Crop Insurance: Savings Would Result from Program Changes and Greater Use of Data Mining
Federally subsidized crop insurance, which farmers can purchase to help manage the risk inherent in farming, has become one of the most important programs in the farm safety net. Under the federal crop insurance program, farmers can choose various levels and types of insurance protection: they can insure against losses caused by poor crop yields, declines in crop prices, or both, for each insurable crop they produce. In 2011, the crop insurance program provided about $113 billion in insurance coverage for about 264 million acres of farmland, for over 1.1 million policies. The federal government’s crop insurance costs include subsidies to pay for (1) part of a farmer’s crop insurance premiums, which averaged about 62 percent of the total premiums in 2011, and (2) administrative and operating expenses (administrative expenses)—provided on behalf of farmers—to insurance companies to cover their expenses for selling and servicing crop insurance policies. The amount of subsidies—for premiums and administrative expenses—is not limited for individuals or legal entities.
The Congressional Budget Office estimates that, for fiscal years 2013 through 2022, the federal government’s crop insurance costs will average $8.9 billion per year. The cost of the federal crop insurance program has come under increased scrutiny because of the nation’s budgetary pressures, particularly when farm income is at record-high levels. For 2011, the U.S. Department of Agriculture (USDA) reported that 2011 net farm income was a record $98.1 billion. For 2012, USDA estimates that net farm income will decline to $91.7 billion—still the second highest level on record. In addition, according to USDA, the top 5 earnings years for the past 3 decades have occurred since 2004, attesting to the recent profitability of farming. Furthermore, farmland values, another measure of farm prosperity, increased by 85 percent from 2003 through 2011.
We and others have reported over the years on the risks for fraud, waste, and abuse in the crop insurance program and recommended ways to minimize these risks, including examining data on crop insurance claims to identify potential abuses. For example, in 2005, we reported on crop insurance fraud cases investigated by USDA that resulted in criminal prosecutions. These cases showed that the farmers, sometimes in collusion with insurance agents and others, falsely claimed weather damage and low production to receive crop insurance payments. Several of these cases also demonstrated the importance of having USDA’s Farm Service Agency (FSA), which administers many farm programs, and Risk Management Agency (RMA), which administers the federal crop insurance program, work together to identify and share information on questionable farming practices and activities. In part to improve compliance with, and the integrity of, the crop insurance program, Congress enacted the Agricultural Risk Protection Act of 2000 (ARPA). This act provided RMA and FSA with new tools for monitoring and controlling program abuses. Among other things, it required the Secretary of Agriculture to use data mining—a technique for extracting knowledge from large volumes of data—to administer and enforce the crop insurance program. Following USDA’s written procedures, developed pursuant to a requirement in ARPA, RMA provides FSA with a list of farmers who have received payments for anomalous claims—that is, claims that are higher or more frequent than others in the same area and that match RMA scenarios of fraud, waste, or abuse. Under the written procedures, staff in FSA county offices are to inspect the fields of the listed farmers and report the inspection results to RMA.
USDA also administers an array of other farm programs to support farm income, assist farmers after disasters, and conserve natural resources. Unlike the crop insurance program, these other farm programs generally have statutory income and payment limits that apply to individual farmers and legal entities, including corporations, estates, and trusts. For example, USDA provides about $5 billion in fixed annual payments—called direct payments—to farmers based on a farm’s crop production history. However, a person or legal entity with an average adjusted gross farm income (over the preceding 3 tax years) exceeding $750,000 is generally ineligible for direct payments. In addition, for direct payments, the annual payment is generally no more than $40,000 per person or legal entity. In anticipation of the next farm bill, farm groups have made proposals that would result in having crop insurance become the centerpiece of the federal farm safety net, with support through traditional commodity programs playing a significantly reduced role.
In this context, you asked us to identify additional opportunities for reducing the cost of the crop insurance program. Our objectives were to determine (1) the effect on program costs of applying limits on farmers’ federal crop insurance subsidies, as payment limits are applied to other farm programs, and (2) the extent to which USDA has used key data mining tools to prevent and detect fraud, waste, and abuse in the crop insurance program.
To address the first objective, we reviewed eligibility standards, such as adjusted gross income limits and payment limits, in the provisions of the Food, Conservation, and Energy Act of 2008 (2008 farm bill); other statutes; and USDA regulations. We also interviewed FSA and RMA officials regarding eligibility standards and payment limits. To determine the distribution of crop insurance subsidies among farmers who participate in the program, we analyzed RMA data for 2010 and 2011 on the number and percentage of farmers receiving various levels of subsidies and the locations of farmers who received higher subsidies. We selected $40,000 as an example of a potential subsidy limit because it is the payment limit for direct payments, which is one of the largest components of the farm safety net. We also reviewed USDA and others’ studies that examined participation in the crop insurance program and premium subsidies. In addition, we reviewed USDA data on the financial condition of farms of different sizes. To address the second objective, we interviewed officials at RMA headquarters and RMA’s six regional compliance offices to determine RMA’s current uses of data mining results, including data mining related to farmers with anomalous claim payments, as well as insurance agents and adjusters who had anomalous losses in comparison with their peers in the same geographic area. In addition, we analyzed 2009 and 2010 data on FSA’s completion of field inspections, pursuant to RMA’s data mining list of farmers with anomalous claim payments. We also interviewed officials at FSA headquarters and five FSA state offices—California, Colorado, Florida, North Dakota, and Texas—to obtain information about field inspection processes and obstacles to the completion of these inspections. We selected FSA’s North Dakota office because of its high completion rate of field inspections (96 percent) for 2009 and 2010 and large number of requests for field inspections (378). We selected the other four state offices because, over the 2-year period, they had low completion rates of field inspections (less than 33 percent) and at least 80 requests for field inspections. A more detailed discussion of our scope and methodology is presented in appendix I.
We conducted this performance audit from January 2011 to March 2012 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives.