Can insurance provide the US dairy farm safety net?

Wolf, Christopher A.; Widmar, Nicole J. Article Can insurance provide the US dairy farm safety net? Agricultural and Food Economics Provided in Cooperation with: Italian Society of Agricultural Economics (SIDEA) Suggested Citation: Wolf, Christopher A.; Widmar, Nicole J. (2015) : Can insurance provide the US dairy farm safety net?, Agricultural and Food Economics, ISSN 2193-7532, Springer, Heidelberg, Vol. 3, pp. 1-13, https://doi.org/10.1186/s40100-015-0031-8 This Version is available at: https://hdl.handle.net/10419/179049 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ RESEARCH Open Access Can insurance provide the US dairy farm safety net? Christopher A Wolf 1,3* and Nicole J Widmar 2 * Correspondence: [email protected] 1 Department of Agricultural, Food and Resource Economics, Michigan State University, East Lansing, MI 48824, USA 3 317A Morrill Hall of Agriculture, 446 West Circle Drive, East Lansing, MI 48824-1039, USA Full list of author information is available at the end of the article Abstract US farm programs have been moving towards using insurance to provide the farm safety net for many years but this policy transformation has not been realized for dairy farms. Historically, support for dairy farmers focused on milk price but has been declining in real terms for decades. Dairy policy in the US has been in flux because of the recent increases in feed price levels and increases in export markets that have made historic milk price supports essentially irrelevant. Recently a program of subsidized insurance for the margin between milk and feed price has met with limited success. We analyze the potential demand for margin insurance using estimates from a survey of US dairy farmers. Dairy farmers with larger herds and more education were willing to pay more for the insurance. In addition, those dairy farm operators who had used milk and feed price risk instruments in the past were willing to pay more for margin insurance than those who had not. However, the willingness-to-pay for these contracts by farmers who had not used risk management tools was low, calling into question the feasibility of using margin insurance as the foundation of policies intended to support all dairy farms. Keywords: Dairy farms; Dairy policy; Margin insurance; Price support; Willingness-to-pay Introduction U.S. agricultural policy has been experiencing a long-term shift towards insurance and away from traditional commodity programs with price supports and deficiency payments. Total government spending on farm safety net programs which include both commodity programs and crop insurance fell by two-thirds from fiscal years 2000 to 2012 (Crop Insurance in America, 2013). This reduction took place as spending on commodity programs was replaced by crop insurance. Acknowledging this shift the U.S. Secretary of Agriculture stated that “…at the end of the day crop insurance is going to be the lynchpin of the safety net” (Vilsack, 2013). This sentiment was echoed by the Administrator of the U.S. Department of Agriculture Risk Management Agency who noted that in the 2014 farm bill insurance would be “…the only title that will see a significant increase in spending”(Willis, 2014). Congress first authorized crop insurance in the 1930’s in the period after the Great Depression that was the beginning of many agricultural policies that still exist today (Glauber, 2013). Crop insurance was mostly limited to major crops (e.g., corn) in the primary producing areas until the Federal Crop Insurance Act of 1980 (U.S. Department of Agriculture, Risk Management Agency USDA-RMA 2013c). The 1980 Act expanded crops and areas eligible for insurance and authorized a 30 percent subsidy of the © 2015 Wolf and Widmar; licensee Springer. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. Wolf and Widmar Agricultural and Food Economics (2015) 3:14 DOI 10.1186/s40100-015-0031-8 crop insurance premium. Even though crop insurance expanded, ad hoc disaster payments remained common. In 2000 legislation was enacted which expanded the role of the private sector in developing new contracts, premium subsidies were increased, and restrictions on the development of livestock (and dairy) insurance products were removed. The result is that acres covered by crop insurance increased from 45 million in 1981 to 272 million in 2008 (Goodwin and Smith, 2013). The majority of U.S. corn, soybeans and wheat acres have been insured in recent years. Thus, the largest crops by acres cultivated and sales have been largely covered by crop insurance policies. The same cannot be said, however, for specialty crops, livestock, and dairy production. With respect to livestock and dairy enterprises, the development and adoption of insurance following the easing of restrictions in 2000 has been slow. An effort was made following the 2008 Farm Bill to increase availability of insurance instruments for livestock and dairy to provide a foundation for the safety net policies. Livestock Gross Margin Insurance for Dairy (LGM-Dairy) was introduced to a limited set of states in 2008 and expanded broadly in 2010 (Gould and Cabrera, 2011). Dairy farmer participation has been limited and calls into question the ability of this type of voluntary insurance tool to be the foundation of US dairy safety net programs. Key considerations include whether dairy farmers would purchase margin insurance broadly, how much dairy farmers would pay for margin insurance, and what socio-economic factors relate to those farm operators who would purchase margin insurance. In order to answer these questions a survey of dairy farmers was conducted in five states that together accounted for more than 40% of US production. In addition to collecting farm and operator demographic data, contingent valuation questions were employed to allow estimation of producer willingness-to-pay for dairy margin insurance. The next section briefly reviews major US dairy policies and the trend away from government support. The third section discusses the survey and data, respondent demographics, estimation of willingness-to-pay for margin insurance, and examines estimation results and policy implications. The final section summarizes our findings and draws conclusions relevant for dairy producers, policy makers, and agricultural industries more broadly. Background U.S. federal dairy policies in large part originated in Farm Bill legislation of the 1930’s as part of the New Deal (Manchester, 1983; Chouinard, et al., 2010). While there are many policies that relate to the dairy industry, the policies that most directly supported the farm milk price include the dairy price support program and the Milk Income Loss Contract (MILC) program a . The U.S. dairy price support program existed in some form from 1949 until 2014 b . For many years, in particular the 1980’s, the support price was set above market clearing levels resulting in a surplus of milk production (Helmberger and Chen, 1994). After the 1980’s the support price was decreased to a level where it rarely directly interfered with market clearing allowing increased levels of farm milk price volatility (Figure 1). The Dairy Price Support Program (DPSP) was an open offer from the Commodity Credit Corporation (a federal governmental agency) to purchase butter, cheese, and nonfat dry milk at set product prices. Under the DPSP the intention was to support Wolf and Widmar Agricultural and Food Economics (2015) 3:14 Page 2 of 13 farm milk prices at a floor—$9.80 per hundredweight for 3.5% butterfat milk in recent years—by purchasing the aforementioned dairy products. Chavas and Kim (2004) concluded that the Dairy Price Support Program reduced milk price volatility and affected milk price levels even during periods when the support prices were not binding. The DPSP was criticized in recent years for curbing product diversity and innovation, supporting world dairy product prices, and being generally insufficient for U.S. farm prices at current feed costs (Shields, 2011; Food and Agricultural Policy Research Institute Dairy Policy Alliance, 2010). Increased farm milk price volatility and higher feed costs are primary drivers of the recent efforts to support dairy margins between milk and feed prices rather than only milk price. The Milk Income Loss Contract (MILC) program originated in the 2002 Farm Security and Rural Investment Act (2002 Farm Bill). The MILC program was a counter-cyclical dairy income support that paid producers a proportion (45 or 34 percent depending on the year) of the difference between the reference and actual milk prices (i.e., a deficiency payment). Farm payments were initially capped at 2.4 million pounds annually. The MILC program was renewed in the 2008 Farm Bill when payments limits were increased to 2.985 million pounds per farm and an adjuster was added that increased payments when feed costs were above a trigger level (U.S. Department of Agriculture, Farm Service Agency, 2013). The financial assistance provided by the program meant that total milk production would not decline as much when milk prices were depressed as herds would have less incentive to cut back production (Jesse, 2005). Payment limits meant that the MILC program disproportionately assisted smaller herds as they could receive payments on most (or all) of their production. Additionally, increasing herd sizes and concentration of farm-level milk production made the MILC program less relevant to total milk production over time. The average U.S. milk cow produces approximately 20,000 pounds of milk per year, thus the 2.4 million pound annual cap translated to 120 milk cows and the 2.985 million pounds translated to approximately 150 milk cows. As Figure 2 demonstrates, the percent of operations that likely had all milk production covered in the period the MILC program existed has declined over time but still constitute a majority. That is, operations with less than 100 5 10 15 20 25 30 35 40 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 $/cwt Support Price Manufactured Milk Price Figure 1 US real milk support price and real manufactured milk farm price, 1980-October 2013 (2013 dollars), source: U.S. Department of Agriculture, National Agricultural Statistics Service, 2013a. Wolf and Widmar Agricultural and Food Economics (2015) 3:14 Page 3 of 13 milk cows represented 78.3 percent of all operations with milk cows in 2003 and declined to 74.1 percent of production by 2012. Meanwhile, large operations (with 1,000 or more milk cows) represented only 1.5 percent of all operations with milk cows in 2003 and doubled to 3.0 percent in 2012 (U.S. Department of Agriculture and National Agricultural Statistics Service USDA-NASS 2013b). The amount of milk production from the herds completely covered by MILC, however, is a different story. As illustrated by Figure 3, operations with less than 100 milk cows produced 23.7 percent of all milk production in 2003 and only 13.7 percent in 2012 (U.S. Department of Agriculture, National Agricultural Statistics Service, 2013a). Operations with 1,000 or more milk cows produced only 31.5 percent of all milk production in 2003 but a majority, 50.6 percent, in 2012. Thus, for operations producing a majority of U.S. milk production, the income protection offered by the MILC program covered only a fraction of farm-level production. The MILC program was repealed in the 2014 Farm Bill. As can be surmised by the declines in the real price of the dairy price support program and percent of milk production covered by the deficiency payments in the MILC program, government support for U.S. dairy farms has declined in recent years. Using OECD data, we can measure how much this support declined. Figure 4 displays the percent of producer single commodity transfers to U.S. milk producers relative to gross 0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700 0.800 0.900 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 <100 100-199 200-499 500-999 1000+ Figure 2 Percent of operations with milk cows by herd size, 2003 –2012. 0.000 0.100 0.200 0.300 0.400 0.500 0.600 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 <100 100-199 200-499 500-999 1000+ Figure 3 Percent of milk production by herd size, 2003 –2012. Wolf and Widmar Agricultural and Food Economics (2015) 3:14 Page 4 of 13 farm milk receipts from 1986 through 2012 (OECD, 2013). Producer Single Commodity Transfers are defined as the annual monetary value of gross transfers from consumers and taxpayers to agricultural producers, measured at the farm gate level, arising from policies linked to the production of a single commodity such that the producer must produce the designated commodity in order to receive the transfer (OECD, 2013). Percentage Producer Single Commodity Transfers (%PSC) are the commodity transfers as a share of gross receipts for the specific commodity. For U.S. milk producers, the %PSC declined to zero in 2008 and has been below 10 percent in recent years. This decline in price support and direct payments fits the general movement in U.S. agriculture policy to move away from or end direct payments. Coinciding with this decrease in support was an increase in milk and feed price volatility. One measure of the variation of milk and feed prices at the farm level is income over feed cost, a commonly used proxy for dairy farm profitability. The margin between milk price and feed cost is the amount available to pay for all other expenses, including labor and returns to management, capital, and unpaid labor. Measuring income over feed cost as the US all milk price less a weighted cost for corn and soybeans from 1990 through 2012, this monthly income over feed cost measure averaged $11 per hundredweight. From 1990 through 1999 the monthly coefficient of variation—the standard deviation divided by the mean as a measure of percent variation—of this margin was 13.6%. From 2000 through 2012 the coefficient of variation increased to 20.4%. This variation in profit margin occurred because variation increased in both the milk and feed prices. Increased volatility in milk and feed prices in recent years has financially stressed many dairy farms and led to calls for government programs to assist them. Dairy margin insurance has been proposed as one option to hold the dairy farm safety net c . Dairy margin insurance In 2007, the Federal Crop Insurance Corporation approved the establishment of the Livestock Gross Margin insurance program for dairy farms (LGM-Dairy). This program became available in August 2008 for dairy producers in 31 states. In July 2010, LGM- Dairy availability was expanded to all of the lower 48 states (Gould and Cabrera, 2011). LGM-Dairy allows dairy farm operators to purchase insurance to protect against 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 %PSC Figure 4 OECD US milk producer single commodity transfers, 1986–2012. Wolf and Widmar Agricultural and Food Economics (2015) 3:14 Page 5 of 13 unanticipated low gross milk margins defined as the difference between estimated milk revenue and feed costs. Under this insurance policy, an indemnity at the end of coverage period is the difference, if positive, between the market anticipated gross margins at contract purchase and actual gross margins. The milk revenues and feed costs protected are based on current futures contract settle prices that exist at insurance sign-up. Chicago Mercantile Exchange (CME) Class III milk, corn, and soybean meal futures and options markets are used as an information source in the setting of premiums and to determine insurance indemnities at the end of the contract. The contract premiums are established to be actuarially sound in the sense that over the long-run, the expected indemnities paid equal premiums before subsidy is applied (Gould and Cabrera, 2011). Past research has highlighted the importance of premium subsidies for dairy margin insurance (Valvekar et al., 2011). Advantages of LGM-Dairy over traditional milk price supports and deficiency payments include that the insurance program: protects margin between milk and feed; has premiums which mean that farmers share in the cost which should help align behavioral incentives; the premiums are set by milk and feed futures and options market which incorporate current market information; allows for the amounts of milk and feed protected to change over time as the farmer desires; allows dairy farmers to scale it to their needs rather than the contract sizes at the exchanges, and this insurance instrument is expected to have no more than minimal effects on production decisions. In addition, from the farmer’s perspective as there are no actual futures market transactions there are no corresponding futures market margin calls. Disadvantages of LGM-Dairy in holding the dairy farm safety net include: limited availability and funding for premium subsidy; lack of knowledge and understanding in the dairy farm industry regarding the insurance instrument; lack of institutional experience in dairy markets using such instrument products; and the local milk and feed prices can vary substantially from the national benchmarks used (basis risk). LGM-Dairy sales summary statistics are available since the 2008 market year. Over this period the peak sales year was 2010 when the 46.2 million hundredweight of coverage sold represented 2.4% of the amount of milk produced in the U.S. that year (U.S. Department of Agriculture, Risk Management Agency 2014). Part of the low participation was due to lack of funding levels in some periods. Other reasons farmers have not used LGM-Dairy include lack of knowledge and perceived difficulty of use (Wolf and Widmar, 2014). Given this relatively low level of participation, can this instrument hold the safety net? Would this insurance be widely purchased? What are the characteristics of the farms that would be willing to pay for it? And how much would they be willing to pay? In order to answer these questions, we surveyed dairy farms in five major milk producing states. Methods Data were collected using a mail survey of 2,419 dairy farms from California, Florida, Indiana, Michigan, and Wisconsin. The list of licensed milk producers was obtained from each State’s respective Department of Agriculture in April 2012. After adjusting for bad addresses and farm exits the number of operations surveyed by state were 468 from California (25% of operations with a license to ship milk), 115 from Florida (100% Wolf and Widmar Agricultural and Food Economics (2015) 3:14 Page 6 of 13 of licensed operations), 410 from Indiana (25% of licensed operations), 450 from Michigan (25% of licensed operations), and 984 from Wisconsin (10% of licensed operations). Together these states accounted for 40.9% of U.S. milk production in 2012 (U.S. Department of Agriculture and National Agricultural Statistics Service USDANASS 2013b). The tailored survey method was employed with an initial survey in May 2012 followed by a reminder card two weeks later and a second survey two weeks after the reminder (Dillman et al. 2009). Respondents identified dairy farm operator characteristics (age, education, experience), and operation characteristics including herd size, crop acreage, percent of feed purchased, farm business organization, cooperative membership, and farm financial solvency situation. Dairy producer respondents were also asked whether they had utilized milk or feed price instruments at any point in the past. Of the 2,419 farms surveyed, 662 were returned for a 27.4% response rate. By state the response rate ranged from 16.5% in California to 34.1% in Indiana. The survey also included basic information on margin insurance and a question to elicit willingness-to-pay for margin insurance as follows: “Currently, there exists insurance to protect the margin between milk and feed price. In this section, we are interested in whether you would purchase an insurance policy that protects a margin between Class III milk price and corn and soybean meal prices. There are many ways that the margin between milk and feed price can be calculated. LGM-Dairy, for example, uses the Class III milk price futures from the Chicago Mercantile Exchange as well as corn and soybean meal futures from the Chicago Board of Tradetocreateamargin. Consider an insurance policy similar to LGM-Dairy that establishes a minimum income over feed cost for the next 12 months using Class III milk, corn, and soybean meal futures. Themarginguaranteeisthefloorforthedifferencebetweenmilkandfeedpricesgiventhe quantity of feed protected. That is, if the actual margin between closing futures milk and feed prices falls below the margin guarantee, you will be paid the difference. The premium rate is the amount that you will pay—per hundredweight of milk protected—for the insurance regardless of whether you end up collecting an indemnity or not. Experience from previous similar surveys is that people often agree to pay more in hypothetical questions than they would actually pay. It is important that you answer as if you were actually facing these choices in your farm management decisions. If an insurance policy is offered with a margin guarantee between milk price and feed cost of $X/cwt and the premium (price) is $Y/cwt on covered milk production, would you be willing to purchase the insurance?” The statement about previous experience and answering as if it was an actual question was part of a “cheap-talk strategy”to attempt to reduce hypothetical bias by simply informing participants of the bias before their participation in the experiment (Lusk, 2003). Wolf and Widmar Agricultural and Food Economics (2015) 3:14 Page 7 of 13 Because the survey offers a hypothetical decision without the exchange of goods or money, the following cheap talk statement was provided to respondents, “Experience from previous similar surveys is that people often agree to pay more in hypothetical questions than they would actually pay. It is important that you answer as if you were actually facing these choices in your farm management decisions.”This statement is part of a “cheap-talk strategy”to attempt to reduce hypothetical bias by simply informing participants of the bias before their participation in the experiment (Lusk, 2003). Using information for LGM-Dairy insurance that was available at the time the survey was conducted (April 2012), the relevant range of margins and premiums offered by the market at that time was derived—plus an allowance on each side to encompass reasonable possibilities. The result was that the margin (X) varied from $2 to $14 per hundredweight of milk production in $2 increments. The premium (Y) varied from $0.10 to $1.30 per hundredweight of milk in $0.30 increments. With seven margins and five premiums, there were 35 potential combinations which were assigned randomly across the surveys. The probability of a “Yes”response to the single-bound question depends on the probability that the willingness-to-pay (WTP) with random error is greater than the premium given Pr YesðÞ¼Pr WTP þe>$y=cwtðÞ:ð1Þ When eis distributed normally the discrete choice estimation is performed using a probit model. All else equal (i.e., margin), the probability of a “Yes”response is expected to vary inversely with the premium cost. The probit model is defined as Pr(y≠0|x)=Φ(xb)where Φis the standard cumulative normal distribution. The probit model has the estimation form y = b’x + e, where y= 0 for farmers who answered “No”to the insurance purchase question; y= 1 for farmers who answered “Yes”to the insurance purchase question; xis a vector of explanatory variables; and bis a vector of coefficients. Results and discussion Summary characteristics of the sample farms are displayed in Table 1. These statistics reflect expected patterns in operation and operator characteristics consistent with national summaries from USDA. For example, the largest herds on average were managed by respondents in California and Florida. Indiana respondents had the smallest average herd size and youngest average operator age. Operator education was, on average, higher in California and Florida respondents. The percent of sole proprietors (as opposed to partnerships, limited liability companies, and incorporated operations) were higher in the Upper Midwest states than in California and Florida. California and Florida also purchased a larger percent of their feed requirements and had higher debt to asset ratios on average. The operation and operator characteristics summarized in Table 1 are used below as explanatory variables in the estimations to explain margin insurance purchase decisions. Table 1 displays summary statistics conditional on whether the respondents had used any milk or feed price risk management tools to date. About one-third of respondents had used milk or feed price risk tools while the remaining two-thirds had not. Those respondents who had used price risk management tools operated larger farms, both in terms of herd size and acres operated. Mean values for all operator and operation characteristics were statistically different between the two groups with the exception of operator age (t-tests, 95%). Respondents who had used price risk tools also had herds with higher milk production per cow, were much less likely to have their business Wolf and Widmar Agricultural and Food Economics (2015) 3:14 Page 8 of 13