Investigations, Number of Referrals to the Department of Justice, and Case Disposition, Fiscal Years 1996 to 2005: Dollars in millions. RMA referrals to USDA's OIG; 1996: 12; 1997: 20; 1998: 12; 1999: 18; 2000: 37; 2001: 28; 2002: 16; 2003: 14; 2004: 14; 2005[A]: 8. OIG investigations opened; 1996: 27; 1997: 19; 1998: 13; 1999: 24; 2000: 40; 2001: 18; 2002: 16; 2003: 8; 2004: 12; 2005[A]: 11. Disposition of OIG referrals to the Department of Justice: Referred; 1996: 1; 1997: 6; 1998: 3; 1999: 8; 2000: 3; 2001: 28; 2002: 13; 2003: 14; 2004: 7; 2005[A]: 5. Accepted; 1996: 1; 1997: 6; 1998: 3; 1999: 7; 2000: 2; 2001: 14; 2002: 7; 2003: 2; 2004: 1; 2005[A]: 3. Declined; 1996: 0; 1997: 0; 1998: 0; 1999: 1; 2000: 1; 2001: 14; 2002: 6; 2003: 10; 2004: 3; 2005[A]: 1. Pending; 1996: 0; 1997: 0; 1998: 0; 1999: 0; 2000: 0; 2001: 0; 2002: 0; 2003: 2; 2004: 3; 2005[A]: 1. Department of Justice disposition: Indictments; 1996: 10; 1997: 6; 1998: 12; 1999: 2; 2000: 11; 2001: 13; 2002: 6; 2003: 15; 2004: 15; 2005[A]: 2. Convictions; 1996: 9; 1997: 11; 1998: 2; 1999: 6; 2000: 4; 2001: 5; 2002: 14; 2003: 8; 2004: 9; 2005[A]: 6. Dollar impact[B]; 1996: $1.4; 1997: $1.7; 1998: $0.1; 1999: $1.9; 2000: $2.0; 2001: $14.0; 2002: $1.9; 2003: $0.7; 2004: $1.7; 2005[A]: $9.7. Source: USDA's OIG. [A] Data as of August 2005. [B] Includes recoveries/collections, restitutions, fines, claims established to demand repayment of USDA benefits, and cost avoidance. [End of table] As table 4 also shows, while the number of referrals to the Department of Justice has increased, the Department of Justice has declined more cases than it has accepted since 2000.[Footnote 16] According to Department of Justice officials, the factors considered when accepting a case include sufficiency of the evidence, complexity of the case, whether the fraudulent activity is part of a pattern or scheme, and workload and resources that would be needed to investigate and prosecute the case. These officials told us that crop insurance fraud cases are highly complex and involve a significant number of documents that must be reviewed and presented in court. Furthermore, the dollar value of crop insurance cases frequently is not as large as in other cases, such as drug trafficking or some white-collar crimes. Finally, the officials noted, some cases require a full-time auditor to guide the prosecutors in reviewing the insurance and financial documents to facilitate presentation to the jury in the trial. Insurance agents we surveyed and company officials we contacted believed that RMA needs to more aggressively seek to penalize those producers, agents, and adjusters that abuse the program. (See app. IV for a summary of the results of our survey of crop insurance agents.) Other Weaknesses Affect the Crop Insurance Program's Vulnerability: We found two other weaknesses in the crop insurance program that leave it vulnerable to abuse. First, while RMA has made some improvements to verify data in its information system, the system still contains inaccurate data and does not always identify inaccurate claims payments. Consequently, RMA has a greater risk of accepting policies that have erroneous information and of paying for excessive losses. Second, production yields can change when producers change farming practices, but RMA may not also respond promptly to the resulting change in yields, which can lead to excessive claims payments. RMA's Insurance Information System Contains Inaccurate Data and Does Not Always Identify Inaccurate Claims Payments: RMA uses its insurance information system to reduce its vulnerability to fraud and abuse. Among other things, this system is to provide a means of validating data to ensure that reimbursements are made on accurate information. OMB guidance states that financial management systems shall be designed with consistent internal controls over data entry, transaction processing, and reporting to ensure that information is valid and that federal resources are protected.[Footnote 17] Without proper controls, an agency risks the possibility of processing irregularities. RMA has made improvements in its verification checks to try to ensure accurate information, but some weaknesses remain. Even though RMA is aware of the need for accurate data, we found that, at times, RMA's insurance information system contained inaccurate data. The system contained inaccuracies because RMA had not established adequate verification checks in making annual adjustments to reflect changes to the crop insurance program. Each year RMA's program automation group reviews system requirements for needed system changes in response to annual program and policy changes. In addition, the group seeks input on needed improvements, based on prior years' problems, from RMA program users and insurance company representatives. This process has been helpful in improving the overall accuracy of the data in the system. For example, RMA has made the following changes to its insurance information system since 1999: * In 1999, RMA implemented a verification check to identify policies that are on the same acreage but have two different insurance providers. * In 2001, RMA began weekly automated reporting on producers with duplicate policies. * In 2002, RMA implemented a verification check to (1) identify producers with unrealistic crop yield reports and (2) ensure that crop yield would be verified when producers changed to a new insurance provider. * In 2003, RMA implemented a verification check to validate producers' claims that they were new participants in the crop insurance program.[Footnote 18] * In 2004, RMA implemented a verification check to identify and eliminate duplicate policies for the same producer with more than one insurance provider. Each of these improvements addressed a specific information system weakness that had been identified in prior years, and each improvement reduced the likelihood of improper crop insurance payments. Nevertheless, we found that certain insurance policies, called written agreements--unique policies RMA regional offices develop to meet a local producer's specific needs--would bypass all of the verification checks that other policies undergo. Policy information from the written agreements is provided to the insurance companies, but not all the specific policy data are entered into RMA's information system. In fact, we found that some of the policies had extremely low insurance premium rates, resulting in understated premiums. For example, a policy we reviewed showed that the total premium was $1,555 for liability coverage of about $520,000, but the correct total premium should have been $155,473. For crop years 2003 and 2004, RMA had 8,511 written agreement policies with a total liability of over $400 million in its insurance information system. Because these types of policies bypass all system checks, other errors could occur. After we advised RMA of this problem, it reported that it changed the information system to check for unusually low premium rates. However, in order to conduct all the necessary verification checks on written policies, RMA will have to conduct time-consuming coordination efforts with its regional offices and the program automation group. We also found, in cases of a partial loss, claims payments were made that were higher than a specific unit's insurance liability. RMA officials stated that the insurance information system contains an edit check to ensure that the total claim is not greater than the total liability. However, we found that the system did not have an edit check to ensure that, in cases of a partial loss, claims paid for each insured optional unit were not higher than the total liability for those fields. When we reported this issue to RMA, it said that it would modify its system. However, due to a number of complexities associated with this change, RMA said that the change would not be implemented until the 2006 crop year. In addition, for 2001 to 2004, we found 14 producers enrolled in the crop insurance program who RMA had determined were ineligible to participate in the program. RMA officials stated that the insurance information system contains an edit check to identify producers determined ineligible to participate in the crop insurance program. Nevertheless, our analysis found that RMA's system does not identify all producers ineligible to participate in the crop insurance program. These ineligible producers received about $145,000 in claims payments. RMA Did Not Always Account for Changes in Farming Practices in a Timely Manner: According to a 2003 RMA study, RMA overpaid claims between 2000 and 2002 in wheat-producing counties in Oregon and Washington because of a program vulnerability. Overpayments occurred because RMA did not begin reducing producers' relatively high insurance guarantees to take into consideration a change in farming practices that began in 1996. This change resulted in lower yields on insured fields that had a higher yield history and insurance guarantee. RMA began to take this change into account in farming practices with the 2004 crop year, but it does not expect to fully resolve this issue until about 2014. RMA officials told us that if they were to fully adjust producers' insurance guarantees to reflect the lower yields in just a year, the agency would still be legally obligated to provide the higher guarantee because guarantees are based on a 10-year historical average. Under the Federal Crop Insurance Act, as amended, RMA is to provide yield coverage based on the actual production history of the crop over at least the past 4 years, building up to the previous 10-year period. Before 1996, producers could insure their wheat crop for a higher yield if they agreed to allow insured fields to lie fallow for 1 to 2 years between plantings, a practice called "summer fallow," rather than plant these fields every year (continuous cropping). This practice is used in semiarid regions, primarily to conserve moisture for the next season. By not planting, producers could allow the soil to recover moisture and, it is expected, produce a higher yield when the field is later planted. Until 1996, RMA knew which practices producers followed. However, in 1996, USDA's National Agricultural Statistics Service (NASS) changed the way it reported information on producers' farming practices.[Footnote 19] NASS had been collecting and reporting county data on wheat yields by whether producers allowed their fields to lie fallow in alternate years or planted them every year. In 1996, when NASS stopped reporting yield data by type of production practice, RMA stopped distinguishing between producers' production practices. RMA allowed producers to continue to insure their wheat at the higher yield level associated with summer fallow practices, whether or not the producers periodically let fields lie fallow or planted them every year. Under the Federal Crop Insurance Act, producers are assigned a yield based on production records. Between 1995 and 2000, many wheat producers in Oregon and Washington shifted their farming practices to planting fields every year while using the higher summer fallow production records to establish their insured yield. During this period, the number of insured acres in the Oregon counties alone rose from 4,535 to 108,569. However, RMA did not adjust its coverage to take into account the lower yields associated with fields planted every year. Consequently, producers received an insurance guarantee based on a history of yields from fields that had been fallow in alternate years, even though now they planted these fields every year, which made them unlikely to achieve the higher yields of a summer fallow practice. According to RMA's data, a summer fallow practice provides producers yields that are up to 33 percent higher than annual planting practice. For example, a producer who grew an average of 40 bushels of wheat per acre using the summer fallow production practice may have the potential to grow only 30 bushels per acre using annual planting practice. Since RMA allows producers to use production history from summer fallow practices to establish insurable yields for annual crop production, the producer in this example can grow 30 bushels annually and make an insurance claim for the other 10 bushels (although over time the actual production history will decrease, reducing the producer's ability to file a claim). RMA's data mining showed that producers took advantage of this program vulnerability. Excessive insurance guarantees for some producers may have contributed to higher claims. Beginning with the 2004 crop year, RMA decided to offer insurance for wheat in these counties by the practice producers employed, either a summer fallow practice or continuous cropping practice. While this decision should reduce program vulnerability, the problem will only be eliminated gradually. RMA did not require producers to recertify their historical acreage and production by separate practice to correct the insurance guarantee. RMA officials agreed that some producers will continue receiving claims payments based on an inflated "normal" yield history until the production history is corrected with actual yields over the next 10 years. RMA's Regulations and Statutory Requirements Hinder RMA Officials' Efforts to Reduce Abuse in the Crop Insurance Program: RMA's regulations, as well as statutory mandates, have created a program design that can impede RMA officials' efforts to prevent and detect fraud, waste, and abuse in a number of ways. First, in terms of RMA's regulations, producers can insure their fields individually instead of insuring all fields combined, which makes it easier for them to switch production among fields, either to make false insurance claims or to build up a higher yield history on a particular field in order to increase its eligibility for higher future insurance guarantees. In addition, companies participating in the crop insurance program bear minimal risk on some of the policies they sell and service, giving the companies little incentive to rigorously challenge questionable claims on these policies. In terms of statutory requirements, RMA must offer producers "prevented planting" coverage-- coverage if an insured crop is prevented from being planted--but it is often difficult to determine whether the producer had the opportunity to plant a crop. Furthermore, statutorily established premium subsidies are high and, therefore, may shield high-risk producers from the full effect of paying higher premiums. Option to Allow Producers to Insure Each of Their Fields Separately May Contribute to Program Abuse: Many patterns of producer fraud, waste, and abuse are possible if producers manipulate how they report production from separately insured units. Under RMA's regulations, producers can insure production of a crop on each optional unit or insure an entire basic unit. With separately insured optional units, for example, if hail damages a crop on one field, producers receive an insurance indemnity to cover the hail losses. However, if producers insured their entire crop in a single basic insurance unit, the hail losses may not have caused the production yield of all units combined to have been below the level guaranteed by the insurance and, therefore, would not warrant an indemnity payment. However, separately insured optional units make it easier for producers to report production from one field that was actually produced on a second field in order to make false insurance claims or to build up a higher yield history on a particular field to increase its eligibility for higher future insurance guarantees.[Footnote 20] Since claims payments for optional units are based upon the yield in each field, rather than the yield for the entire farm, the result of this misreporting is to generate or increase claims on the first field while enhancing the yield for future insurance guarantees on the second field. In a future period, the producer reallocates production from the second field to the first field, thus increasing indemnities on the second field while rebuilding the yield of the first field. Insurance companies or RMA could increase inspection activity in an attempt to reduce occurrences of production switching, but increased activity would raise the costs of administering the program. According to a 2002 RMA study, relative losses per unit increase as the number of separately insured optional units increases.[Footnote 21] Furthermore, given the similarities in a producer's separately insured units, the study could not identify any credible reasons, in the absence of fraud, waste, or abuse, that the losses should increase with increases in the number of separately insured units. Finally, the study concluded that such loss patterns are unlikely to occur naturally. According to an RMA official, gathering the evidence to support a yield- switching fraud case requires considerable resources, especially for large farming operations. Furthermore, the official noted, in order to prove production switching, adjusters would need to appraise all of a producer's fields just before harvest. In 2003, RMA identified 2,371 suspicious claims, 273 of which (about 12 percent) had patterns associated with switching production among fields. Furthermore, in our review of claim files, we identified 10 producers with patterns of claims associated with this type of fraud. Table 5 highlights a pattern of claims suggesting yield switching, as shown by the production history for a producer farming over 2,000 acres of irrigated cotton in west Texas. Generally, this producer insured a yield of about 700 pounds per unit. To the extent an individual unit reports production below the insurance guarantee, the producer is paid an indemnity. Table 5: Production Reported by an Irrigated Cotton Producer Indicating Yield Switching: Unit identifier: 101; Production per acre (in pounds): Crop year 2001: 1,419; Production per acre (in pounds): Crop year 2002: 113; Production per acre (in pounds): Crop year 2003: 184. Unit identifier: 102; Production per acre (in pounds): Crop year 2001: 156; Production per acre (in pounds): Crop year 2002: 1,769; Production per acre (in pounds): Crop year 2003: 366. Unit identifier: 103; Production per acre (in pounds): Crop year 2001: 208; Production per acre (in pounds): Crop year 2002: 230; Production per acre (in pounds): Crop year 2003: 1,523. Unit identifier: 104; Production per acre (in pounds): Crop year 2001: 303; Production per acre (in pounds): Crop year 2002: 387; Production per acre (in pounds): Crop year 2003: 183. Unit identifier: 105; Production per acre (in pounds): Crop year 2001: [A]; Production per acre (in pounds): Crop year 2002: 445; Production per acre (in pounds): Crop year 2003: 166. Claim payment received; Production per acre (in pounds): Crop year 2001: $539,233; Production per acre (in pounds): Crop year 2002: $450,077; Production per acre (in pounds): Crop year 2003: $639,457. Source: GAO analysis of RMA's claim data. [A] Unit number 105 was not insured in crop year 2001. [End of table] For example, in 2001 the producer harvested the crop on Unit 101 with a reported yield of 1,419 pounds of cotton per acre and reported losses on the remaining three units, thereby obtaining claims payments of over $500,000. It appears some of the production from the three units with claims for losses was shifted to the unit with the high production. By building up a higher yield history on Unit 101, the producer increased the insurance guarantee on this unit for 2002 and beyond. In 2002, the producer claimed a loss on Unit 101, as well three other units, obtaining claims payments of over $400,000. In 2003, the pattern was repeated, resulting in claims payments of more than $600,000. FSA's field inspection of Unit 101 in 2003, conducted in September just prior to harvest, found that the "cotton looked good and was comparable with other fields in the area." Nonetheless, the insurance company paid the claims on the units to the producer in December 2003 and January 2004. Moreover, we found that the producer in this case--a farming operation set up as a general partnership--leased land from the owner of the cotton gin where the farming operation sold its cotton and where the cotton gin recorded the production levels that the farming operation used to substantiate its claims. The owner of the gin also provided the partnership with loan security to obtain operating capital. Furthermore, one of the partners in the farming operation, who had a power of attorney to sign documents on behalf of the partnership, was also employed in the office of the cotton gin. We referred this case to RMA for follow-up investigation, which reported that there was not enough evidence of abuse to refer the case for sanction or prosecution. In his 2003 loss review, however, the claims adjuster questioned the producer's farming practices, prompting the insurance company to perform a preharvest inspection in 2004. The producer did not file a claim in 2004. In 2005, RMA and insurance company representatives performed joint preharvest appraisals on this producer's fields in anticipation of a filing for a claim. No claim had been filed for 2005 at the time we completed our review. Minimal Risk Sharing on Some Policies May Not Provide Insurance Companies Strong Incentive to Carry Out Their Responsibilities under the Program: Insurance companies participating in the crop insurance program share a percentage of the risk of loss or opportunity for gain on each insurance policy they write, but the federal government ultimately bears a high share of the risk. Under the SRA, insurance companies are allowed to assign policies to one of three risk funds--assigned risk, developmental, or commercial. The SRA provides some criteria for designating policies to these funds. For the assigned risk fund, the companies cede up to 85 percent of the premium and associated liability for claims payments to the government and share a limited portion of the gains and losses on the policies they retain. Economic incentives to control program costs associated with fraud, waste, and abuse are commensurate with financial exposure. Therefore, for policies placed in the assigned risk fund, companies have far less incentive to investigate claims than the federal government would. For example, in one claim file we reviewed, an insurance company official characterized the producer as filing frequent, questionable claims; however, the company paid a claim of over $500,000. The official indicated that if the company vigorously challenged the claim, the producer would have defended his claim just as vigorously, and the company would have potentially incurred significant litigation expenses, which RMA does not reimburse. In the company's opinion, it was less costly to pay the claim. In 2003, companies placed about 19 percent of the policies they wrote in the assigned risk fund and about 69 percent in the commercial fund. However, for those producers on RMA's spot-check list, about 47 percent of the policies were in the assigned risk fund, and 38 percent were in the commercial fund. RMA and Insurance Companies Have Difficulty Determining Potential Abuse Associated with Prevented Planting Coverage: Under the Federal Crop Insurance Act, as amended, RMA must offer prevented planting coverage. Under the act and its implementing regulations, RMA allows claims for prevented planting if producers cannot plant due to an insured cause of loss that is general in the surrounding area and that prevents other producers from planting acreage with similar characteristics.[Footnote 22] Claims for prevented planting are paid at a reduced level, recognizing that producers do not incur all production costs associated with planting and harvesting a crop. However, determining whether producers can plant their crop may be difficult. Annually, RMA pays about $300 million in claims for prevented planting. In written comments on our survey, 25 FSA inspectors reported that they believe some producers in their county who claimed prevented planting losses never intended to plant or did not make a good faith attempt to plant their crop. Additionally, in some cases, it appears that the insurance company's claims adjusters may not exercise due diligence in evaluating prevented planting claims. For example, a producer in south Texas received claims payments of over $21,000 for prevented planting claims for corn in 2003 and 2004. The producer claimed that excess rainfall made his fields too wet to plant. However, according to a June 2004 FSA field inspection report, there was no evidence the producer had made any attempt to prepare the fields for planting in either the 2003 or 2004 growing seasons. The FSA inspection report noted, and photographs showed, the fields contained permanent grasses and 5-foot tall weeds, as well as large hay bales from the prior growing season. In addition, rainfall for the county in April and May, 2003, was well below normal, and there was no evidence that the producer had ordered seed in anticipation of planting. Moreover, in 2003, of the 66 corn policies in the county, the producer's policy had the only claim for prevented planting. Because the cause of loss was not general to the area, the producer should not have received payment on the claim. Similarly, in 2004, the producer filed one of only three claims for prevented planting of the 55 corn policies in the county. According to an official of the insurance company that sold and serviced this policy, prevented planting claims are paid early in the growing season, and because information on other companies' claims experience is unavailable, it is difficult to assess whether producers' claims are due to an insured cause of loss that is general in the surrounding area and that prevents other producers from planting acreage with similar characteristics. On the basis of our review, RMA investigated the 2003 and 2004 prevented planting claims for this producer and subsequently directed the insurance company to seek reimbursement for the 2003 claim payment. High Premium Subsidies May Inhibit RMA's Ability to Control Program Abuse: To encourage program participation, ARPA increased premium subsidies-- the share of the premium paid by the government--but this increase may hamper RMA's ability to control program waste and abuse. Premium subsidies are calculated as a percentage of the total premium, and producers pay only between 33 to 62 percent of the policy premium, depending on coverage level. High premium subsidies shield producers from the full effect of paying higher premiums. Because premium rates are higher in riskier areas and for riskier crops, the subsidy structure transfers more federal dollars to those who produce riskier crops or farm in riskier areas. In addition, premium rates are higher for producers who choose to insure their fields separately under optional units, rather than all fields combined, because the frequency of claims payments is higher on the separately insured units. Again, however, because of high premium subsidies, producers pay only a fraction of the higher premium. Thus, the subsidy structure creates a disincentive for producers to insure all fields combined. Over one-half (56 percent) of the crop insurance agents responding to our survey believed that charging higher premiums for producers with a pattern of high or frequent claims would discourage fraud, waste, and abuse in the crop insurance program. Finally, in disaster years, ARPA increases insurance protection by allowing producers to substitute a percentage of the historical average county yield for actual yields. As a result of ARPA, RMA sets a floor under producers' annual yields so that yields in any year cannot fall below 60 percent of the historical average county yield (called the transitional yield) for that crop. Consequently, the amount of crop insured against loss is at least 60 percent of the average county yield, giving producers higher coverage than experience would allow. Although RMA sets a higher premium for producers because of actual production losses, because of high premium subsidies producers pay only a fraction of the higher premium. Thus, the subsidy structure creates an incentive for producers to insure at the higher level of protection. Recently Prosecuted Crop Insurance Fraud Cases Highlight Program Vulnerabilities: Eight recent crop insurance fraud cases that were investigated by USDA's OIG and resulted in criminal prosecution between June 2003 and April 2005 reflect some of the issues we identified. The cases show how producers, sometimes in collusion with others, falsely report planting, claims of damage and production to try to circumvent RMA's procedures. In some cases, producers hid production or switched it from one field to another. Several of these cases also demonstrate the importance of having FSA and RMA work together to identify and share information on questionable farming practices/activities. Table 6 summarizes these eight cases, which accounted for $3.1 million in fraudulent claims payments. These cases, which were researched and analyzed by our Office of Forensic Audits and Special Investigations, are described here and in more detail in appendix V. Table 6: Crop Insurance Fraud Cases Investigated by the USDA/OIG and Resulting in Criminal Prosecution, June 2003 to April 2005: Case: 1; Fraud allegation: Failure to plant; How detected: OIG/RMA/FSA identified irregularities through joint data mining effort and follow-up inspection; Collusion: Possible. Insurance adjuster indicted for falsely verifying losses; Fraudulent claims payments: $57,155. Case: 2; Fraud allegation: False claim of crop damage from hail, heat, and drought; How detected: RMA and FSA received complaints and initiated review; Collusion: Possible. Insurance policy purchased from agency owned by a sister-in-law; Fraudulent claims payments: $39,826. Case: 3; Fraud allegation: False claim of crop damage from excessive moisture; How detected: OIG initiated. Fraud detection survey of grain elevator disclosed irregularities; Collusion: No; Fraudulent claims payments: $435,087. Case: 4; Fraud allegation: Failure to plant; How detected: FSA filed complaint with RMA; Collusion: Yes. Insured was also agent and issued policies through his agency. Insurance adjusters falsified forms. Seed dealers also provided false receipts; Fraudulent claims payments: $630,000. Case: 5; Fraud allegation: False claim of crop damage; How detected: RMA noticed suspicious adjustments in grain quality by grain elevator company; Collusion: Yes. Farmer and grain elevator operator; Fraudulent claims payments: $1,000,000. Case: 6; Fraud allegation: False crop yield history to inflate insurance claim; How detected: OIG hotline complaint; Collusion: Yes. Insurance agents pled guilty to falsifying insurance documents; Fraudulent claims payments: [A]. Case: 7; Fraud allegation: No ownership interest in crops; underreporting of crop yield; How detected: OIG hotline complaint; Collusion: No; Fraudulent claims payments: $19,000. Case: 8; Fraud allegation: Failure to plant; false claim of moisture damage; concealing production; How detected: Bankruptcy fraud investigation revealed insurance fraud; Collusion: Ongoing investigation of insurance representatives; Fraudulent claims payments: $912,364. Sources: GAO's analysis of USDA and U.S. Department of Justice case information. [A] Data not available. [End of table] These eight crop insurance cases are described as follows: * Case 1. The subject of this investigation, a producer in Tennessee, in 1999 improperly obtained crop insurance coverage for his tomato crop and received a claims payment for losses that had not occurred. Moreover, this producer was ineligible to participate in the crop insurance program because he had not paid a past premium. In order to hide the fact that he was the true grower of 1999 tomato crops in two Tennessee counties, he used his wife's name on crop insurance documents. In addition, his wife filed a report with the insurance company claiming a higher level of acreage planted to inflate the value of any subsequent insurance claim. An insurance adjuster assisted the producer by fraudulently signing forms showing he inspected and measured the nonexistent crops and that his observations supported the wife's claimed loss. * Case 2. A producer planted a wheat crop after the planting deadline, which made the crop ineligible for crop insurance. He reported, however, that the crop had been planted before the deadline and falsely claimed crop losses because of hail, heat, and drought. Furthermore, the producer did not have an ownership interest in the crop. Instead, the producer's brother leased the farm land and paid the cost of planting, and the brother's wife owned the insurance agency that issued the insurance policy on the crop. FSA filed a complaint with RMA because FSA officials had observed that the crop was planted past the planting deadline. * Case 3. Two producers conspired to file fraudulent crop insurance claims, stating that their bean crops had been damaged by excessive moisture. They underreported the crop yield to the insurance company and hid production by delivering the harvest to processing plants using false names. The scheme was discovered by OIG during a fraud detection survey at a grain elevator. Investigators reviewed the sales of uninsured crops and identified production sold under other names that actually belonged to the two producers. * Case 4. A farming partnership filed fraudulent insurance claims of crop losses for cotton, wheat, and grain sorghum that were never planted. The producer, who also owned an insurance agency formed a farming partnership with other family members for these acres, and the producer's insurance agency wrote insurance policies for the farming partnership. Two insurance adjusters, who did not visit any of the fields, filed false appraisal and production worksheets verifying the losses. A seed dealer also prepared false receipts to support the producer's planting claims. However, inconsistent statements on documents submitted to FSA led to an inspection of the farming operation, and inspectors found little evidence of planting. A subsequent investigation resulted in admissions of guilt. * Case 5. The manager of a grain elevator conspired with producers to sell their wheat at discounted prices by providing them with false documentation showing that a large portion of their crop was damaged by weather and below weight. The manager also falsified documents, stating that the crop was of lower quality and provided falsified samples of severely damaged wheat to mislead insurance adjusters. Producers could then collect crop insurance and disaster payments from the federal government. * Case 6. Two crop insurance agents conspired with producers to inflate actual production histories, which allowed the producers to receive higher indemnity payments on insurance claims. The agents backdated an insurance application, created a false insurance policy based on a fictitious yield rate, and had the producers sign blank insurance documents. This fraud was identified through an OIG hotline complaint. * Case 7. An investigation was initiated following an OIG hotline complaint that a producer had reported different crop yields to FSA and RMA. OIG determined that the producer had, among other things, (1) filed four false insurance claims, stating that he had experienced a failed harvest and needed to replant; (2) filed claims on crops in which he had no ownership interest; (3) inflated the size of a corn crop loss; and (4) filed a claim in which he underreported the yield. * Case 8. Producers falsely claimed they (1) were prevented from planting because of excess moisture and (2) had planted crops that they did not plant and claimed losses on these crops. The producers also filed claims in which they underreported the yield and for crops in which they had no interest. In order to report a crop loss and manipulate their yields, the producers sold crops using other people's names. These schemes were discovered during the course of a bankruptcy fraud investigation involving some of the producers. RMA's Failure to Follow Its Guidelines Has Resulted in Program Losses for Some New and Expanded Crop Insurance Products: RMA has not always developed or expanded crop insurance products according to its guidelines, thereby contributing to program losses. RMA's guidelines (1) identify RMA's mission for expanding the crop insurance program; (2) outline the process and procedures by which the RMA responds to requests to add a new insurance program; (3) specify data submission requirements for analysis of the new program request; and (4) establish the framework RMA will use to implement, maintain, and evaluate a new program expansion.[Footnote 23] According to RMA, these guidelines are intended to ensure producer interest and crop suitability and to minimize exposure to loss. We found that, in some instances when RMA did not follow its guidelines, it had higher claims and loss ratios. Most of the newly developed and expanded products we reviewed--15 of the 16 products--were developed under the guidelines RMA had in place before ARPA. Under these guidelines, RMA officials are to obtain information documenting, among other things, the following: * significant grower interest in the insurance coverage; * the crop's economic significance; * actuarial sufficiency and data availability (e.g., producer acreage, crop yield, production cost, and weather data); * a risk profile and analysis (e.g., perils affecting the crop, production experience, available markets, and product viability); * agronomic, aquatic, and horticultural suitability (e.g., commercial life cycle of the crop, rotation requirements, whether the crop is an annual, biennial, or perennial); * marketing potential (e.g., market characteristics, risks, and competition); and: * implementation parameters (e.g., crop year for implementation and number of states). Under ARPA, the FCIC Board of Directors is required to have actuarial and underwriting experts independently review policies, plans of insurance, and related materials before approving new products. Of the 16 crop insurance products we reviewed, 11 were newly developed, and 5 were expansions of an existing crop into a new geographic area.[Footnote 24] RMA's overall loss ratio for the 16 pilot crop products was about 2.0 (or $2.00 in claims paid for every $1.00 in premium). Four of the 16 cases we reviewed had loss ratios of 1.0 or less. (See app. VI for a comparison of loss ratios for these 16 crop insurance products.) In 12 of the 16 cases, it appeared that RMA followed its guidelines in terms of documenting the proposed insured crops' past production experience, crop suitability, potential for loss, and implementation parameters. However, for the cases in which RMA did not follow its guidelines, the products experienced claims in excess of premiums of over $50 million. For example, FCIC's Board of Directors approved coverage of fall- planted watermelons as part of a broader watermelon insurance pilot program covering watermelon production in seven states and 15 counties for spring-and fall-planted watermelons--without considering all the factors called for under its guidelines--such as a horticultural study. Such a consideration would have provided assurance that the proposed product was actuarially sound, according to the OIG.[Footnote 25] Overall, the watermelon pilot program had over $51 million in claims payments in 1999, its first year of operations, and had a loss ratio (claims paid divided by premiums) of 5.8. Of the claims payments in 1999, $21.1 million, or 44 percent, was for fall-planted watermelons in the three Texas counties. According to the Inspector General, a horticultural analysis would have shown that fall watermelons were a high-risk crop in that region of Texas. The product was discontinued after a year. In responding to the Inspector General about its offering of insurance for fall-planted watermelons in Texas, RMA commented that a shortage of experienced staff was a major factor contributing to the agency's lack of adherence to its guidelines. In another case--the apple quality option policy--the FCIC Board of Directors quickly approved this product despite a number of questions suggesting the need for additional study. The approved product, which insured a higher grade of apples than existing apple crop policies, experienced large losses. In determining whether to approve this product, the Board of Directors contracted for studies from five independent companies. One of the studies reported 10 major concerns about the product, including incomplete documentation, the large number of counties in the pilot program (approximately 86 percent of the apple crop), and an understated premium rate. Three of the other four reviewers raised similar concerns. The other reviewer supported the proposal with minor reservations. Even with these concerns, the FCIC Board of Directors quickly approved the pilot without additional analyses. The claims for this product for crop years 2001 through 2003 were about $4.4 million, and the product had a loss ratio of 2.6. RMA officials contend that the loss ratio can primarily be attributed to some apple producers in California, who may have abused the system. According to RMA officials, the Board of Directors discussed the concerns raised by the studies but still approved the request for the pilot. A delay in approval might have delayed the pilot's implementation for a year. As of August 2005, RMA was moving to contract with independent reviewers to evaluate the apple quality option pilot program including its actuarial soundness and whether it effectively meets the needs of apple producers. Upon approval by the FCIC Board of Directors, new products have a probationary period--generally 1 to 3 years. Under RMA guidelines, newly developed products are to be examined annually to see that they are meeting performance goals--such as producer participation by year, state, county, and insurance company; loss ratios; and appropriate premiums. RMA is to make adjustments if warranted and determine whether the product should be continued or terminated. According to RMA officials, in lieu of a formal review, RMA informally collects data on each new insurance product by at least annually corresponding with RMA regional offices and outside sources. They said that annual reviews are only of limited value in providing the information RMA needs to determine the viability of a product because it takes several years to get a clear picture of how well an insurance product will perform. The lack of an annual evaluation did not appear to significantly affect future years' loss experience for most of the 11 new products we reviewed. However, timely annual evaluation of the sweet potato program might have saved the federal crop insurance program several million dollars. The sweet potato program's loss ratio from 1998 through 2002 ranged from 2.3 to 5.2, but RMA waited until 2003 before making changes to it and made additional changes for the 2004 crop year. From 1998 to 2003, the sweet potato program recorded claims of $47 million compared with premiums of about $12 million. An evaluation completed in 2003 suggested that the high dollar claims for this product could not fully be attributed to weather conditions but rather suggested potential fraud or abuse. Earlier, more timely reviews could have identified the irregular claims activity in selected counties and might have averted the claims payments. In October 2004, the FCIC Board of Directors terminated the existing sweet potato pilot program and implemented a new program that included reduced coverage and increased growing experience requirements for participation. In May 2005, the Inspector General reported on RMA's failure to follow its procedures for performing new product reviews, including the sweet potato pilot program, and recommended that RMA improve the timeliness of its evaluations for new products.[Footnote 26] Conclusions: Federal crop insurance plays an invaluable role in assuring the nation's farmers that their crops will be protected from natural disasters. However, the importance of the program and the significant role it plays in U.S. agriculture is frequently overshadowed by controversy associated with fraud, waste, and abuse in the program. In recent years, with the assistance of the new tools in ARPA, RMA has made progress in strengthening a number of program elements and thereby reducing fraud, waste, and abuse, as well as the amount of funds paid in error. Still, we identified weaknesses in how RMA, FSA, and insurance companies carry out their program responsibilities, and these weaknesses continue to leave the program vulnerable to questionable claims and missed opportunities to prevent losses to the federal government. Lack of timely, useful communication between RMA and FSA has resulted in insufficient information sharing between the two agencies, as well as with the insurance companies and insufficient inspections of land with suspicious claims. Furthermore, RMA has not been given access by FSA to key information on producers who have a beneficial interest in one or more farming operations. As a result, many of the largest farming operations are not closely scrutinized in the crop insurance program. Many of these farming entities fail to disclose producers having an ownership interest in the entity and who may be ineligible to participate in the federal crop insurance program. In addition, insurance companies' quality control reviews of claims are weak because RMA does not effectively oversee the companies' quality assurance efforts leaving the crop insurance program susceptible to fraud and abuse. Finally, RMA has used one of its key enforcement tools--sanctions against producers, agents, and companies--on a very limited scale because, 4 years after ARPA, it has not promulgated regulations establishing how it will impose some of the additional sanctions authorized under ARPA. We recognize that the crop insurance regulations or statutory provisions are intended to strengthen protection for producers. However, in many cases, RMA has difficulty ensuring that these provisions are not abused. RMA has opportunities to improve the program's design by eliminating the option of insuring fields (optional units) for producers who have a history of high losses in comparison with other producers growing the same crop in the same area. The program's high premium subsidies, specified in the Federal Crop Insurance Act, as amended by ARPA, may also limit RMA's ability to control program abuse because the subsidies shield producers from the full effect of paying higher premiums associated with frequent or larger claims. Periodic lapses in program management when developing and expanding new crop insurance products limit the effectiveness of RMA's guidelines and can cause unnecessary losses to the crop insurance program. Generally, when RMA followed its guidelines, new products incurred fewer losses. ARPA provided RMA with a new, seemingly more rigorous process to review new products. However, the new process cannot succeed unless RMA more closely follows its guidelines. The crop insurance program is designed to accommodate the needs of all of America's producers. Fraud, waste, and abuse can cause producers to pay more for crop insurance and hurt the reputation of the program. Further reducing vulnerability to fraud, waste, and abuse in the crop insurance program will require a coordinated effort among the agencies and companies managing the program--RMA, FSA, and the participating insurance companies. Each agency and the companies have an important role in monitoring agent, adjuster, and producer actions and in sharing key program information with one another. Matter for Congressional Consideration: To better protect the crop insurance program from fraud, waste, and abuse, Congress should consider allowing RMA to reduce premium subsidies--and hence raise the insurance premiums--for producers who consistently have claims that are irregular in comparison with other producers growing the same crop in the same location. Recommendations for Executive Action: To better ensure that field inspections are used to the maximum effect to address fraud, waste, and abuse in the federal crop insurance program, we recommend that the Secretary of Agriculture take the following eight actions. Specifically, we recommend that the Secretary direct the Administrators of RMA and FSA to create an action plan to ensure that: * FSA field offices conduct all inspections called for under agency guidance; * RMA informs FSA field inspectors of the suspect claim patterns that they are to investigate; and: * FSA inspections are conducted in a timely manner, and inspection results are reported expeditiously to insurance companies. We further recommend that the Secretary of Agriculture: * promulgate regulations to implement its expanded authority under ARPA to impose sanctions; * direct FSA to share producer-derived information with RMA for data mining to administer and enforce the requirements of the crop insurance program; * direct RMA to determine if payments have been made to ineligible producers or to entities that failed to disclose ownership interests and, if so, to recover the appropriate amounts; * direct RMA to strengthen its oversight of the insurance companies' implementation of the quality control review system; and: * direct RMA to reduce the insurance guarantee or eliminate optional unit coverage for producers who consistently have claims that are irregular in comparison with other producers growing the same crop in the same location. Agency Comments and Our Evaluation: We provided USDA with a draft of this report for its review and comment. We received written comments from USDA's Under Secretary for Farm and Foreign Agricultural Services. The department agreed to act on most of our recommendations, including (1) ensuring that inspections are conducted in a timely manner, and that inspection results are reported expeditiously to insurance companies; (2) directing FSA to share producer-derived information with RMA; (3) directing RMA to determine if payments have been made to ineligible producers or to entities that failed to disclose ownership interests and, if so, to recover the appropriate amounts; and (4) directing RMA to strengthen its oversight of the insurance companies' implementation of the quality control review system. With respect to issuing regulations to implement its expanded authority under ARPA to impose sanctions, USDA agreed that promulgation of ARPA- based sanction regulations would enhance RMA's sanctions efforts. However, USDA stated that it was incorrect to suggest that the lack of regulations is a critical impediment to imposing sanctions. USDA also stated that there has been a learning curve since ARPA was enacted but that it has been imposing sanctions under its ARPA authority since 2000 and that it has made "significant and steady progress" in both the numbers and types of sanctions imposed. Our report indicates that USDA has imposed some sanctions since the enactment of ARPA. However, the number of sanctions imposed by RMA has not increased appreciably since the enactment of ARPA. For example, RMA imposed an average of less than 20 sanctions annually from 1996 to 2000, and an average of less than 20 sanctions annually from 2001 to 2005, except for 2004 (67), which was an exception. While not all questionable claims payments are necessarily sanctionable, RMA has identified about 3,000 questionable payments annually since beginning data mining in 2001. We continue to believe RMA has not fully exercised its new authority. Under ARPA, RMA has new authority to impose sanctions on agents, loss adjusters, approved insurance providers, and others who willfully and intentionally fail to comply with an FCIC program requirement. In April 2005, RMA officials told us that the number of sanctions has not substantially increased in part because regulations have not been promulgated under ARPA. Subsequently, an official from USDA's Office of General Counsel told us that RMA had not imposed any sanctions under its new authority to do so on the basis of a failure to comply with an FCIC program requirement. This official indicated that regulations were needed to establish what constitutes an FCIC program requirement, how USDA will determine that a material violation has occurred, and what process will be followed before imposing sanctions under this provision. USDA does not dispute the report's findings that no sanctions have been imposed under this sanction provision of ARPA. USDA disagreed with our recommendation to ensure that FSA field offices conduct all inspections called for under agency guidance because FSA did not have sufficient resources to complete all of these inspections. Given the potentially high payoff from providing greater assistance to RMA, we believe that FSA may want to conduct a study to determine the costs and benefits of making staff available for crop insurance inspections. USDA also disagreed with our recommendation to reduce the insurance guarantee or eliminate optional unit coverage for producers who consistently file claims that are irregular in comparison with other producers growing the same crop in the same location. USDA stated that this recommendation represented a disproportionate response, considering the small number of producers identified as yield switching each year and that adoption of our recommendation would not be cost- effective. We agree that the number of policies identified annually as having patterns consistent with yield switching is small in comparison with the number of policies in the crop insurance program. However, we believe it is possible to narrowly tailor an underwriting rule so that it would target only a few producers each year and would entail few resources. Such a tool would provide RMA with another means to discourage producers from abusing the program. Thus, we continue to believe it is reasonable for RMA to reduce the insurance guarantee or eliminate optional unit coverage for producers who consistently file and receive questionable claims payments. USDA also provided technical corrections, which we have incorporated into this report as appropriate. USDA's written comments are presented in appendix VII. As agreed with your office, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will send copies of this report to appropriate Congressional Committees; the Secretary of Agriculture; the Director, Office of Management and Budget; and other interested parties. In addition, this report will be available at no charge on the GAO Web site at [Hyperlink, http://www.gao.gov]. If you have any questions about this report, please contact me at (202) 512-3841 or [Hyperlink, robinsonr@gao.gov]. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made major contributions to this report are listed in appendix VIII. Sincerely yours, Signed by: Robert A. Robinson: Managing Director, Natural Resources and Environment: [End of section] Appendixes: Appendix I: Objectives, Scope, and Methodology: At the request of the Chairman of the Senate Committee on Homeland Security and Governmental Affairs, we examined the U.S. Department of Agriculture (USDA) Risk Management Agency's (RMA) procedures for assuring integrity in the crop insurance program. Specifically, we agreed to (1) assess the effectiveness of USDA's procedures and processes to prevent and detect fraud, waste, and abuse in selling and servicing crop insurance policies; (2) determine the extent to which program design issues may make the program more vulnerable to fraud, waste, and abuse; and (3) determine the effectiveness of USDA's procedures to assure program integrity in developing new crop insurance products. To assess the effectiveness of USDA's procedures and processes to prevent and detect fraud, waste, and abuse in the selling and servicing of crop insurance policies, we examined a nonrandom sample of 120 insurance claims from the 2,794 claims that RMA identified as warranting a field inspection in 2004. Of these 120 claims, 100 were the largest claims paid, and 20 were selected to ensure representation of all data analysis selection criteria. To review the claims, we visited seven insurance companies and contacted three others to request the claim files. We also examined the guidance that RMA and USDA's Farm Service Agency (FSA) field offices use to implement the requirements of the Agricultural Risk Protection Act of 2000 (ARPA), including the FSA/RMA Handbook FCIC Program Integrity, 4-RM. In addition, we examined guidance that RMA uses to ensure compliance with the Federal Crop Insurance Act, including relevant laws; regulations and agency policy, including the 2003 Crop Insurance Handbook; Manual 14, Guidelines and Expectations for Delivery of the Federal Crop Insurance Program; loss adjustment manuals; crop insurance handbooks; and related managers' bulletins and notices. In addition, we conducted two surveys. To assess the effectiveness of RMA and FSA's coordinated monitoring efforts at the local level, we conducted a Web-based survey of all 829 FSA county officials who were responsible for conducting field inspections in 2003. The survey contained 17 questions that asked for opinions and assessments of (1) preharvest and growing season inspections; (2) training and education on crop insurance; (3) the findings and actions taken by RMA; and (4) FSA's role in reducing fraud, waste, and abuse in crop insurance. In developing the questionnaire, we met with officials in FSA headquarters to gain a thorough understanding of the coordinated plan FSA uses to assist RMA in monitoring the crop insurance program. We also shared a draft copy of the questionnaire with these officials, who provided us with comments, including technical corrections. We then pretested the questionnaire with FSA officials in two county offices each in Texas and Georgia, as well as with officials in one office each in Maryland, Minnesota, and North Dakota. During these pretests, we asked the officials to complete the Web-based survey as we observed the process. After completing the survey, we interviewed the respondents to ensure that (1) questions were clear and unambiguous, (2) terms we used were precise, (3) the questionnaire did not place an undue burden on the FSA county officials completing it, and (4) the questionnaire was independent and unbiased. On the basis of the feedback from the pretests, we modified the questions, as appropriate. The questionnaire was posted on GAO's survey Web site. When the survey was activated, the officials who had been selected to participate were informed of its availability with an e-mail message that contained a unique user name and password. This allowed respondents to log on and fill out a questionnaire but did not allow respondents access to the questionnaires of others. The survey was available from December 9, 2004, until March 11, 2005. We received responses for 92 percent of the 829 FSA officials. Results of the survey to FSA county officials are summarized in appendix III.[Footnote 27] To solicit crop insurance agents' views on control weaknesses and suggestions for improving oversight of the crop insurance program, we developed a questionnaire that was mailed to crop insurance agents. In developing the questionnaire, we met with officials in RMA headquarters to gain a thorough understanding of internal controls in the crop insurance program. We also shared a draft copy of the questionnaire with these officials who provided us comments, including technical corrections. We then pretested the questionnaire with five crop insurance agents in Texas, five agents in Georgia, and five agents in North Dakota. During each pretest we interviewed the respondent to make sure that the (1) questions were clear and unambiguous, (2) questionnaire terms were precise, (3) questionnaire did not place an undue burden on the agents completing it, and (4) questionnaire was independent and unbiased. The survey was mailed to the crop insurance agents on December 21, 2004. Because of the large number of agents (more than 13,000) that participate in the crop insurance program, we used a probability sample. Before sampling, agents were grouped by the dollar amount of policies sold in 2003. The number of agents sampled from each group was proportionate to the number of agents in the group. The final number of agents selected was 951. However, because of incorrect addresses or other information, the final sample was 935. We received responses from 76 percent of these 935 insurance agents. The results from the survey are presented as estimates, each with a measurable precision, or sampling error, that may be expressed as a plus/minus figure. By adding the sampling error to and subtracting it from the estimate, we developed upper and lower bounds for each estimate. The range between these two estimates is called the confidence interval. Sampling errors and confidence intervals are stated at a certain confidence level, in this case, 95 percent. For example, a confidence interval at the 95 percent confidence level means that in 95 out of 100 cases, the sampling procedure would produce a confidence interval that would include the estimated value. Results of the survey are summarized in appendix IV.[Footnote 28] At our request, the Center for Agribusiness Excellence in Stephenville, Texas, conducted data mining on RMA's crop insurance databases and FSA's computer databases for farming operations receiving commodity program payments. For these operations, the databases contain detailed information on the individuals that are members or beneficiaries, their share of payments, and additional organizational details. We asked the Center for Agribusiness Excellence to determine whether (1) policyholders report all farming operations in which they have a substantial beneficial interest, (2) certain farming operations manipulate production among affiliated farming entities to file questionable crop insurance claims, and (3) conflicts of interest exist for members of farming operations who are also agents or claims adjusters. In addition, for policyholders identified as not having reported all farming operations in which they have a substantial beneficial interest, we asked the Center for Agribusiness Excellence to determine whether any of these policyholders had been identified by RMA as ineligible to participate in the crop insurance program. To determine the extent to which program design issues may make the program more vulnerable to fraud, waste, and abuse, we conducted a qualitative assessment of economic studies. We also discussed these issues with USDA program officials in headquarters and with field office officials who conduct inspections during the growing season. To examine practices recently employed by producers, agents, and adjusters to defraud the federal crop insurance program, GAO's Office of Forensic Audits and Special Investigations (FSI) asked USDA's Office of Inspector General (OIG) to provide the names of all crop insurance fraud cases they had investigated in the preceding 2 years that resulted in criminal prosecution. OIG identified eight cases of crop insurance fraud prosecuted between June 2003 and April 2005. FSI reviewed OIG's case files for these cases, spoke with representatives from the U.S. Department of Justice, and reviewed relevant reports, court papers, and other documentation on these cases. FSI conducted its investigation from February through June 2005 in accordance with quality standards for investigations as set forth by the President's Council on Integrity and Efficiency. Finally, to determine the effectiveness of USDA's procedures in assuring program integrity in developing new crop insurance programs, we evaluated the agency's policies, procedures, and other pertinent documents to determine what controls were in place to assure program integrity. Specifically, for developmental products, we reviewed RMA's New Program Development Handbook and, for expansion of existing products, we reviewed RMA's General Guidelines and Criteria for Submitting County Crop Program Expansion Requests. We reviewed these RMA guidelines to determine the types of analyses expected and the process that was to be followed in order to establish a new or expansion product. We also spoke with RMA's product development and expansion officials. We selected a nonrandom sample of 16 products that were developed (11 products) or expanded (5 products) during 1998 and 2002 and compared the work performed and documented with the appropriate RMA guidance. The 11 products included 5 products with the highest loss ratios (over $3 claims paid for every $1 of premiums) and 6 products with low loss ratios (less than $3 in claims paid for every $1 of premiums). We conducted our review from July 2004 through August 2005 according to generally accepted government auditing standards, which included an assessment of data reliability and internal controls. [End of section] Appendix II: Selected Information on the Federal Crop Insurance Program, 1981-2004: Acres and dollars in millions. Year: 1981; Policies: (in thousands): 416.8; Acres: 45.0; Liability: $5,981.2; Total premium: $376.8; Subsidy: $47.0; Producer premium: $329.8; Claims paid: $407.3; Total loss ratio[A]: 1.08; Producer loss ratio[B]: 1.23. Year: 1982; Policies: (in thousands): 386.0; Acres: 42.7; Liability: $6,124.9; Total premium: $396.1; Subsidy: $91.3; Producer premium: $304.8; Claims paid: $529.1; Total loss ratio[A]: 1.34; Producer loss ratio[B]: 1.74. Year: 1983; Policies: (in thousands): 310.0; Acres: 27.9; Liability: $4,369.9; Total premium: $285.8; Subsidy: $63.7; Producer premium: $222.1; Claims paid: $583.7; Total loss ratio[A]: 2.04; Producer loss ratio[B]: 2.63. Year: 1984; Policies: (in thousands): 389.8; Acres: 42.7; Liability: $6,619.6; Total premium: $433.9; Subsidy: $98.3; Producer premium: $335.6; Claims paid: $638.4; Total loss ratio[A]: 1.47; Producer loss ratio[B]: 1.90. Year: 1985; Policies: (in thousands): 414.6; Acres: 48.6; Liability: $7,159.9; Total premium: $439.8; Subsidy: $100.1; Producer premium: $339.7; Claims paid: $683.1; Total loss ratio[A]: 1.55; Producer loss ratio[B]: 2.01. Year: 1986; Policies: (in thousands): 406.9; Acres: 48.7; Liability: $6,230.0; Total premium: $379.7; Subsidy: $88.1; Producer premium: $291.6; Claims paid: $615.7; Total loss ratio[A]: 1.62; Producer loss ratio[B]: 2.11. Year: 1987; Policies: (in thousands): 433.9; Acres: 49.1; Liability: $6,094.9; Total premium: $365.1; Subsidy: $87.6; Producer premium: $277.5; Claims paid: $369.8; Total loss ratio[A]: 1.01; Producer loss ratio[B]: 1.33. Year: 1988; Policies: (in thousands): 461.0; Acres: 55.6; Liability: $6,964.7; Total premium: $436.4; Subsidy: $108.0; Producer premium: $328.4; Claims paid: $1,067.6; Total loss ratio[A]: 2.45; Producer loss ratio[B]: 3.25. Year: 1989; Policies: (in thousands): 948.6; Acres: 101.6; Liability: $13,535.8; Total premium: $814.3; Subsidy: $205.0; Producer premium: $609.3; Claims paid: $1,212.2; Total loss ratio[A]: 1.49; Producer loss ratio[B]: 1.99. Year: 1990; Policies: (in thousands): 894.8; Acres: 101.4; Liability: $12,828.4; Total premium: $836.5; Subsidy: $215.3; Producer premium: $621.2; Claims paid: $973.0; Total loss ratio[A]: 1.16; Producer loss ratio[B]: 1.57. Year: 1991; Policies: (in thousands): 706.8; Acres: 82.4; Liability: $11,216.0; Total premium: $737.0; Subsidy: $190.1; Producer premium: $546.9; Claims paid: $955.3; Total loss ratio[A]: 1.30; Producer loss ratio[B]: 1.75. Year: 1992; Policies: (in thousands): 663.4; Acres: 83.1; Liability: $11,334.1; Total premium: $758.8; Subsidy: $196.7; Producer premium: $562.1; Claims paid: $918.2; Total loss ratio[A]: 1.21; Producer loss ratio[B]: 1.63. Year: 1993; Policies: (in thousands): 679.2; Acres: 83.7; Liability: $11,353.4; Total premium: $755.7; Subsidy: $200.0; Producer premium: $555.7; Claims paid: $1,655.5; Total loss ratio[A]: 2.19; Producer loss ratio[B]: 2.98. Year: 1994; Policies: (in thousands): 800.9; Acres: 99.6; Liability: $13,608.4; Total premium: $949.4; Subsidy: $254.9; Producer premium: $694.5; Claims paid: $601.1; Total loss ratio[A]: 0.63; Producer loss ratio[B]: 0.87. Year: 1995; Policies: (in thousands): 2,034.3; Acres: 220.5; Liability: $23,728.5; Total premium: $1,543.3; Subsidy: $889.4; Producer premium: $653.9; Claims paid: $1,567.7; Total loss ratio[A]: 1.02; Producer loss ratio[B]: 2.40. Year: 1996; Policies: (in thousands): 1,615.2; Acres: 204.9; Liability: $26,876.8; Total premium: $1,838.6; Subsidy: $982.1; Producer premium: $856.5; Claims paid: $1,492.7; Total loss ratio[A]: 0.81; Producer loss ratio[B]: 1.74. Year: 1997; Policies: (in thousands): 1,319.8; Acres: 182.2; Liability: $25,459.0; Total premium: $1,775.4; Subsidy: $902.8; Producer premium: $872.6; Claims paid: $993.6; Total loss ratio[A]: 0.56; Producer loss ratio[B]: 1.14. Year: 1998; Policies: (in thousands): 1,242.7; Acres: 181.8; Liability: $27,921.4; Total premium: $1,875.9; Subsidy: $946.3; Producer premium: $929.6; Claims paid: $1,677.5; Total loss ratio[A]: 0.89; Producer loss ratio[B]: 1.80. Year: 1999; Policies: (in thousands): 1,288.8; Acres: 196.9; Liability: $30,939.5; Total premium: $2,310.1; Subsidy: $1,394.0; Producer premium: $916.1; Claims paid: $2,434.7; Total loss ratio[A]: 1.05; Producer loss ratio[B]: 2.66. Year: 2000; Policies: (in thousands): 1,323.2; Acres: 206.5; Liability: $34,443.8; Total premium: $2,540.2; Subsidy: $1,365.8; Producer premium: $1,174.4; Claims paid: $2,594.8; Total loss ratio[A]: 1.02; Producer loss ratio[B]: 2.21. Year: 2001; Policies: (in thousands): 1,297.9; Acres: 211.3; Liability: $36,730.3; Total premium: $2,961.8; Subsidy: $1,771.7; Producer premium: $1,190.1; Claims paid: $2,960.2; Total loss ratio[A]: 1.00; Producer loss ratio[B]: 2.49. Year: 2002; Policies: (in thousands): 1,259.5; Acres: 214.9; Liability: $37,335.0; Total premium: $2,916.3; Subsidy: $1,741.8; Producer premium: $1,174.5; Claims paid: $4,066.9; Total loss ratio[A]: 1.39; Producer loss ratio[B]: 3.46. Year: 2003; Policies: (in thousands): 1,241.5; Acres: 217.4; Liability: $40,619.0; Total premium: $3,431.2; Subsidy: $2,041.9; Producer premium: $1,389.3; Claims paid: $3,254.6; Total loss ratio[A]: 0.95; Producer loss ratio[B]: 2.34. Year: 2004; Policies: (in thousands): 1,228.8; Acres: 221.1; Liability: $46,615.5; Total premium: $4,186.2; Subsidy: $2,477.4; Producer premium: $1,708.8; Claims paid: $3,110.9; Total loss ratio[A]: 0.74; Producer loss ratio[B]: 1.82. 1981-94 average; Policies: (in thousands): 565.2; Acres: 65.2; Liability: $8,815.8; Total premium: $569.0; Subsidy: $139.0; Producer premium: $429.9; Claims paid: $800.7; Total loss ratio[A]: 1.41; Producer loss ratio[B]: 1.86. 1995-04 average; Policies: (in thousands): 1,385.2; Acres: 205.8; Liability: $33,066.9; Total premium: $2,537.9; Subsidy: $1,451.3; Producer premium: $1,086.6; Claims paid: $2,415.4; Total loss ratio[A]: 0.95; Producer loss ratio[B]: 2.30. Source: RMA. Note: GAO analysis of RMA's data. [A] Claims paid (indemnity) divided by total premium. [B] Claims paid (indemnity) divided by producer premium. [End of table] [End of section] Appendix III: Results of Implementation of the Agricultural Risk Protection Act of 2000: Survey of FSA County Directors of USDA: Q1. Do you feel that you clearly understand RMA's expectations regarding conducting growing season inspections (spot checks)? Definitely yes (percent): 49.7%; Probably yes (percent): 43.1%; Uncertain (percent): 4.7%; Probably not (percent): 1.7%; Definitely not (percent): 0.8%; No response (percent): 0.0%; Number of respondents: 743. [End of table] Q2. In your opinion, are most producers in your county aware of growing season inspections that are conducted on the fields of other producers? Definitely yes (percent): 6.5%; Probably yes (percent): 25.6%; Uncertain (percent): 15.3%; Probably not (percent): 43.6%; Definitely not (percent): 8.7%; No opinion (percent): 0.1%; No response (percent): 0.1%; Number of respondents: 743. [End of table] Q3. In your opinion, have growing season inspections in your county increased compliance of most producers with crop insurance provisions? Definitely yes (percent): 3.4%; Probably yes (percent): 25.2%; Uncertain (percent): 32.2%; Probably not (percent): 32.0%; Definitely not (percent): 5.5%; No opinion (percent): 1.5%; No response (percent): 0.3%; Number of respondents: 743. [End of table] Q4. During the past two years have you observed a potential non- compliance situation during the normal course of your official duties? Yes (percent): 28.1%; No (percent): 68.4%; No response (percent): 3.5%; Number of respondents: 743. [End of table] Q5. During the past two years, when you observed a potential non- compliance situation during the normal course of your official duties, how often did you initiate CCC Form 2007 (Report of Crop Insurance Non- compliance)? Always or almost always (percent): 52.8%; Most of the time (percent): 11.5%; About half of the time (percent): 2.0%; Some of the time (percent): 5.6%; Never or almost never (percent): 13.9%; No response (percent): 14.3%; Number of respondents: 252. [End of table] Q6. When did you receive RMA's request(s) for 2003 growing season inspection(s) (spot checks)? Before planting season; Percent: 40.5%; Number of respondents: 743. During planting season; Percent: 25.2%; Number of respondents: 743. Less than 30 days after final planting date; Percent: 6.5%; Number of respondents: 743. More than 30 days after final planting date but before harvest; Percent: 14.9%; Number of respondents: 743. During or after harvest; Percent: 2.0%; Number of respondents: 743. Don't remember and there is no record of time request was received; Percent: 2.7%; Number of respondents: 743. Other ( Please specify below); Percent: 7.0%; Number of respondents: 743. No response; Percent: 1.1%; Number of respondents: 743. [End of table] Q6 other. If you checked "Other" in Question 6 above, please explain in the textbox below. Writing comment (percent): 8.1%; Number of respondents: 743. [End of table] Q7. During the past 4 years, have you received any training/education on federal crop insurance? Yes (percent): 82.9%; No (percent): 15.5%; No response (percent): 1.6%; Number of respondents: 742. [End of table] Q8. How adequate was the training/education you received on federal crop insurance? Very adequate (percent): 28.6%; Somewhat adequate (percent): 52.8%; Neither adequate nor inadequate (percent): 9.9%; Somewhat inadequate (percent): 5.9%; Very inadequate (percent): 0.8%; No response (percent): 1.9%; Number of respondents: 625. [End of table] Q9. Do you think that you need more training/education on federal crop insurance? Definitely yes (percent): 16.0%; Probably yes (percent): 49.1%; Uncertain (percent): 13.6%; Probably not (percent): 17.9%; Definitely not (percent): 2.7%; No opinion (percent): 0.5%; No response (percent): 0.1%; Number of respondents: 742. [End of table] Q10. During the past 4 years, have you received any training/education (either in-class or on-line) on conducting growing season inspections (spot checks)? Yes (percent): 56.0%; No (percent): 41.2%; No response (percent): 2.8%; Number of respondents: 741. [End of table] Q11. How adequate was the training/education you have received on conducting growing season inspections (spot checks)? Very adequate (percent): 19.6%; Somewhat adequate (percent): 55.2%; Neither adequate nor inadequate (percent): 11.4%; Somewhat inadequate (percent): 6.0%; Very inadequate (percent): 1.1%; No response (percent): 6.7%; Number of respondents: 464. [End of table] Q12. Do you think that you need more training/education on conducting growing season inspections (spot checks)? Definitely yes (percent): 17.3%; Probably yes (percent): 44.3%; Uncertain (percent): 10.0%; Probably not (percent): 23.3%; Definitely not (percent): 4.0%; No opinion (percent): 0.7%; No response (percent): 0.4%; Number of respondents: 741. [End of table] Q13. Do any of the following make it difficult for you to conduct growing season and preharvest inspections? a. Quality of the guidance from FSA%; Yes (percent): 12.5%; No (percent): 78.9%; Uncertain (percent): 6.5%; No response (percent): 2.2%; Number of respondents: 738. b. Quality of the guidance from RMA%; Yes (percent): 25.4%; No (percent): 63.2%; Uncertain (percent): 8.6%; No response (percent): 2.9%; Number of respondents: 736. c. Inspection requests received late%; Yes (percent): 30.7%; No (percent): 62.3%; Uncertain (percent): 3.9%; No response (percent): 3.2%; Number of respondents: 727. d. Not a priority with your supervisor%; Yes (percent): 3.4%; No (percent): 86.3%; Uncertain (percent): 4.8%; No response (percent): 5.6%; Number of respondents: 735. e. Not a priority with state office%; Yes (percent): 3.3%; No (percent): 86.5%; Uncertain (percent): 5.9%; No response (percent): 4.4%; Number of respondents: 735. f. Findings not acted on%; Yes (percent): 17.1%; No (percent): 60.9%; Uncertain (percent): 16.6%; No response (percent): 5.5%; Number of respondents: 731. g. No explanation or follow-up from RMA on findings%; Yes (percent): 33.7%; No (percent): 50.0%; Uncertain (percent): 11.4%; No response (percent): 4.9%; Number of respondents: 736. h. You are uncomfortable in assisting RMA with enforcement%; Yes (percent): 17.9%; No (percent): 75.3%; Uncertain (percent): 5.3%; No response (percent): 1.5%; Number of respondents: 738. i. Not enough time%; Yes (percent): 49.1%; No (percent): 45.7%; Uncertain (percent): 2.8%; No response (percent): 2.4%; Number of respondents: 740. j. Other (Please specify below.)%; Yes (percent): 28.8%; No (percent): 33.5%; Uncertain (percent): 2.1%; No response (percent): 35.6%; Number of respondents: 340. [End of table] Q13 other. If you checked "Other" in Question 13 above, please explain in the textbox below. Writing comment (percent): 13.9%; Number of respondents: 743. [End of table] Q14. Would feedback from RMA following growing season inspections help you be more effective in reducing fraud, waste, and abuse in the crop insurance program in your county? Definitely yes (percent): 34.5%; Probably yes (percent): 44.6%; Uncertain (percent): 11.1%; Probably not (percent): 8.8%; Definitely not (percent): 0.7%; No response (percent): 0.4%; Number of respondents: 740. [End of table] Q15. In your opinion, do any of the following areas of the FSA/RMA Handbook FCIC Program Integrity 4-RM need to be improved? a. Part 1: Basic provisions (pp. 1.1 - 1.7); Yes (percent): 5.6%; No (percent): 70.8%; Uncertain (percent): 17.8%; No response (percent): 5.7%; Number of respondents: 734. b. Part 2: Referrals and Investigations (pp. 2.1 - 2.59); Yes (percent): 17.1%; No (percent): 58.5%; Uncertain (percent): 18.8%; No response (percent): 5.6%; Number of respondents: 733. c. Part 3: Claims Audit (pp. 3.1 - 3.9); Yes (percent): 12.2%; No (percent): 57.1%; Uncertain (percent): 24.0%; No response (percent): 6.8%; Number of respondents: 722. d. Part 5: State Technical Committee (STC) Consultation (pp. 5.1 - 5.59); Yes (percent): 6.6%; No (percent): 57.6%; Uncertain (percent): 26.9%; No response (percent): 8.9%; Number of respondents: 728. e. Part 6: Data Reconciliation (pp. 6.1 - 6.185); Yes (percent): 19.0%; No (percent): 52.9%; Uncertain (percent): 21.4%; No response (percent): 6.7%; Number of respondents: 733. f. Other (Please specify below.); Yes (percent): 8.2%; No (percent): 36.3%; Uncertain (percent): 13.0%; No response (percent): 42.6%; Number of respondents: 331. [End of table] Q15 other. If you checked "Other" in Question 15 above, please explain in the textbox below. Writing comment (percent): 4.4%; Number of respondents: 743. [End of table] Q16. In your opinion, how effective are RMA and FSA in coordinating their efforts to reduce fraud, waste, and abuse in the crop insurance program? Extremely effective (percent): 1.6%; Very effective (percent): 16.2%; Moderately effective (percent): 31.4%; Somewhat effective (percent): 27.6%; Slightly effective or not effective (percent): 17.6%; No opinion (percent): 5.1%; No response (percent): 0.5%; Number of respondents: 740. [End of table] Q17. In your opinion, with respect to reducing fraud, waste, and abuse in the crop insurance program, how effective is the coordination of your State FSA Office with your County Office? Extremely effective (percent): 9.4%; Very effective (percent): 43.3%; Moderately effective (percent): 25.3%; Somewhat effective (percent): 11.3%; Slightly effective or not effective (percent): 6.1%; No opinion (percent): 4.2%; No response (percent): 0.4%; Number of respondents: 742. [End of table] Q18. If you would like to make any additional comments concerning the crop insurance program or RMA's and FSA's efforts to combat fraud, waste, and abuse and ensure integrity and compliance, please enter your comments in the textbox below. Writing comment (percent): 40.8%; Number of respondents: 743. [End of table] [End of section] Appendix IV: Results of Improving Compliance and Integrity in the Federal Crop Insurance Program: Survey of Crop Insurance Agents: This appendix presents the questionnaire that was sent to crop insurance agents and the results of our survey. The results are presented in terms of the estimates of the responses we would have received to the questions in our survey if we had surveyed all agents selling crop insurance. Because of the large number of crop insurance agents, we used a sample (probability sample). Before sampling, agents were grouped by the dollar amount of policies sold in 2003. The number of agents sampled from each group was proportionate to the number of agents in the group. The final sample consisted of 935 agents. The survey results that we present are estimates, each with a measurable precision, or sampling error, that may be expressed as a plus/minus figure. By adding the sampling error to and subtracting it from the estimate, we develop upper and lower bounds for each estimate. The range between these two estimates is called the confidence interval. Sampling errors and confidence intervals are stated at a certain confidence level, in this case, 95 percent. For example, a confidence interval at the 95 percent confidence level means that in 95 out of 100 cases, the sampling procedure would produce a confidence interval that would include the estimated value. [See PDF for image] [End of figure] [End of section] Appendix V: Crop Insurance Fraud Cases Criminally Prosecuted, June 2003 to April 2005: This appendix presents GAO's Office of Forensic Audits and Special Investigations description of eight cases of crop insurance fraud investigated by U.S. Department of Agriculture's (USDA) Office of Inspector General (OIG) and criminally prosecuted between June 2003 and April 2005. Case 1: Based on results from data mining, OIG, and USDA's Farm Service Agency (FSA) and Risk Management Agency (RMA) conducted field inspections on a producer in Tennessee. Problems identified during the inspection of the producer's farm were reported to the insurance company that provided the policy. In 1999, the producer improperly obtained crop insurance coverage for his tomato crop and received a claims payment for losses that had not occurred. The producer was ineligible to participate in the crop insurance program because he had not paid a past premium. In order to hide the fact that he was the true grower of 1999 tomato production in two Tennessee counties, he used his wife's name on crop insurance documents. In addition, his wife filed a report with the insurance company claiming a higher level of acreage planted to inflate the value of any subsequent insurance claim. An insurance adjuster assisted the producer by fraudulently signing forms showing he inspected and measured the crops and that his observations supported the wife's claimed loss. The producer's wife received $57,155 in crop insurance payments for losses claimed on the 1999 tomato crop. In September 2003, the producer pled guilty to six counts of making false statements on loan or credit applications and one count each of conspiracy to defraud the government, making false statements, and making false claims. He was sentenced to 3 years probation and ordered to pay $57,155 in restitution to RMA. Case 2: Following up on OIG hotline complaints, USDA initiated an investigation of a wheat producer in Georgia. In February 2000, the Georgia producer claimed a loss of 400 acres of wheat because of heavy hail damage and a hot and dry growing season. He had previously reported that the wheat had been planted by the planting deadline--December 15, 1999. The lease for the land, however, was not effective until January 2000, and the owner of the land confirmed that he did not allow any crops to be planted before the lease began. Other witnesses corroborated that the wheat on this land was not planted until sometime in early 2000--after the planting deadline. The producer received a $39,826 claims payment for the reported loss. The USDA investigation revealed that the producer's brother worked for the insurance company that insured the wheat crop, and the policy was actually sold through a crop insurance agency owned by the brother's wife. Records subpoenaed from the wife's insurance agency disclosed additional policies issued to the subject. Crop insurance claims payments on those policies for the 2000 crop year exceeded $400,000. The investigation also determined that the insurance adjusters for these claims were contract adjusters for the brother's employer. On June 30, 2003, the producer pled guilty in federal district court to making false statements on loan or credit applications. He was sentenced to 1 day in prison and ordered to pay $39,826 in restitution to RMA. Case 3: An OIG-initiated fraud detection survey of grain elevators disclosed irregularities at a Minnesota grain elevator and led to the investigation of two Minnesota producers. In 1997 and 1998, these producers grew pinto bean and black turtle bean crops that were covered by federal crop insurance. The producers falsely claimed that their crops had been damaged by excessive moisture and underreported the amount of beans that were harvested and sold for processing. They also arranged for the harvested beans to be sold using false names so that the insurance adjuster would not be able to link them to this production. Production information was also omitted from production worksheets. The producers received claims payments of $435,087 for losses. In June 2003, each of the producers pled guilty to felony theft and entered into civil agreements under the False Claims Act. The producers were disqualified from the crop insurance program for 1 year and paid restitution of $435,087 and civil penalties of $10,000. Case 4: This case originated when FSA filed a complaint with RMA about a crop insurance agent who was also a producer in Wilbarger County, Texas. The producer and other family members were in a farming partnership and obtained six crop insurance policies for crop year 1999 that covered about 6,500 acres of wheat, cotton, and grain sorghum. All of the policies were issued through the producer's own insurance agency. The producer gave an adjuster a schedule of insurance forms with predetermined figures to be used on appraisal and production worksheets for the wheat, cotton, and grain crops. The adjuster later admitted that he did not visit any of the fields to conduct appraisals or to verify crop production. In addition, the producer had a seed dealer prepare false receipts to support his claim that he had purchased seed to plant the crops. The producer and adjuster falsified crop insurance loss documents and collected $630,000 in fraudulent claims payments for crops that the producer had not planted. In February 2004, the producer was convicted of multiple counts of making false claims to the government and a related conspiracy charge. He was sentenced to 41 months in prison, to be followed by 3 years of supervised release and ordered to pay $448,000 in restitution to RMA. In November 2004, the insurance adjuster pled guilty to one count of conspiracy and received a sentence of 2 years probation, including in- home confinement and was ordered to pay $447,230 in restitution to RMA. Case 5: RMA investigated the manager of a northwestern Minnesota grain elevator when it noticed suspicious adjustments in grain quality. The manager of the grain elevator was charged with providing North Dakota farmers with false documents that were used to obtain over $1 million in fraudulent crop insurance payments from RMA and over $350,000 in improper crop disaster payments from FSA. The manager persuaded producers to sell their wheat at discounted prices to his grain elevator company and then provided the producers with false documents that purposefully undervalued the wheat's weight and quality. The manager also directed company employees to create samples of severely damaged wheat to mislead insurance adjusters and to create records that misled USDA about the amount of wheat at the elevator. This falsification enabled the producers to fraudulently collect crop insurance and crop disaster payments. The manager was convicted of conspiracy to defraud USDA and two counts of false statements in February 2003 and was subsequently sentenced to 46 months in prison and ordered to make restitution in the amount of $751,758. He also was to remain on supervised release for 3 years after serving his prison term. Case 6: Following an OIG hotline complaint, investigators found that two crop insurance agents had falsified crop insurance documents for a producer, thereby enabling the producer to file false crop insurance claims. For example, the agents backdated the producer's crop insurance application, created a false insurance policy based on a fictitious yield rate, and allowed the producer to report fictitious actual yield rates on their insurance application. On August 30, 2004, the agents pled guilty to one count of conspiracy to submit false statements to RMA. Both agents were sentenced to a year's probation and ordered to pay fines of $5,000 and $2,000, respectively. Case 7: OIG followed up on an FSA referral that an Ohio producer reported different crop yields to FSA and RMA. The producer filed four false crop insurance claims and received payments totaling approximately $19,000 between May and December 1999. The producer claimed losses for crops that he did not own. The producer also inflated an insurance claim for a loss on his corn crop. In September 2004, the producer pled guilty to one count of making a false insurance claim. He was sentenced to 24 months probation, 40 hours of community service, and ordered to pay $2,899 in restitution. Case 8: This case involves fraudulent crop insurance and farm program payments and was triggered by a fraud investigation following a bankruptcy filing in July 2000. Between 1997 and 2003, an Iowa couple, aided by other family members, friends, and employees, executed fraud schemes that resulted in crop insurance payments of $912,364.[Footnote 29] The co-conspirators made false representations. Specifically, they falsely certified to: * a prevented planting loss of 1,478 acres in corn and soybeans in several counties because of excessive moisture; they received about $100,000 in claims payments; and satellite imagery and testimonial evidence later confirmed that this claim was false; and: * a loss of approximately 332 acres of soybeans in one county in Iowa when, in fact, satellite imagery evidence indicated only approximately 71 acres of soybeans was planted; $17,000 was paid on this claim; and RMA also paid $88,000 for false crop losses in three other counties in 2001. During this period, the family was also submitting false claims for crop disaster assistance and received $86,000 in payments. The investigation is continuing, although the co-conspirators have pled guilty to some charges, agreed to pay restitution in some instances, and are barred from participating in USDA's farm programs. Officials involved in the investigation and prosecution of this case note that these fraudulent activities were facilitated by the direct participation of the crop insurance agent who provided the policy to the couple. [End of section] Appendix VI: Comparison of Loss Ratios for Crop Development and Expansion Products: Dollars in millions. Newly developed products: Apple quality option; Years with data: 2001-2003; Total premium: $ 1.7; Liability: $ 19.9; Claims paid: $ 4.4; Loss ratio: 2.6. Avocados; Years with data: 1998-2003; Total premium: $23.1; Liability: $148.3; Claims paid: $1.9; Loss ratio: 0.1. Blueberries; Years with data: 2000-2003; Total premium: $6.7; Liability: $81.4; Claims paid: $3.2; Loss ratio: 0.5. Cherries; Years with data: 1999-2003; Total premium: $22.7; Liability: $247.7; Claims paid: $22.2; Loss ratio: 1.0. Livestock; Years with data: 2002-2003; Total premium: $4.2; Liability: $89.7; Claims paid: $6.2; Loss ratio: 1.5. Malting barley option B[A]; Years with data: 1998-2003; Total premium: $14.0; Liability: $132.9; Claims paid: $38.7; Loss ratio: 2.8. Pecan-revenue; Years with data: 1998-2003; Total premium: $16.9; Liability: $207.0; Claims paid: $7.1; Loss ratio: 0.4. Rangeland-group risk plan; Years with data: 1999-2003; Total premium: $6.9; Liability: $154.9; Claims paid: $29.5; Loss ratio: 4.3. Sweet potatoes; Years with data: 1998-2003; Total premium: $11.6; Liability: $126.3; Claims paid: $47.0; Loss ratio: 4.1. Watermelons; Years with data: 1999; Total premium: $8.8; Liability: $67.1; Claims paid: $51.1; Loss ratio: 5.8. Wheat-income protection; Years with data: 1998-2003; Total premium: $7.9; Liability: $68.8; Claims paid: $31.4; Loss ratio: 4.0. Expansion products: Onions in Georgia; Years with data: 1998-2003; Total premium: $15.2; Liability: $151.5; Claims paid: $21.8; Loss ratio: 1.4. Forage in Minnesota; Years with data: 1998-2003; Total premium: $4.5; Liability: $68.3; Claims paid: $10.3; Loss ratio: 2.3. Soybeans in Oklahoma; Years with data: 1999-2003; Total premium: $1.2; Liability: $6.0; Claims paid: $5.0; Loss ratio: 4.2. Onions in Texas; Years with data: 1998-2003; Total premium: $16.4; Liability: $119.1; Claims paid: $31.5; Loss ratio: 1.9. Peanuts in Texas; Years with data: 1998-2002; Total premium: $1.8; Liability: $11.9; Claims paid: $8.4; Loss ratio: 4.6. Source: RMA. [A] Malting barley option B provides producers who grow malting barley under contract with a brewer additional price and quality insurance beyond feed barley coverage. [End of table] [End of section] Appendix VII: Comments from the U.S. Department of Agriculture: USDA: United States Department of Agriculture: Office of the Secretary: Washington, D.C. 20250: Mr. Robert A. Robinson: Managing Director, Natural Resources and Environment: Government Accountability Office: Washington, DC: Dear Mr. Robinson: Attached is the Farm and Foreign Agricultural Service's (FFAS) response to the draft report titled, "CROP INSURANCE: Actions Needed to Reduce Program's Vulnerability to Fraud, Waste, and Abuse." Thank you for the opportunity to provide comments. If you have any questions regarding our response, please contact Alan Sneeringer at 202-720-8813. Sincerely, Signed by: J.B. Penn: Under Secretary: Farm and Foreign Agricultural Service: Attachment: U.S. Department of Agriculture Statement of Action on the U.S. Government Accountability Office Draft Report GAO-05-528, "Crop Insurance: Actions Needed to Reduce Program's Vulnerability to Fraud, Waste, and Abuse: September 13, 2005: Federal crop insurance protects producers against financial losses caused by natural disasters. The United States Department of Agriculture's (USDA) Risk Management Agency (RMA) administers this program with private insurers. During this study, GAO assessed the effectiveness of USDA's processes to address fraud, waste, and abuse in the program and the extent to which program design may make the program vulnerable to abuse. As a result of the study, GAO recommends that Congress consider the reduction of premium subsidies to producers who repeatedly file questionable claims. GAO also developed eight recommendations for USDA. The following addresses those recommendations. General Comments: The USDA and RMA would like to thank the GAO for the opportunity to review this draft and for the work done by the auditors to assess the Federal crop insurance program and recommend changes for improving the controls on waste, fraud, and abuse. Following are some general comments as well as specific comments to each of your recommendations. The current draft letter to Chairman Collins, as well as the "Highlights" section of the report, contains certain information, references, and conclusions that should be corrected prior to issuing the report to avoid possible confusion by the reader. In listing the four major findings, RMA takes exception in part to the second and fourth items. Regarding "RMA's data analysis excludes the largest farming operations", the text would lead readers to believe that data mining excludes these policies altogether as opposed to the fact that GAO is recommending that we improve our ability to detect errors in reporting interests in farming operations. Furthermore, the highlight statement unequivocally states that these entities "received $74 million in fraudulent claim payments". RMA agrees that the data mining protocol needs to be improved to assess interest discrepancies between RMA and FSA, but we were not provided information that would indicate GAO actually validated "fraudulent" activity on the 21,000 policies. RMA had several discussions with GAO regarding the general misconception that because a policy is identified as being anomalous it does not mean that the policyholder did anything wrong. The statement incorrectly maligns these producers without the benefit of a review. GAO's own analysis in this case shows that several thousand policies had identification numbers that did not match because the number was off by a single digit. RMA experience would indicate that the vast majority of these cases can be attributed to data entry error and are not "fraudulent" in any sense of the word. RMA will conduct some preliminary validation tests on the data obtained from GAO prior to sending these policies to the companies for review. The fourth bullet in the "Highlights" section of the report could also mislead the reader as it indicates that RMA imposed sanctions on 114 cases over four years out of some 3,000 questionable claims identified annually through data mining. The 3000 "suspicious" policyholders were identified based on their historical claim activity. The claims used to identify the producers are not necessarily fraudulent nor is there any information to indicate that they are suitable for sanctions. As stated above, there is a general misconception that policies identified as anomalous are in error. In fact, identified anomalies must be validated and are subject to further review and scrutiny at several levels before they would be considered for sanctions. By law, RMA must have specific evidence to prove every allegation of fraud and for every sanction. The spot-check list that is produced annually by our data mining contractor is used to pre-empt payments by monitoring anomalous producers throughout the growing season for the year they are identified. Presumably, when these producers are notified that they are being watched by FSA and RMA they change whatever behavior caused them to be anomalous, and their indemnities decrease. That has been our experience as noted by GAO, but since not paying a claim to a producer does not constitute an offense, the comparison of these numbers is inappropriate. Furthermore, the GAO draft report recognizes the delays that occur in the sanction process, so to compare policies identified during a specific timeframe to sanctions imposed on policies from prior crop years further compounds the misconception. This is discussed further the response to Recommendation 4 below. GAO Recommendation 1: Ensure that FSA field offices conduct all inspections called for under agency guidance. USDA Response: Resources are not available in FSA to complete 100% of the inspections referred by RMA. Because of FSA's budget cuts and the current staff reductions, it is reasonable to expect that it will not be possible to complete as much work for RMA as it has done in the past without additional funding for FSA. Unless funding is increased, RMA will continue to rely on FSA's local knowledge of producers and conditions in determining which inspections to conduct during the year. GAO Recommendation 2: Ensure that RMA informs FSA field inspectors of the suspect claim patterns that they are to investigate. USDA Response: Meetings were held between RMA and several State FSA Committees and FSA employee associations to present the results of the referral and spot check process, including the program cost savings associated with the results. The 2005 spot check list, submitted to the FSA State Offices and Insurance Providers, included the nature of suspected abuse. RMA and FSA will continue to work together on the spot check process to derive the maximum benefits from the resources expended in this effort. The partnership to improve program integrity between the two agencies and the insurance companies is recognized by all three parties as an effective tool in strengthening the crop insurance program. GAO Recommendation 3: Ensure that inspections are conducted in a timely manner, and that inspection results are reported expeditiously to insurance companies. USDA Response: Currently, RMA only provides the spot check list annually on April 1 of each year for both fall seeded and spring seeded crops. A meeting of the joint RMA/FSA Investigation/Referral Team held February 15, 2005, addressed this issue. Beginning with the 2006 insurance year, two separate lists will be generated to accommodate fall and spring planted crops. This will provide FSA with a timelier list for fall-seeded crops. Other areas of improvement discussed during the joint meeting included the development of a FTP or URL on the web that will allow both agencies to post findings of inspections electronically. Currently, referrals are sent by mail or faxed between agencies that can delay response time. The enhancement of a shared site will also allow the National FSA Office to monitor the status of completed inspections, as identified in this report. Inspection forms are also being modified to allow the use of GPS in documenting the location of where an inspection was completed so digital pictures can be documented by location. Insurance companies will also benefit from the above actions in terms of more complete and timely reports. GAO Recommendation 4: Promulgate regulations to implement its expanded authority under ARPA to impose sanctions. USDA Response: It has been suggested that RMA's lack of regulations to implement its expanded authority under ARPA is a critical impediment to imposing sanctions. This is incorrect. In fact, prior to ARPA, RMA had authority to disqualify producers for up to ten years, and used its suspension and debarment authority under 7 C.F.R. 3017 et seq. to deter fraud committed by producers and agents. In fact, the agency has been imposing sanctions under its ARPA authority each year since 2002. RMA has imposed a total of 67 Sanctions in 2004, an increase of over 300% over 2002 and 2003. In 2005, to date 15 sanctions have been imposed, with an additional 64 pending final disposition. Over two thirds of the pending cases are being considered under ARPA authority. In addition, there has been a learning curve, not only by RMA staff, but also by OGC. Since ARPA was enacted, adjustments to internal procedures have been necessary, but clearly, RMA has made significant and steady progress in both the numbers and types of sanctions imposed. Although promulgation of sanctions regulations will surely enhance RMA sanctions efforts, it is inaccurate to indicate that the lack of these regulations has prohibited RMA from currently using the ARPA-based authority to impose sanctions. As to the decline in the number of sanctions referrals, a learning curve was involved in preparing and referring ARPA-related sanctions. Following the enactment of ARPA, RMA Headquarters was inundated with referrals for potential sanctions cases from all of its Compliance offices, which were eager to test RMA's new authority. However, many cases simply did not meet the standards for a Sanctions case; and/or involved violations that had passed the statute of limitations. As Compliance field staff became more proficient at making referrals, the overall number of referrals declined. While fewer cases were referred in recent years, those cases which were referred were timely (within the statute of limitations) and legally sufficient to impose and defend any proposed sanctions. GAO Recommendation 5: Direct FSA to share producer-derived information with RMA for data mining purposes to administer and enforce the requirements of the crop insurance program. USDA Response: Even though the current RMA analysis during data mining might not contain data as a result of producers not following reporting requirements, the RMA and FSA data for producers and the reported shares were compared in the initial RMA/FSA attempts at data reconciliation. Numerous mismatches were determined. FSA researched and made applicable changes to their records and referred the remaining mismatches to RMA. It is a fact that data mining has been successful for RMA to perform analysis after the fact and to make projections. However, a process is being developed jointly by RMA and FSA to compare the data more immediately after being reported by producers. This eliminates the need to review the data after the fact. This process is being developed within the scope of the Common Component of the Common Information Management System (CIMS). Doing the comparison on a near "real time" basis versus months or years after the data is reported to the agencies will be more effective and efficient. The FSA database containing this information is currently accessible by both RMA and NRCS. Therefore, the GAO recommendation that the Administrators be directed to share the data is a moot point. It is being shared with NRCS. Under CIMS, it is anticipated that as data is reported to one agency, it will be compared to the data reported to the other agency, and a discrepancy report will be created for use by both agencies. This process will be validated in the near future in selected counties. As a result of the GAO work in this area, RMA intends to request the information from FSA in order to compare 2005 producer certifications and identify discrepancies for review. GAO Recommendation 6: Direct RMA to determine if payments have been made to ineligible producers or to entities that failed to disclose ownership interests and, if so, to recover the appropriate amounts. USDA Response: RMA agrees with the recommendation to determine if payments have been made to the 14 ineligible producers and will notify companies to recover any amounts due. RMA will also create a new application that will generate a weekly report to determine and identify policies that were previously accepted in RMA's system, but which were later determined by the company to be ineligible for the crop year for which the policies were accepted. Companies will be notified if any policies are identified. Due to the excessive numbers of policies with incorrect or missing entity interest information, RMA intends to conduct a preliminary review of a small number of policies to validate the information. Provided our review does not reveal any overt data-mining errors or issues, RMA will use the authority under the 2005 Standard Reinsurance Agreement to require the companies review their respective policies and correct any premium or indemnity payments as appropriate. GAO Recommendation 7: Direct RMA to strengthen its oversight of the insurance companies' implementation of the quality control review system. USDA Response: As an integral part of its review of a company's Plan of Operations, the Reinsurance Services Division (RSD) assesses the completeness and adequacy of Exhibit 22, which is a company's quality control plan. For 2006, RSD stepped up the rigor with which it evaluates these plans by developing and using written evaluation procedures. As a result, several companies were required to revise their quality control plans before approval for the 2006 reinsurance year was granted. After RSD evaluates and approves a company's quality control plan, RMA's Compliance Office has the responsibility of detecting failures of a company to implement required quality control measures through National Operational Reviews. Any failures detected are brought to the attention of RSD. RSD then imposes sanctions or penalties on an insurance provider that are appropriate to the quality control failures. RSD is also strengthening its quality control oversight through the development and implementation of a quality performance indicator system. The development of such a system of performance indicators is nearing completion through a RMA-sponsored contract with Trivalent Solutions, Inc. of Arlington Heights, IL. Once this system is operational, RMA will have information that will allow it to detect potential quality problems with individual companies. RSD will be able to identify potential problems to investigate and address in some cases well before discovery through a national operational review. GAO Recommendation 8: Direct RMA to reduce the insurance guarantee or eliminate optional unit coverage for producers who consistently have claims that are irregular in comparison with other producers growing the same crop in the same location. USDA Response: RMA respectfully disagrees with the above-noted GAO recommendation, as it represents a disproportionate response considering the prevalence and importance of yield switching. Consider that in crop year 2003, RMA reinsured 1,241,491 Federal crop insurance policies. [NOTE 1] Of that number, a total of 395,889 policies (31.9%) were indemnified, a relatively typical percentage for the Federal Crop Insurance program. Of the almost 396,000 indemnified policies, RMA identified a total of 2,371 suspicious claims, or 0.60% of the total number of indemnified policies and only 0.19% of the total number of policies in force. Of the 2,371 suspicious claims, 273 had patterns consistent with yield switching activity, i.e., 0.07% of total number of indemnified policies and 0.02% of the total number of policies in force. GAO's recommendation is targeted at a group of policyholders that constitute only 0.02% of BMA's business. RMA also notes that the unit division study [NOTE 2] by Drs. Knight and Coble clearly supports the low incidence of yield switching activity among policyholders. The study also indicates that the current 10% premium surcharge is actuarially appropriate for the additional indemnities that are attributable to optional units. This relatively modest surcharge provides further evidence that the RMA has effective underwriting rules and monitoring programs for use of optional units. Secondly, RMA does not believe that adoption of the GAO recommendation represents a prudent and cost-effective use of Agency resources or taxpayer funds. In order for RMA to reduce the guarantee or eliminate optional unit coverage for producers whose yield history evidences possible yield switching activity, as per the GAO recommendation, RMA would have to either: 1. Promulgate criteria that will invariably have a negative impact on a much larger number of growers whose losses are likely legitimate; or: 2. Gather sufficient evidence that an individual insured has engaged in yield switching activities to support the penalty upon appeal by the insured. The second option, however, is not particularly realistic because once the needed evidence was obtained, RMA would more likely use it to pursue a criminal case for fraud against the insured or alternately, administrative sanctions, rather than simply impose the penalties recommended by GAO. To gather such evidence requires RMA to devote substantial resources to each suspect claimant, including the necessity of conducting a pre-harvest appraisal - potentially applying to about 0.02% of RMA's business. Because of the impact on RMA resources, the Agency generally requires strong indications of egregious violations before an in-depth investigation is launched. Once it does so, the more meaningful approach is that of pursuing criminal or civil penalties for fraud rather than "black listing" for a potential forfeiture of optional units or a reduced guarantee, both of which can be averted by a change in entity. NOTES: [1] RMA Summary of Business, data as of August 29, 2005. [2] "Unit Division Structure Review", Tom Knight, Barry Goodwin, Keith Coble, Roderick Rejesus, Contract Number GS-23F-0167K, completed January 23, 2004. These comments reflect RMA's prior experience gained from the attempt to implement a similar procedure in the early 1990s. Revisiting this intrinsically flawed endeavor, as recommended by GAO, would not appear to be prudent or cost-effective. Other Comments: As discussed, RMA will provide you with technical comments to the draft separately. The following are GAO's comments on the letter from the U.S. Department of Agriculture dated September 15, 2005. GAO Comments: 1. We have clarified the language in the Highlights section of this report to note the distinctions in the analyses. As we state in this report, the U.S. Department of Agriculture's (USDA) Risk Management Agency (RMA) is using data mining to administer and enforce the crop insurance program and to analyze patterns that suggest fraudulent activity, such as unusually high or frequent claims payments. However, RMA's analysis is incomplete with regard to the largest farming operations--those that include multiple partnerships and joint ventures. RMA's analysis excludes comparisons of farming operations' reported ownership interest with data that has been validated by USDA's Farm Service Agency (FSA). Because it does not know the ownership interests in these farming operations, RMA cannot readily identify potential fraud. For example, producers who are members of more than one farming operation may have the opportunity to move production from one operation to another to file unwarranted claims, without RMA's knowledge that these producers participate in more than one farming operation. 2. We have changed the text in the Highlights section to more accurately reflect our findings, deleting the term "fraudulent." As detailed in this report, RMA regulations require policyholders to report individuals and entities with a beneficial interest in their farming operation. If a policyholder fails to disclose an ownership interest, then the policyholder must repay the amount of the claims payment that is proportionate to the interest of the person that was not disclosed. Our findings indicate that USDA should be able to recover up to $74 million in such claims payments for 2003. 3. We have clarified the language in the Highlights section. We recognize in this report that the annual spot-check list of 3,000 questionable claims is used to prevent unwarranted claims payments by monitoring anomalous producers during the growing season for the year they are identified. The 3,000 questionable claims represent producers who consistently file claims and receive payments that are irregular in comparison with other producers growing the same crop in the same location. While not all of these policy irregularities are necessarily sanctionable, the 3,000 questionable claims provide a general reference in comparison with the 114 sanctions RMA imposed from 2001 through 2004. 4. The Agricultural Risk Protection Act of 2000 (ARPA) requires the Secretary of Agriculture to develop and implement a coordinated plan for FSA to assist RMA in the ongoing monitoring of the crop insurance program, including conducting fact-finding into allegations of program fraud, waste, or abuse and reporting the results of any such fact- finding to RMA. USDA guidance states that FSA county offices are to conduct growing season inspections on the larger of the first 10 producers or the top 5 percent of the producers identified by RMA. However, as we report, FSA conducted only 64 percent of the monitoring inspections RMA requested between 2001 and 2004, and FSA offices in nine states did not conduct any of the inspections RMA requested in one or more of these years. Given the potentially high payoff from assisting RMA in monitoring the crop insurance program and conducting inspections, FSA may want to conduct a study to determine the costs and benefits of making staff available for crop insurance inspections. 5. We agree that RMA has imposed sanctions since the enactment of ARPA. However, we continue to believe RMA has not fully exercised its new authority. Under ARPA, RMA has the authority to impose sanctions on agents, loss adjusters, approved insurance providers, and others who willfully and intentionally (1) provide false or inaccurate information or (2) fail to comply with other Federal Crop Insurance Corporation (FCIC) requirements. The number of sanctions imposed by RMA has not increased appreciably since enactment of ARPA. For example, RMA imposed an average of less than 20 sanctions annually from 1996 to 2000, and an average of less than 20 sanctions annually from 2001 to 2005, except for 2004 (67), which was an aberration. While not all questionable claims are necessarily sanctionable, RMA has identified about 3,000 questionable payments annually since beginning data mining in 2001. In discussing this report's findings with RMA in April 2005, officials told us that the number of sanctions has not substantially increased, in part because regulations have not been promulgated under ARPA. According to an official in USDA's Office of General Counsel, RMA had imposed sanctions under the provisions of ARPA that were similar to RMA's previous authority to impose sanctions. RMA had authority, prior to ARPA, to impose sanctions on individuals who willfully and intentionally provide false information and had promulgated regulations for imposing sanctions on this basis. (See 7 C.F.R. §§ 400.454, 718.11, and 1405.8.) However, according to this official, it had not imposed any sanctions under its new authority to do so for failure to comply with an FCIC program requirement. This official indicated that regulations were needed to establish what constituted an FCIC requirement, how USDA will determine that a material violation has occurred, and what process would be followed before imposing sanctions. USDA does not dispute the report's findings that no sanctions have been imposed under this provision of ARPA. 6. Our recommendation that FSA share producer-derived information with RMA for data mining to administer and enforce the requirements of the crop insurance program is based on a current lack of information sharing between FSA and RMA. As we report, many farming operations do not always certify individuals or entities who hold or acquire a beneficial interest of 10 percent or more in the insured operation, as required by RMA regulations. RMA was unaware that these entities had failed to fully disclose ownership interest because it had not been given access by FSA to the data file identifying a producer's ownership interest in other farming operations. Furthermore, although USDA is developing a system--called the Common Information Management System-- -to allow FSA and RMA to share producer information, as of September 2005, USDA had not decided whether the data elements necessary to identify a producer's ownership interest in a farming operation would be included in the new system. Therefore, we continue to believe that FSA needs to make this information available to RMA and that RMA should conduct data mining on this information to identify producers having a pattern of anomalous claims payments. 7. Our recommendation is directed at providing RMA with an additional tool to address producers who seem to experience losses year after year when other similarly situated producers do not. Such a tool would complement RMA's current focus on preventing producers from committing future abuses by providing an incentive to not commit the abuse in the first place. For example, eliminating optional unit coverage for a year for a producer who exhibits a clear pattern of yield switching, as we present in this report, would discourage the abuse. Removing the means that enables a producer to commit abuse may act as a deterrent. Furthermore, the costs of fraud, waste, and abuse to the crop insurance program from farms with optional units can be significant, according to a 2002 RMA study.[Footnote 30] The study estimated the additional indemnities associated with fraud, waste, and abuse were $131 million for 1996 to 2000 for farms insuring with optional units on wheat and barley in North Dakota and cotton in west Texas. We agree with USDA that the number of policies identified annually as having patterns consistent with yield switching is small in comparison with the number of policies in the crop insurance program. However, we believe it is possible to narrowly tailor an underwriting rule so that it would target only a few producers each year and would entail few resources. Such a rule would also reduce reliance on RMA's broad-brush approach of assessing a 10 percent premium surcharge on all producers with optional units to cover the additional indemnities that are attributable to optional units. Thus, we continue to consider it reasonable for RMA to reduce the insurance guarantee or eliminate optional unit coverage for producers who consistently have claims that are irregular in comparison with other producers growing the same crop in the same location. [End of section] Appendix VIII: GAO Contact and Staff Acknowledgments: GAO Contact: Robert A. Robinson (202) 512-3841: Staff Acknowledgments: In addition to the individual named above, Jeanne Barger, Thomas Cook, Ronald E. Maxon Jr., Lynn Musser, Carol Herrnstadt Shulman, and Amy Webbink made key contributions to this report. Important contributions were also made by Stephen Brown and Paul Desaulniers. We also wish to give special tribute to our dear friend and colleague, Cleofas Zapata Jr., who died tragically near the conclusion of our work. Cleo dedicated his life to public service, first as a member of the U.S. Air Force, and then as an analyst for 24 years with GAO. Cleo's career with GAO was characterized by his strong desire to make government programs more effective and efficient and to weed out fraud, waste, and abuse. But it was his humor, kindness, and respect for all human beings that inspired his co-workers, who held him in the highest esteem and miss him greatly. [End of section] Related GAO Products: Crop Insurance: USDA Needs to Improve Oversight of Insurance Companies and Develop a Policy to Address Any Future Insolvencies. [Hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-04-517] Washington, D.C.: June 1, 2004. Crop Insurance: USDA Needs a Better Estimate of Improper Payments to Strengthen Controls of Claims. [Hyperlink, http://www.gao.gov/cgi- bin/getrpt?GAO/RCED-99-266] Washington, D.C.: September 22, 1999. Crop Insurance: USDA's Progress in Expanding Insurance for Specialty Crops. [Hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO/RCED-99-67] Washington, D.C.: April 16, 1999. Crop Insurance: Federal Program Faces Insurability and Design Problems. [Hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO/RCED-93-98] Washington, D.C.: May 24, 1993. Crop Insurance: Program Has Not Fostered Significant Risk-Sharing by Insurance Companies. [Hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO/RCED-92-25] Washington, D.C.: January 13, 1992. (360445): FOOTNOTES [1] According to the USDA Inspector General, fraud is commonly perpetrated through false certification of one or more of the basic data elements essential for determining program eligibility and amounts of benefits. In RMA cases, the scheme typically involves a conspiracy between an insurance company representative and a producer. Abuse is more subjective and occurs when a participant's actions defeat the intent of the program, although no law, regulation, or contract provision is actually violated. Waste, on the other hand, occurs when there are flaws in the program design that inevitably invite abuse by the program participants. GAO has previously reported on the potential for fraud, waste, and abuse in the federal crop insurance program. See Related GAO Products. [2] Pub. L. No. 107-300, 116 Stat. 2350 (2002). [3] In general, RMA permits producers to establish optional units by land section or FSA farm serial number and by irrigated and nonirrigated practices. Optional units may be established only if each optional unit is located on noncontiguous land, unless otherwise allowed by written agreement. In addition, producers who insure all their fields together in a basic unit receive a 10 percent discount on the premium they pay. [4] Additionally, ARPA requires USDA to subsidize revenue insurance products at the same rate as the level of subsidy provided for a basic crop insurance policy. Revenue insurance products provide coverage to producers against lost revenues (or incomes) caused by low prices, low yields, or a combination of low prices and low yields. An indemnity is paid to a producer when any combination of yield and price results in revenue that is less than a pre-specified revenue guarantee. [5] FSA/RMA Handbook, FCIC Program Integrity, 4-RM. [6] FSA/RMA Handbook, FCIC Program Integrity, 4-RM. The final planting date is the date contained in the special provisions for the insured crop by which the crop must initially be planted in order to be insured for the full production guarantee or the amount of insurance per acre. [7] A company official stated that because this producer was on the spot-check list, the company had contacted the local FSA office at the beginning of the growing season requesting copies of growing season inspection reports when they were completed. However, the company did not follow up with FSA at the end of the growing season. In addition, although the company received RMA's referral too late to conduct preharvest appraisals, the company received the information before the producer filed the claim (December 2, 2003) and company paid it (January 16, 2004). However, it does not appear that the company used the referral information to question the producer's claim. [8] The Center for Agribusiness Excellence conducted this analysis at the request of GAO. The Center, located at Tarleton State University in Stephenville, Texas, provides research, training, and resources for data warehousing and data mining of agribusiness and agriculture data. The Center provides data mining of crop insurance data for RMA. [9] 7 C.F.R. § 457.8. [10] In addition, RMA guidance Manual 14, Guidelines and Expectations for Delivery of the Federal Crop Insurance Program states that insurance companies must conduct conflict-of-interest reviews for all crop insurance claims of individuals directly associated with the federal crop insurance program. However, without knowledge that these insurance agents held a substantial beneficial interest of 10 percent or more in entities that received claims payments, insurance companies may not have conducted the reviews in 2003 and 2004. As of August 2005, RMA could not confirm that these reviews had been conducted. [11] We also identified an additional 12 claims adjusters who adjusted 13 policies in 2003 and 2004 with claims payments of $173,292 for farming operations in which they held a beneficial interest of 10 percent or more and who disclosed this information to RMA. In May 2005, we referred the names of these 12 adjusters to RMA for further investigation. RMA found that 11 of the adjusters did not adjust policies for farming operations in which they held a beneficial interest, but that erroneous information in RMA's databases made it appear that the adjusters had engaged in conflict-of-interest activities. As of August 2005, RMA had not completed its investigation for the remaining claims adjuster. [12] RMA Loss Adjustment Manual (LAM) Standards Handbook, 2003 and Succeeding Years. [13] Final Research Report For Multiple Year Coverage, Task Order # RMA- RED-01-06, Watts and Associates, Inc., June 27, 2002. [14] See U.S. Department of Agriculture, Office of Inspector General, Monitoring of RMA's Implementation of Manual 14 Reviews/Quality Control Review System, Audit Report No. 05099-14-KC (Washington, D.C.: Mar. 15, 2002). [15] GAO, Crop Insurance: USDA Needs a Better Estimate of Improper Payments to Strengthen Controls of Claims, GAO/RCED-99-266 (Washington, D.C.: Sept. 22, 1999). [16] The federal government has sought indictments based on a conspiracy to defraud the government or making false statements to the federal government under 18 U.S.C. § 371 and 18 U.S.C. § 1014, respectively. [17] See Office of Management and Budget, Financial Management Systems, Circular No. A-127 Revised, (Washington, D.C.: July 23, 1993). [18] New participants, who have no history of production in the crop insurance programs, get assigned the county average yield for determining their insurance guarantee, which also affects their premium costs. [19] NASS collects and reports production data for major crops in most counties nationwide. RMA uses these data to establish a normal crop yield. [20] RMA regulations state that optional units are not available to a producer who does not provide acceptable production reports for at least the most recent crop year. [21] Final Research Report For Multiple Year Coverage, Task Order # RMA- RED-01-06, Watts and Associates, Inc., June 27, 2002. [22] 7 C.F.R. § 457.8. [23] USDA Risk Management Agency, New Program Development Handbook, FCIC-23010, October 1997. [24] For example, in 1999, RMA introduced a program for sweet cherries, which had not previously been eligible for crop insurance. RMA had experience with the other five products we reviewed but expanded these products to new counties in selected states. For example, in 1998, RMA expanded crop insurance on onions in Texas to producers in 11 new counties that previously did not have the option of insurance for this crop. [25] See U.S. Department of Agriculture, Office of Inspector General, Risk Management Agency Viability of Fall Watermelons in Texas and Their Inclusion in the 1999 Watermelon Insurance Pilot Program, Audit Report No. 05601-8-Te (Washington, D.C.: Sept. 30, 2002). [26] See U.S. Department of Agriculture, Office of Inspector General, Risk Management Agency Survey of Pilot Programs, Audit Report No. 05601- 12-Te (Washington, D.C.: May 24, 2005). [27] In addition to responding to our survey questions, many of these field officials also provided us with written comments. Because these written comments were voluminous, they have not been included in appendix III. [28] In addition to responding to our survey questions, many of these insurance agents provided us with written comments. Because these written comments were voluminous, they have not been included in appendix IV. [29] These producers and others also received $746,700 in federal farm program benefits through fraudulent schemes. [30] Final Research Report For Multiple Year Coverage, Task Order # RMA- RED-01-06, Watts and Associates, Inc., June 27, 2002. GAO's Mission: The Government Accountability Office, the investigative arm of Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO's commitment to good government is reflected in its core values of accountability, integrity, and reliability. Obtaining Copies of GAO Reports and Testimony: The fastest and easiest way to obtain copies of GAO documents at no cost is through the Internet. GAO's Web site ( www.gao.gov ) contains abstracts and full-text files of current reports and testimony and an expanding archive of older products. The Web site features a search engine to help you locate documents using key words and phrases. You can print these documents in their entirety, including charts and other graphics. Each day, GAO issues a list of newly released reports, testimony, and correspondence. GAO posts this list, known as "Today's Reports," on its Web site daily. The list contains links to the full-text document files. To have GAO e-mail this list to you every afternoon, go to www.gao.gov and select "Subscribe to e-mail alerts" under the "Order GAO Products" heading. Order by Mail or Phone: The first copy of each printed report is free. Additional copies are $2 each. A check or money order should be made out to the Superintendent of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more copies mailed to a single address are discounted 25 percent. Orders should be sent to: U.S. Government Accountability Office 441 G Street NW, Room LM Washington, D.C. 20548: To order by Phone: Voice: (202) 512-6000: TDD: (202) 512-2537: Fax: (202) 512-6061: To Report Fraud, Waste, and Abuse in Federal Programs: Contact: Web site: www.gao.gov/fraudnet/fraudnet.htm E-mail: fraudnet@gao.gov Automated answering system: (800) 424-5454 or (202) 512-7470: Public Affairs: Jeff Nelligan, managing director, NelliganJ@gao.gov (202) 512-4800 U.S. Government Accountability Office, 441 G Street NW, Room 7149 Washington, D.C. 20548: