SELECTED TOPICS IN PEANUT PRODUCTION: ECONOMIC FEASIBILITY OF AN ENERGY CROP ON A SOUTH ALABAMA COTTON-PEANUT FARM AND DO ECONOMIES OF SIZE EXIST ON PEANUT FARMS IN THE SOUTHEAST? Except where reference is made to the work of others, the work described in this thesis is my own or was done in collaboration with my advisory committee. This thesis does not include proprietary or classified information. ________________________________________ Edward Todd Frank Certificate of Approval: ______________________ ______________________ C. Robert Taylor Patricia A. Duffy, Chair Professor Professor Agricultural Economics Agricultural Economics and Rural Sociology and Rural Sociology ______________________ ______________________ David Bransby Stephen L. McFarland Professor Acting Dean Agronomy and Soils Graduate School SELECTED TOPICS IN PEANUT PRODUCTION: ECONOMIC FEASIBILITY OF AN ENERGY CROP ON A SOUTH ALABAMA COTTON-PEANUT FARM AND DO ECONOMIES OF SIZE EXIST ON PEANUT FARMS IN THE SOUTHEAST? Edward Todd Frank A Thesis Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirements for the Degree of Master of Science Auburn, AL December 16, 2005 iii SELECTED TOPICS IN PEANUT PRODUCTION: ECONOMIC FEASIBILITY OF AN ENERGY CROP ON A SOUTH ALABAMA COTTON-PEANUT FARM AND DO ECONOMIES OF SIZE EXIST ON PEANUT FARMS IN THE SOUTHEAST? Edward Todd Frank Permission is granted to Auburn University to make copies of this thesis at its discretion, upon request of individuals or institutions and at their expense. The author reserves all publication rights. _________________________ Signature of Author _________________________ Date of Graduation iv VITA Edward Todd Frank, son of Edward and Patsy Frank, was born on December 6, 1979 in Elberta, Alabama. He graduated from Foley High School in 1998 and entered Auburn University in September 1998. He graduated with a Bachelor of Science in Agricultural Economics with a minor in Agronomy and Soils in May 2002. He entered Graduate School in August of 2002 and was employed as a Graduate Research Assistant in the Department of Agricultural Economics and Rural Sociology. v THESIS ABSTRACT SELECTED TOPICS IN PEANUT PRODUCTION: ECONOMIC FEASIBILITY OF AN ENERGY CROP ON A SOUTH ALABAMA COTTON-PEANUT FARM AND DO ECONOMIES OF SIZE EXIST ON PEANUT FARMS IN THE SOUTHEAST? Edward Todd Frank Master of Science, December 16, 2005 (B.S., Auburn University, 2002) 53 Typed Pages Directed by Patricia A. Duffy The possibility of low prices farmers receive for peanuts at the farm gate may cause them to implement alternative production practices. This thesis is comprised of two separate papers, formatted for publication with the help of Patricia Duffy, Robert Taylor, and David Bransby, for the purpose of analyzing two possible alternatives. In the first paper, linear programming and enterprise budgeting are used to analyze possible crop rotations that include a bioenergy crop (velvet bean) on a cotton-peanut farm in southeast Alabama. The price for velvet bean is parameterized from a point where a velvet bean- peanut rotation is optimal and then increased to a point where continuous velvet bean vi production is optimal. The second paper uses a regression model to estimate the influence of peanut acres, education, age, and primary occupation on average cost per unit of output and to determine whether economies of size exist on peanut farms located in the Southeast. vii ACKNOWLEDGMENTS The author would like to thank his committee chair, Dr. Patricia Duffy, for her help on the research and writing process of the thesis. Also, she had provided a tremendous amount of guidance and encouragement during the tenure that author worked on graduate studies. Thanks are also given to the committee members Dr. C. Robert Taylor and Dr. Bransby for their insight and help during the thesis process. Finally, the author would like to his parents for their encouragement and guidance. viii Style manual or journal used: Journal of Agricultural & Applied Economics Computer software used: Microsoft Word 2002, Microsoft Excel 2002, and SAS version 9.1 ix TABLE OF CONTENTS LIST OF TABLES xi INTRODUCTION 1 CHAPTER 1: ECONOMIC FEASIBILITY OF AN ENERGY CROP ON A SOUTH ALABAMA COTTON-PEANUT FARM 3 INTRODUCTION DATA AND METHOD 7 RESULTS 9 DISCUSSION 10 REFERENCE 11 CHAPTER 2: DO ECONOMIES OF SIZE EXIST ON PEANUT FARMS IN THE SOUTHEAST? 20 INTRODUCTION PROBLEM STATEMENT 24 CONCEPTUAL MODEL 26 EMPIRICAL SPECIFICATION OF THE MODEL 26 DATA 27 REGRESSION RESULT 30 DISCUSSION 32 x REFERENCES OVERALL CONCLUSION 34 41 xi LIST OF TABLES Table 1.1: Actual Yields of Cotton, Peanut, and Velvet Beans Grown Grown on Wiregrass Experiment Farm Located in Headland Alabama???????????????????????????13 Table 1.2: Yields from Experiment Farm Adjusted to Stimulate Real Farm Condition????????????????????????.14 Table 1.3: Enterprise Budget for Velvet bean/Grain Sorghum Mix????????..15 Table 1.4: Linear Programming Model for Velvet Bean, Cotton, and Peanut Crops Grown in South Alabama??????????????...16 Table 1.5: Results of Linear Programming Model for Parameterized Velvet Bean Prices???????????????????????17 Table 1.6: Results of Price Sensitivity Analysis for Velvet Bean-Velvet Bean-Peanut Rotation?????????????????????...18 Table 1.7: Results of Price Sensitivity Analysis for Continuous Velvet Bean Rotation?????????????????????????19 Table 2.1: Estimated Average Variable Cost per Pound of Peanuts Produced on Dryland Farms???????????????????.35 Table 2.2: Estimated Average Fixed Cost per Pound of Peanuts Produced on Dryland Farms???????????????????????.36 Table 2.3: Estimated Average Total Cost per Pound of Peanuts Produced on Dryland Farms???????????????????????.37 Table 2.4: Estimated Average Variable Cost per Pound of Peanuts Produced on Irrigated Farms???????????????????.38 Table 2.5: Estimated Average Fixed Cost per Pound of Peanuts Produced on Irrigated Farms???????????????????????.39 Table 2.6: Estimated Average Total Cost per Pound of Peanuts Produced on Peanut Farms????????????????????????40 1 INTRODUCTION Peanut farms in southeast Alabama (the Wiregrass) have had to face low peanut prices since The Farm Security and Rural Investment Act of 2002 which eliminated the production quota system for peanuts. Without the quota system, peanuts are sold on the open market, which can be very volatile. Also lowering the price of peanuts is the expansion of peanut production in areas where peanuts were not grown. Producers in these new areas of production began growing peanuts as a result of lower cotton and grain prices, land that was free of peanut pests, and the elimination of the quota system. With little disease pressure, producers in these areas had, on average, higher yields compared to traditional peanut producing counties in southeast Alabama. The increase in peanut acres and the elimination of the quota system has decreased the average price of peanuts from $532 per ton in the 2000 marketing year to $328 per ton in 2002, according to the Alabama Agricultural Statistic 2004 Bulletin. With lower prices, peanut producers in southeast Alabama may need to look at alternative production practices. Producers may look into crop rotations with non-traditional crops that may be more profitable and may increase peanut yields. One crop that could benefit producers is velvet beans. In the section ?Economic Feasibility of an Energy Crop on a South Alabama Cotton-Peanut Farm,? the economic feasibility of velvet bean, peanut, and cotton rotation is analyzed. A rotation with velvet bean could be profitable for producers if there is an 2 increase in demand for its biomass. If new markets open, such as feedstock for energy production, producers could switch production from continuous peanuts or cotton to velvet bean rotations. Also, research has shown that velvet bean can improve soil productivity and negatively affect soil pests. Velvet beans suppress or control organisms such as root-knot nematode, Reniform nematode, and white mold. By decreasing the incidence of soil pests, producers could possibly lower production costs by applying fewer insecticides and fungicides to peanuts or cotton. Also producers could realize an increase in peanut or cotton yields with this rotation. In the chapter, a level of prices received for velvet bean biomass is parameterized to maximize velvet bean-peanut, velvet bean-velvet bean-peanut and continuous velvet bean rotations. Peanut producers in southeast Alabama could also decide whether or not an increase in peanut acres would help increase net farm income. In the section ?Do Economies of Size Exist on Peanut Farms in the Southeast?,? an average cost per unit of output (pounds of peanuts) model is used to decide whether economies of size exist on peanut farms in the Southeast region of the United States. Variables that may also influence average cost per unit of output include total peanut acres, education, age of the producer, and the primary occupation of the producer. 3 ECONOMIC FEASIBILTY OF AN ENERGY CROP ON A SOUTH ALABAMA COTTON-PEANUT FARM Introduction The U.S. Department of Energy (DOE) is seeking alternative fuel sources, such as energy crops, so that the United States can become less reliant on foreign fuel and also to reduce pollution associated with fossil fuels. The biomass from velvet bean production can be used as an alternative source of fuel. Also, because of velvet beans' effects on soil pests and its soil building properties, a rotation with velvet beans can increase the yields of peanuts and cotton. Unfortunately, there is not currently enough information about the net returns of velvet bean biomass to allow producers to make educated decisions about including this crop in a rotation. The objectives of this paper are to analyze the economic attractiveness of producing velvet beans in south Alabama and to find the prices at which velvet beans would be a competitive source of income for producers in this area. The DOE has been searching for alternative energy sources since the energy crisis of the 1970?s. Recently, the department has been looking at ways to meet the demands of electrical generation facilities. In early tests, waste biomass feedstocks -- such as logging residues, wood processing mill residue, urban wood wastes, and selected agricultural residues -- were used to meet some of the demand. Currently, certain locations have 4 limited quantities of these biomass feedstocks available. To meet the demand of these electrical generation facilities and to expand the biomass industry, other sources of biomass must be considered. Biomass supply from crop residue could be used to meet these needs. Crop residue could potentially displace about 12.5 percent of petroleum imports or 5 percent of electricity consumption (Gallagher et al., 2003). One crop that could be used for its biomass qualities is velvet beans. The velvet bean was first produced in the southern portion of the United States in the late 1800?s. Its uses have included livestock feeding and grazing, and it has been recognized as having the potential to improve soil productivity and to have a depressive effect on soil pests. Velvet beans suppress root-knot nematode (Meloidogyne arenaria) in cotton, peanut, soybeans, and other crops. The crop also provides some suppression of Southern blight, (Sclerotium rolfsii, also known as White mold), and it is the only known means of control of Reniform nematode (Rotylenchulus reniformis). Articles in old literature suggest that velvet beans can be used to control weeds such as bermudagrass and Johnson grass, but there are no current experiments in Alabama that back these claims (Taylor and Rodriguez-Kabana, 1998). On average, velvet beans produce about 15,692 kilograms of biomass per hectare (7 ton/ac). Recently southeastern producers have turned away from growing velvet beans since there is a limited market for the crop, strictly involving livestock feeding. Instead, producers exclusively grow conventional crops, such as peanuts and cotton, which have higher net returns at current prices. A peanut-cotton rotation is a fairly common practice in southeastern Alabama. 5 The benefits of crop rotation have been well documented in professional journals. Taylor and Rodriguez-Kabana (1999a) noted the suppression of root-knot nematode and Southern blight fungus in cotton-peanut and velvet bean-peanut rotations. Both organisms can be suppressed with pesticides, but control with this method is becoming less efficacious and more expensive. According to their findings, the limited effect of peanuts following one year of cotton increased the peanut yield by 691.9 kg/ha and two years of cotton followed by peanuts increased peanut yields by 1227 kg/ha. One year of velvet beans preceding peanuts increases peanut yields by 654.5 kg/ha, while two years of the crop prior to peanuts increases the peanut yield by 1096 kg/ha (Taylor and Rodriguez-Kabana, 1999a). Conway (1996) states that ?one of the major objectives of sustainable agricultural systems is to reduce inputs into crop production.? One way to lower inputs is to include crop rotations in monoculture agricultural systems. He states that crop rotation is a natural type of soil sanitation. Conway referenced Sumner, Doupnik, and Boosalis (1981) when showing the benefits of 2-4 year rotations of non-host crops (such as corn or sorghum) with peanuts, which can significantly decrease inoculum of Southern blight fungus following a severe outbreak and reduce the occurrence of root-knot nematodes (Conway, 1996). Under the Farm Security and Rural Investment Act of 2002, producers enjoy greater flexibility with respect to their planting decisions. Under the 1996 FAIR Act, cotton program payments were decoupled from planting decisions, but peanuts remained under a marketing quota system. The Farm Security and Rural Investment Act of 2002 substantially changed the farm policy for peanuts. Peanut quotas were eliminated, and 6 the current program provides for payments to producers based on past production. Producers may now plant alternative crops on their historical peanut acreage without loss of farm program payments. Also, there are sections in the current farm bill that may benefit producers such as subsidies that promote soil conservation and environmental awareness. An alternative use for velvet bean, such as for a biomass feedstock, improved soil quality, and higher subsequent cotton and peanut yields, may help persuade producers to include this low-input crop in their current crop rotations. In addition, the 2002 Farm Bill contains a section that allocates funds for biomass research and development. Nineteen projects received $23 million in fiscal year 2003. A total of $22 million is offered for 2004 (Duffield and Shapouri, 2003). Other legislation also affects the economic feasibility of energy crops. The Energy Policy Act of 1992 established a 10-year 1.5 cents per kilowatt-hour production tax credit (PTC) for privately owned as well as investor-owned wind projects and biomass plants using dedicated crops (closed-loop) brought on-line between 1993 and 1994, respectively, and June 30, 1999. In addition, the Act instituted the Renewable Energy Production Incentive (REPI), which provides a 1.5 cents per kilowatt-hour incentive, for generation from biomass and other forms of renewable energy, to tax- exempt publicly owned utilities and rural cooperatives (Moore, 1996). The credit has been adjusted upward for inflation, to reach 1.8 cents per kilowatt-hour. The Economic Security and Recovery Act of 2001 included a two-year extension of the credits, which are now due to expire in December 2003. The Energy Policy Act of 2003, which contains provisions affecting biomass sources for energy, is currently stalled in the Senate. 7 Energy subsidies, provided to plants for use of co-fired biomass, could raise demand for biomass to the point where velvet beans become economically attractive. An issue that has been studied on only a limited basis is how much electric plants would have to pay agricultural producers to ensure an adequate supply of this crop. Data and Methods Enterprise budgeting and linear programming were used to find the prices at which velvet beans become profitable crop alternatives on south Alabama cotton and peanut farms. Yield data for the enterprise budgets came from field tests performed from 1993 to 1998 at the Wiregrass Substation in Headland, Alabama. Rotations considered were: continuous peanut, velvet bean-peanut, cotton-peanut, velvet bean-velvet bean- peanut, and cotton-cotton-peanut. The experimental yields are reported in Table 1.1. Energy crop yield data were not collected for many of the trials, as indicated. For the purpose of the model, the experimental yields were adjusted downward to reflect farm condition yields. Peanut yields were deflated by 22 percent to represent an expected yield under producer?s field conditions (Taylor and Rodriguez-Kabana, 1999a). Machine-picked lint cotton yield was only 20 percent of the hand picked seed cotton yield. The yield from machine-picked cotton is only 50 percent of hand-picked seed cotton and lint cotton is only about 39 percent of seed cotton yield (Taylor and Rodriguez-Kabana, 1999b). The adjusted yields are reported in Table 1.2. The adjusted yields were used to compute the percentage increase in yields resulting from crop rotation effects. 8 Alabama Cooperative Extension System enterprise budgets for south Alabama were used for the variable and fixed costs of producing cotton and peanuts, as well as for the base yields (without rotation) of these crops. Since there is no recent enterprise budget for velvet beans, one was developed using information provided by the Alabama Cooperative Extension System and from the Alabama Agricultural Experiment Station. A yield of 15692 kilograms (about 7 ton/ac) per hectare was used in the budget, the approximate yield achieved in the experimental plots. Because velvet beans need some type of structural support to grow upright, they are intercropped with sorghum and the cost of sorghum seed is included in the budget. The velvet bean budget is reported in Table 1.3. In the budgets, the expected price for peanuts was set at $0.44 ($400/ton) per kilogram and price for lint cotton was $1.28 ($0.58/lb) per kilogram. The price received for dry velvet bean biomass was parameterized, over the range $0.033 ($30/ton) per kilogram to $0.05 ($45/ton). The returns above variable costs from the enterprise budgets were imported into a linear programming model, which included the rotation restrictions under the assumption of equilibrium cropping patterns, and a land restriction. The land available for production was limited to 404.7 hectares (about 1000 acres), which is a representative, average-size, commercial cotton-peanut farm in the southeast region of Alabama. The linear programming model, for the "base" situation of $0.033 per kilogram biomass, is reported in Table 1.4. In the model, with the yield of biomass held constant at 15692 kg/ha, the price for a kilogram of biomass began at the base level of $0.033 and increased parametrically 9 until velvet beans entered the solution as part of a rotation, and then again until continuous velvet bean was found to be optimal. To find price sensitivity, the price received for lint cotton was then varied over the range of $0.99/kg ($0.45/lb) to $1.54/kg ($0.70/lb) and the biomass yield was varied from 8,967 kg/ha (4 ton/ac) to 20,175 kg/ha (9 ton/ac). Results Model results are reported in Table 1.5. When the energy crop is priced below $0.0396 per kilogram of biomass, the farm would select a cotton-cotton-peanut rotation. When the energy crop is priced between $0.0396 and $0.0397 per kilogram of biomass, a velvet bean-velvet bean-peanut rotation is selected. At or above $0.0398 per kilogram, it would be optimal to plant velvet beans only. The increased energy crop prices were associated with increased returns above variable cost for the farm. Table 1.6 and 1.7 show the price sensitivity analysis of the velvet bean-velvet bean- peanut and continuous velvet bean rotations. When cotton prices reach $1.32 per kg., continuous cotton and continuous velvet bean rotations compete for resources. As cotton prices decrease below $1.32 per kg., the velvet bean- velvet bean-peanut and cotton- cotton-peanut rotations are competing for the limited resource (land). The decision between the rotations will depend on the price received for the velvet bean biomass and the potential biomass yield per hectare. Results of this analysis support those found by De La Torre Ugarte et al. (2003). Using a market-level simulation tool (POLYSYS), they found that at a price of $0.044 per dry ton for switchgrass, 17 million hectares of cropped, idled, pasture or CRP acres 10 would be converted to biomass production. Our farm-level analysis discovered a similar energy price would be necessary to convert traditional cropland to biomass production. Discussion Velvet beans are not economically attractive in a rotation on a representative south Alabama cotton and peanut farm until the price reaches $0.0396 per kilogram of biomass. Producers will have to be offered at least this price at the farm gate before they can consider the rotation above a cotton-cotton-peanut rotation. For producers to grow only velvet beans, the price must increase to $0.0398 per kilogram of biomass. As velvet bean yields vary and the price of cotton changes, price received for velvet bean will be the determining factor when producers have to make planting decision. At a lint cotton price of $1.32/kg, the prices at which velvet bean biomass must reach is between $0.032 and $0.73 per kilogram, depending on the biomass yield (between 8,967 and 20,175 kg/ha). As cotton prices increase, the price for the biomass must also increase so that the velvet bean rotation can remain competitive for limited resources. Depending on the distance of a farm from a power plant and the outcome of the pending federal legislation, which includes incentives for co-firing biomass, prices in this range of those found in this study to induce velvet bean production could be realized in the near future. 11 References Alabama Cooperative Extension System. 2003 Budgets for Major Row Crops in Alabama. Auburn, AL: Auburn University, Department of Agricultural Economics and Rural Sociology, AGEC BUD 5-1 May. Conway, Kenneth E. ?An Overview of the Influence of Sustainable Agricultural Systems on Plant Diseases.? Crop Protection 15 (1996):223-228. De La Torre Ugarte, Daniel G., Hosein Shapouri, Marie E. Walsh and Stephen P. Slinsky. The Economic Impacts of Bioenergy Crop Production on U.S. Agriculture. Washington, D.C.: USDA. February, 2003. Duffield, J., and H. Shapouri. ?Bioenergy Policy Options.? Presentation at the National Public Policy Eductation Conference, September 21-24, 2003, Salt Lake City, Utah. USDA. Gallagher, P., M. Dikeman, J. Fritz, E. Wailes, W. Gauther, and H. Shapouri. Biomass from Crop Residues: Cost and Supply Estimates. U.S. Department of Agriculture, Office of Chief Economist, Office of Energy Policy and New Uses. Agricultural Economic Report No. 819, 2003. Moore, T. ?Harvesting the Benefits of Biomass.? EPRI Journal. (1996): 16-25. Nienow, S., K.T. McNamara, A.R. Gillespie, and P.V. Preckel. ?A Model for the Economic Evaluation of Plantation Biomass Production for Co-firing with Coal in Electricity Production.? Agricultural and Resource Economics Review. April 1999:106-118. 12 Sumner, D.R., B. Doupnik, Jr., and M.G. Boosalis. ?Effects of Reduced Tillage and Multiple Cropping on Plant Diseases.? Annual Review of Phytopathology. 19 (1981): 167-187. Taylor, C.R., and R.R. Rodriguez-Kabana. History of the Alabama Velvet Bean and Its Potential for the Future. Auburn, AL: Auburn University, Farming Systems Research Forum, No. 2, 1998. Taylor, C.R., and R.R. Rodriguez-Kabana. ?Optimal Rotation of Peanuts and Cotton to Manage Soil-borne Organisms.? Agricultural Systems, 61 (1999a): 57-68. Taylor, C.R., R.R. Rodriguez-Kabana. ?Population Dynamics and Crop Yield Effects of Nematodes and White Mold in Peanuts, Cotton, and Velvet Beans.? Agricultural Systems, 59 (1999b): 177-191. Table 1.1. Actual Yields of Cotton, Peanut, and Velvet Beans Grown on Wiregrass Experiment Farm Located in Headland, Alabama Ye ar: Yield Crop Yield Crop Yield Crop Yield Crop Yield Crop Yield Crop 2 2875.2 P 4665.4 P 4041.5 P 4231.4 P 2468.3 P 2983.7 P 4 - VB 5424.9 P - VB 4801 P - VB 3526.2 P 6 2550.1 CT 5207.9 P 3987.3 CT 4855.3 P 2902.3 CT 3634.7 P 8 - VB - VB 5560.5 P 15340.3 VB - VB 3987.3 P 9 2658.6 CT 2142.8 CT 5587.6 P 2983.7 CT 2631.1 CT 4068.7 P 1997 Pl o t N u m b e r 1993 19981994 1995 1996 Yield in kilogram per hectare. P = peanuts. VB = velvet beans. CT = cotton. 13 Ye ar: Yield Crop Yield Crop Yield Crop Yield Crop Yield Crop Yield Crop 2 2242.7 P 3639 P 3152.4 P 3300.5 P 1924.6 P 2327.3 P 4 - VB 4231.4 P - VB 3744.9 P - VB 2750.4 P 6 510 CT 4062.7 P 797.5 CT 3787.1 P 580.5 CT 2835.1 P 8 - VB - VB 4337.2 P 15340.3 VB - VB 3110.1 P 9 531.6 CT 428.6 CT 4358.4 P 596.7 CT 526.2 CT 3173.7 P 1993 1994 1995 1996 1997 1998 Pl o t N u m b e r Table 1.2. Yields from Experiment Farm Adjusted to Simulate Real Farm Conditions Yield in kilogram per hectare. P = peanuts. VB = velvet beans. CT = cotton. 14 15 Table 1.3. Enterprise Budget for Velvet Beans/Grain Sorghum Mix Velvet bean for biomass ESTIMATED ANNUAL COSTS PER ACRE FOLLOWING RECOMMENDED MANAGEMENT PRACTICES 7 TON YIELD; ENTERPRISE ACREAGE ======> 1 <===== PRICE OR TOTAL ITEM UNIT QUANTITY COST/UNIT PER ACRE Gross 10 VARIA SEED (v 30 See 6 FER NIT 0 P 0 P 0 LIM 7.425 HER 0 INS 0 TRA 42.86 LA 35.71 INTER 4.57 TO 126.57 Incom 83.43 Receipts Velvetbean/grain sorghum Ton 7 30 2 BLE COSTS elvetbean) 1 30 d (sorghum) LBS. 5 1.2 TILIZER ROGEN LBS. 0 0.28 HOSPHATE LBS. 0 0.2 OTASH LBS. 0 0.15 E (PRORATED) TONS 0.33 22.5 BICIDE ACRE 0 6.75 ECTICIDE ACRE 1 0 CTORS & EQUIPMENT ACRE 1 42.86 BOR (WAGES & FRINGE) HOUR 5.49 6.5 EST ON OP. CAP. DOL. 61.00 0.075 TAL VARIABLE COST e Above Variable Costs Costs of production drawn from Alabama Cooperative Extension System budgets. Table 1.4. Linear Programming Model for Velvet Bean, Cotton, and Peanut Crops in South Alabama year 1 year 2 year 3 vb conpt ct pbrot pcrot p2brot p3c2c1 RHS Objective 83.43 34.82 152.51 103.25 103.64 187.40 194.69 MAX bprotc -1 1 LE 0 cprotcon -1 1 LE 0 bbprotc -0.5 1 LE 0 ccprotc -0.5 1 0 LE Land 1 1 1 1 1 1 1 1000LE 16 17 able 1.5. Results of Linear Programming Model for Parameterized Velvet Bean rices T P Price Per kg. Optimal Rotation Optimal Gross Net of Velvet bean Biomass Return on 1000 Acres Less than $0.0396 Cotton-Cotton-Peanut $125,367.20 Between $0.0396 and $0.0397 Velvet bean-Velvet bean- Peanut $125,413.90-$125,927.20 $0.0398 and above Velvet bean-Velvet bean- Velvet bean $125,985.80 Table 1.6. Results of Price Sensitivity Analysis for Velvet Bean-Velvet Bean-Peanut Rotation $ 18 0.99 .043 $1.10 $1.21 $1.32 $1.43 $1.54 0.053 0.063 0.073 0.083 0.094 11,208 0.034 0.042 0.051 0.058 0.066 0.075 13,450 0.029 0.035 0.042 0.049 0.055 0.062 0.030 0.036 0.042 0.047 0.053 0.026 0.032 0.036 0.042 0.046 20,175 0.019 0.023 0.028 0.032 0.036 0.042 Lint Cotton Prices (Dollars/kg.) Biom ield (kg 8,967 / 0 15, 17, 692 934 0 0 .024 .021 ass Y ha) 19 Production Table 1.7. Results of Price Sensitivity Analysis for Continuous Velvet Bean $0.99 $1.10 $1.21 $1.32 $1.43 $1.54 8,967 0.069 0.069 0.069 0.073 0.083 0.094 11,208 0.056 0.056 0.056 0.058 0.066 0.075 13,450 0.046 0.046 0.046 0.049 0.055 0.062 15,692 0.040 0.040 0.040 0.042 0.047 0.053 17,934 0.035 0.035 0.035 0.036 0.042 0.046 20,175 0.031 0.031 0.031 0.032 0.036 0.042 Biomass Yield (kg/ha) Lint Cotton Prices (Dollars/kg.) 20 DO ECONOMIES OF SIZE EXIST ON PEANUT FARMS IN THE SOUTHEAST? Introduction The theory of economies of size has been discussed for many years in the agricultural industry. According to Arne Hallam (1991), an associate professor at Iowa State University, production agriculture in the U.S. has been characterized as an industry with many firms and few barriers to entry. However, since World War II, farm numbers have decreased while size has increased. This trend may indicate that larger farms are more economically efficient in that they may have decreasing average costs and increased overall output. Given the ongoing changes in agricultural structure, there should be greater interest in competitiveness and social issues. Hallam (1991) included three areas of interest in economies of size and scale. The first issue deals with international competitiveness and changes in trade policies. In one country, economies of size may be exploited in order to maximize domestic welfare, while in another country, economies of size maybe subdued in order to protect an industry and accomplish other social goals (Hallam 1991, p. 155). Another area of interest in economies of size concerns family farms. According to Hallam (1991), ?if research shows that there are no economies of size for firms larger than ?typical? family farms, then policies that protect this entity are more palatable to those who argue for economic efficiency as a primary objective (Hallam 1991, p. 155).? 21 in e agricultural industry. Changes in technology, consumer preferences, and world conditions have forced industries to adapt. If these trends were capable of being predicted and understood, individual firms would be able to reduce uncertainty, consumers? stress about buying decisions would be eased, policy makers would be able to implement programs that are optimal for reaching desired goals, and investors would be able to make more educated decisions on the allocation of resources (Hallam 1991, p. 155-156). Peterson (1997) looked at the efficiency of different size farms in the long run. In his paper ?Are Large Farms More Efficient? he points out three measurement problems that can affect the estimates of return to scale in agriculture. The first problem deals with combining the farm dwellings with capital inputs. He states that the agricultural census asks respondent to report the value of land and all buildings, which includes the house. If the house is included in this category, the house is included as an input but its service flow does not show as farm output. Many small farms main output is from service flows, so there is an upward bias of inputs and a downward bias of output (Peterson, 1997). The next problem with estimates of returns to scale is that yields are usually substantially higher on large farms compared to small farms. According to Peterson, this should be of concern since ?a plant or an animal does not know if it is growing on a large or small farm (Peterson 1997, p.2).? If management and environmental conditions are held constant, the difference in yield can only be attributed to differences in the characteristics of the land. Large farms are usually situated on large, rectangular fields which require less travel time between fields and less turning times per acre. These The third area of interest in economies of size is related to structural changes th 22 factors lower the costs per unit output. Instead of being economies of scales, Peterson terms this situation ?economies of fertile soil and large, level fields (Peterson 1997, p.6).? The final problem is how to deal with off-farm employment. Producers on small farms are more likely to take off-farm work to help subsidize the farm income. With less time available, the level of management becomes less intensive. However, the planting and harvesting crops in a timely manner, which may lower the returns from the farm and lead to a reduction in the off-farm earnings. However, Peterson argues that these producers are not socially wasteful if their take-home pay from the farm and off-farm work is greater compared to what they would have earned from farming only. Peterson also points out that managerial ability is not the same between large and small farms. It is more difficult to manage a large farm than it is a small farm. It is assumed, therefore, that the long run average total curve begins to turn up at some level because of the limitation of managerial skill (Peterson, p.3-4). Peterson used the 90 observations in ten Corn Belt states (IL, IN, IA, MI, MN, MO, NE, OH, SD, and WI) which came from the 1987 agricultural census. He categorized the farms into nine different classes based on total sales. By basic observation he noted the difference in yields between the smaller farms and the larger rm employment, and/or managerial ability. He first ran a model using only variables for the different classes with the middle size as the reference dummy. In a second model, he included two additional variables: ?corn yield? as a proxy for the differences in managerial skill and land quality (such as field size and topography), and ?days of opportunity cost is the risk of not farms. As mentioned above, the differences may be attributed to land quality, off-fa 23 off-farm t tionship tive hip e total cost. Also, the ?corn yield? variable had a negative relationship to the s in se nt as s led work,? which would explain the differences in employment. Average total cos per dollar of output is the dependent variable in both models. In the first model, farms with less than $10,000 of sales had a positive rela to the long run average total cost per dollar of output, while larger farms had nega relationship. The second model, including the ?corn yield? and ?days of off-farm work? variables, resulted in a negative relationship to long run average total cost per unit of output for farms with less than $10,000 of sales. Larger farms had a positive relations to the averag long run average total cost per dollar of output. This implies that land quality and managerial differences influence costs per dollar of output. Peterson notes that the ?days of off-farm work? variable has a positive relationship to long run average total cost per unit of output and is highly significant, which means that part-time farmers may have higher costs due to more repairs and have the lack of timeliness associated to planting and harvesting. Peterson concluded that factors other than size can influence unit cost agriculture. The factors include the quality of land and management and the impact of off-farm employment, which influence output and output costs. When considering the factors, Peterson found that small farms and part-time operators are at least as efficie large farms, and that diseconomies of scale may exist as farm size increase (Peterson 1997, p. 12-13). Tew et al. (1980) looked at the issue that lower prices are paid for variable input if purchased in large quantities. The theory that explains this phenomenon is cal pecuniary internal economies of size. If this theory is true, large farms are not necessarily more physically efficient but they are more cost efficient because of these 24 interrel s and the hase 0, p. 153). nologies ist on many journal papers written that support the contention that, as farm sizes increase, discounts. Tew et al., however, also note that there are other reasons why input prices may vary. He mentions that prices of firms at different locations vary due to ationship between volume of sales and technological economies of size, varying transportation costs between manufacturing and retail outlets and the seasonal nature of agricultural production. Also, prices may vary because of geographic difference specialization of farm firms. It may also be possible that large farms are able to purc inputs during periods of seasonal low prices because of readily accessible financial capital and storage facilities (Tew et al. 1980, p. 151-152). In Tew et al. (1980), results show that there is evidence of pecuniary economies of size provided by some input prices, but there is little effect on total cost per acre. The differences in total costs per acre were negligible for those variable inputs, with a difference in per acre total selected variable input costs being $0.55 between 10 and 500 acres. This study supports the assumption of constant input costs (Tew et al. 198 Technology and management skills have also been identified as variables that may affect average total costs. Larger farms may have more access to new tech and better managers, which may allow these farms to allocate their resources more efficiently. However, it is difficult to quantify and collect data on these variables. Problem Statement The problem that will be looked at in this paper is if economies of size ex peanut farms located in the southeastern region of the United States. There have been 25 nuts), and determine if the costs p d ese size is nt are economies of scope (fixed-cost and variable-cost del variable and fixed costs per unit of output decrease. This paper will look at average variable, fixed, and total costs per unit of output (pounds of pea er unit of output on a peanut farm decrease as farm size increases. The dependent variables studied are variable input costs such as fertilizer, seed, pesticide, and labor an fixed input costs such as depreciation, interest, taxes (property), and insurance. The primary occupation, age of producer, and level of education may affect the result so th variables should also be included in the analysis. In the short run, the typical average total cost curve is often believed to be ?U?- shaped with increasing return to size that leads to a minimum point where return to constant, and then a section of the curve where there is decreasing return to size. This decreasing returns to size section on average total cost curve is due to increasing demand for management and increasing average fixed costs. If the theory of economies of size holds in peanut production, the average total cost curve would be much flatter. This would suggest increasing return to scale as output increases. The problem is that it is difficult to find the proper functional form to model this curve. Mafoua suggests using a flexible fixed cost quadratic model for multi-product firms. This framework allows producers to answer questions such as: ?Are three-crop farms more cost efficient than two-crop or single-crop farms?? ?How importa components) in two-crop and three-crop farms? (Mafoua 2002, p.1)? Since this paper is only looking at one enterprise (peanuts), the flexible quadratic mo may not be the proper form to model economies of size. 26 e ze average Conceptual Model Three conceptual models will be used to analyze economies of size. The dependent variables for the models will be average variable cost, average fixed cost, and average total cost, with respect to a unit of output. Since farms that use irrigation in peanut production have different costs (such as depreciation and interest) compared to dryland production, the data for the models were separated into dryland and irrigated. The conceptual models will be in the following forms 2.) AFC= F(PA,EDU,PO, AGE) Where, AFC=average fixed cost ATC=average total cost PA=peanut acres PO=primary occupation Empirical Specification of the Model The empirical specification of the quadratic model is: AC=? 0 +? 1 PA+? 2 PA 2 +? 3 EDU+? 4 PO+? 5 AGE+? The independent variables may affect the average total cost per unit of output differently. If economies of size do exist, then as peanut acres increase, average total cost should Other researchers have suggested using a log-linear or trans-log model to analyz a set of data accurately. In this paper, a quadratic model will be used to analy cost per unit of output. 1.) AVC=F(PA,EDU,PO, AGE) 3.) ATC= F(PA,EDU,PO, AGE) AVC=average variable cost EDU=level of formal education AGE=age of producer on last birthday 27 the . s compared to part-time and retired rage total cost per unit of output. they may buy rns Survey 2001? which was administered through the National Center for Peanut Competitiveness, National Peanut decrease, which would demonstrate that producers of larger farms are able to produce crop more cost efficiently. The level of education is hypothesized to have a negative affect on average total cost. As the level of education increases, average total cost should decrease since it assumed that a knowledgeable producer is better able to reduce cost relative to output It is hypothesized that farms that are operated by full-time producers would have a lower average total cost per unit of production a producers. The average fixed cost (depreciation and interest) for full-time farmers may be higher than part-time/retired farmers due to the expense of operating newer equipment. However, part-time/retired farers may have higher average variable costs because machinery on these farms may require more repairs. The age of the producer may also affect the ave Younger producers may have higher fixed costs because of higher interest and depreciation costs compared to older producers. The higher interest costs may be due to the reliance on borrowed capital. These younger producers may also be classed as high risk and may have higher interest rates due to the lack of capital assets. Young farmers may also have more depreciation compared to older producers because newer equipment while older producers may choice not to invest in new equipment as they approach retirement. Data The data are from the ?Peanut Farm Cost and Retu 28 costs (variable and fixed), varieties of peanuts, demographic haracteristics of producer and the location of farms (region and state). The data are cross-sectional collected from different peanut producers over one time period (2001). Of the 750 producers surveyed, some respondents did not complete questions that were on the survey needed to run the model. The total number of usable respondents is 152. There are 107 dryland producers and 45 producers who used irrigation in peanut production. The acres in peanut production on the farms surveyed ranged in size from 4 acres to 1300 acres. t production is 111 acres and land in irrigated peanu ost of the producers had at least lege education and others completing 52 and most worked full-time on the farm. fixed costs and total variable pounds of peanuts produced). The ariable expenses were for fertilizer, lime, other nutrients, herbicides, insecticides, , and any other inputs that were used to produce peanuts. The fixed cost as the amount of depreciation, insurance, property taxes, and interest used to produce eanuts. Producers indicated on the survey the percentage of the farm?s total fixed cost at was attributed to peanut production. Board, Southern Peanut Farmers Federation, University of Georgia, Auburn University, and University of Florida. Producers from Alabama, Florida, and Georgia were surveyed about farm size, yields, c The average size in dryland peanu t production averages 230 acres. M completed high school, with some having some col college. The average age of the producers was Average total cost was found by summing total costs. The total cost was then divided by total output ( v fungicides, seed w p th 29 01. This value is used to find the total pounds of peanuts owned . hool. For , ived the value ?1?. Produc is e Respondents of the survey identified the total pounds of peanuts that were produced on their farm in 20 produced by the individuals. Producers indicated on the survey how many acres of peanuts they planted in 2001. The total acres include land rented for free, rented land, share-rent land, and land by the producer. Excluded from this value is land rented to other producers. In the survey, producers were asked how much formal education they received The producers were given 5 possible responses: 1) Did not complete high school, 2) Completed high school, 3) Some college, 4) Completed college, 5) Graduate sc ?education? variable, producers who had at least completed some college level courses completed college, and/or pursued a graduate level degree rece ers who did not go to college received the value ?0?. Producers were asked on the survey about their primary occupation. Th question was used to characterize producers for the ?primary occupation? variable. If the producer indicated that he/she worked full-time on the farm, the value ?1? was assess to that individual. For individuals who indicated that they were retired or had a primary occupation other than farming, they received the value ?0?. In the survey, producers also indicated their age. This is used to determine if ag will affect unit cost. 30 2.1 shows the results for average variable cost per unit of output (pounds of peanuts ble s not otal cost per unit of output in dryland production. The coefficients for this estimated model are not significant and the R- squared is 0.014. The peanut acre coefficient is positive, which will increase the average Regression Results The data for the dependent and independent variables were imported into SAS and OLS estimators were calculated for each independent variable using a quadratic functional form. Tables 2.1 through table 2.3 are the results for the dryland peanut production and tables 2.4, 2.5, and 2.6 are the results for the irrigated peanuts. As mentioned in the ?Empirical Specification of the Model? section, it is hypothesized that peanut acres, education, primary occupation, and age will have negative affects on average cost per unit of output. Table produced) for dryland peanuts. None of the independent variables are significant at the 90% confidence level, but some may be economically important. The estimated peanut acre coefficient is negative which means it would decrease the average varia cost per pound of peanuts produced. The education and primary occupation variables are also negative. Age variable resulted in a positive coefficient. The R-squared is 0.035 which means the estimated model does not fit the data well. In table 2.2, the results for dryland average fixed cost per unit of output show that peanut acres has a positive affect on the average fixed cost. However, the variable i significant. The education, primary occupation, and age estimated coefficient are negative, which was hypothesized, but these are not significant either. The R-squared is 0.018. Table 2.3 is the results for average t 31 ary 6 are the results for average variable cost, average fixed cost, an nts ut icant odel is 0.11. o the 1. sing a test of first and In addition to the regression analyses already reported, t-tests were also performed test the hypothesis that large farms have lower unit costs. The hypothesis of identical er unit costs could not be rejected. total cost per unit of output. Education and age coefficients are negative while prim occupation is positive. Tables 2.4, 2.5, and 2. d average total cost, respectively, per pound of peanuts produced on irrigated peanut farms. In table 2.4, the peanut acre and primary occupation coefficients are negative, while the education and age have positive coefficients. None of the coefficie are significant at the 90% confidence level and R-squared is 0.08. Table 2.5 shows the results for average fixed cost per unit of output. The pean acre coefficient is negative, but not significant. The estimated coefficients for education, primary occupation, and age are positive, which is the opposite of the affect on average fixed cost per unit of output that was hypothesized. These coefficients are not signif at the 90% confidence either. The R-squared for this estimated m The results in table 2.6 for average total cost per unit of output are similar t results in table 2.5. Education, primary occupation, and age coefficients have a positive affect on average total cost, and none of the coefficients are significant. The coefficient for peanut acres is positive but it is not significant. The R-squared for this model is 0.1 All of the models were tested for heteroskedasticity u second moment specification in the SAS program. None of the models tested positive for heteroskedasticity. to p 32 Th ically porta if some have ge total cost per unit of output e ay be of economic importance. The affects of the coeffic data Discussion e results of the estimated quadratic model did not indicate any return to scale for average variable cost, average fixed cost, or average total cost per pound of peanuts produced. None of the coefficients in the six models are significant at the 90% confidence level. However, some of the estimated coefficients may be econom im nt. The coefficients may have an affect on average cost per unit of output variables were included, for example technology. These specification issues may over- or under-estimated the coefficients. Data that could be included in the model to explain technology were not available in the response to the survey. Technology could affect the average cost per unit of output and fit the estimated model to the data. Larger farms may implement newer technology that might increase the farms initial expense. However, these farms may have a higher return on investment, which would lower average cost per unit of output. It may be of interest that the education variable is estimated to have a negative affect on average variable cost, average fixed cost, and avera for dryland production. However, the estimated coefficient for education for irrigated production is positive in the three cost models. The results of the education coefficient cannot be deciphered because none are significant at the 90% confidenc level. Overall, the results m ients may become significant if more data was available. A variable for technology may have helped estimate a better fit for the model. Also, if time-series 33 farms. were included with the cross-sectional data, it may have helped explain changes in the cost structure across all 34 Empirical Measurement.? Review of Agricultural Economics 13,1 (1991):155- e Evidence.? Staff paper, University of Minnesota. October 1991. Mafoua, E.K. ?Economies of Scope and Scale of Multi-Product U.S. Cash Grain Farms: American Agricultural Economics Associations Meetings, July 2002. McClelland, J.W., M.E. Wetzstein, W.N. Musser. ?Returns to Scale and Size in (1986): 129-133. Miller, B.R., D. Hebie, and N. Brooks. ?The Cost Distribution of Producing Peanuts Georgia Agricultural Experiment Stations Research Bulletin Number 429 (July ers? Productivity: Some Evidence from Georgia.? American Journal Agricultural Minnesota. January 1997. Tew, B.V, S. Spurlock, W.N. Musser, and B.R. Miller. ?Some Evidence on Pecuniary ics July 1980: 151-154. References Fletcher, S. ?Peanut Farm Costs and Return Survey 2001.? Hallam, A. ?Economies of Size and Scale in Agriculture: An Interpretive Review of 172. Kislev, Y. and W. Peterson. ?Economies of Scale in Agriculture: Reexamination of th A flexible Fixed-Cost Quadratic (FFCQ) Model Analysis.? Selected paper, Agricultural Economics.? Western Journal of Agricultural Economics 11, 2 According to Acreage Size, Region, and Cost in the U.S. and Georgia.? The 1997). Nelson, M.C., N.B. Brown, Jr. and L.F. Tommer. ?Limited Resource Farm Economics December 1991: 1480-1484. Peterson, W.L. ?Are Large Farms More Efficient?.? Staff paper, University of Economies of Size for Farm Firms.? Southern Journal of Agricultural Econom 35 Dryland Farms Table 2.1. Estimated Average Variable Cost per Pound of Peanuts Produced on Variable Coefficient T-Value R-squared = 0.0355 Constant 0.1759 2.32 PA 1 -1.7563 E-4 -0.58 PA 2 2.5911 E-7 0.45 EDU 3 -0.0319 -1.04 PO 4 -0.0139 -0.34 AGE 5 8.531 E-4 0.73 Note: 1) Peanut acres; 2) Peanut acres squared; 3) Level of education (0=high school or less and 1=college or above); 4) Primary occupation (0=part-time/retired and 1=full- time); 5) age of the producer 36 stimated Average Fixed Cost per Pound of Peanuts Produced on Dryland Farms Table 2.2. E Variable Coefficient T-Value R-squared = 0.017 6 onstant 0.0961 1.43 1 DU 3 AGE 5 346 E-4 -0.87 C PA 1.9609 0.73 PA2 2 -2.7360 E-7 -0.54 E -0.0136 -0.50 PO 4 -0.0076 -0.21 -9.0 Note: 1) Peanut acres; 2) Peanut acres squared; 3) Level of education (0=high school or less and 1=college or above); 4) Primary occupation (0=part-time/retired and 1=full- time); 5) age of the pro ducer 37 Table 2.3. Estimated Average Total Cost per Pound of Peanuts Produced on Dryland Farms Variable Coefficient T-Value R-squared = 0.0135 Constant 0.272 2.57 PA 1 2.046 E-5 0.05 PA2 2 -1.4492 E-8 -0.02 EDU 3 -0.0455 -1.06 PO 4 0.0215 -0.38 AGE 5 -5.028 E-5 -0.03 Note: 1) Peanut acres; 2) Peanut acres squared; 3) Level of education (0=high school or less and 1=college or above); 4) Primary occupation (0=part-time/retired and 1=full- time); 5) age of the producer 38 Table 2.4. Estimated Average Variable Cost per Pound of Peanuts Produced on Irrigated Farms Variable Coefficient T-Value R-squared = 0.08 Constant 0.0630 0.76 PA 1 -9.89 E-6 -0.06 PA2 2 -2.5647 E-8 -0.18 EDU 3 0.0496 1.57 PO 4 -0.0013 -0.03 AGE 5 0.0011 0.91 Note: 1) Peanut acres; 2) Peanut acres squared; 3) Level of education (0=high school or less and 1=college or above); 4) Primary occupation (0=part-time/retired and 1=full- time); 5) age of the producer 39 Table 2.5. Estimated Average Fixed Cost per Pound of Peanuts Produced on Irrigated Farms Variable Coefficient T-Value R-squared = 0.11 Constant -0.0074 -0.32 PA 1 -4.766 E-5 -1.04 PA2 2 2.4960 E-8 0.65 EDU 3 0.0141 1.62 PO 4 0.0132 1.01 AGE 5 3.9560 E-4 1.15 Note: 1) Peanut acres; 2) Peanut acres squared; 3) Level of education (0=high school or less and 1=college or above); 4) Primary occupation (0=part-time/retired and 1=full- time); 5) age of the producer . 40 Table 2.6. Estimated Average Total Cost per Pound of Peanuts Produced on Irrigated Farms Variable Coefficient T-Value R-squared = 0.11 Constant 0.0556 0.62 PA 1 -5.755 E-5 -0.32 PA2 2 -6.876 E-10 -0.00 EDU 3 0.0638 1.87 PO 4 0.0119 0.23 AGE 5 0.0015 1.14 Note: 1) Peanut acres; 2) Peanut acres squared; 3) Level of education (0=high school or less and 1=college or above); 4) Primary occupation (0=part-time/retired and 1=full- time); 5) age of the producer 41 verall Conclusion sts hen f ith lower farm gate prices. Two ways cers rea s e through crop rotation and expansion of peanut th cr rotat ease peanut yields, have higher ecause of the f economies f size exist on peanut farms, producers can also increase peanut acres in order to lower In ?Economic Feasibility of an Energy Crop on a South Alabama Cotton-Peanut? ction, velvet beans are analyzed as possible rotation crop with peanuts and cotton in southeast Alabama. Velvet beans could be profitable for producers if the energy industry realize its economic benefits when used in the co-firing process of electrical generation. Producers would not shift from a cotton-cotton-peanut rotation to a velvet bean-velvet bean-peanut until a price of $41.00 per ton of dry velvet bean biomass. At a price of $45.00 per ton of dry biomass, continuous velvet bean production would be optimal for producers. The regression results in the section ?Do Economies of Size Exist on Peanut Farms in the Southeast?? are not significant as to whether economies of size exist at the O Peanut producers in southeast may have to decide on innovative ideas in order to maximize profits and minimize co w aced w that produ can ch these goal ar acres. Wi op ions, producers may be able to incr returns when producing nontraditional crops, and lower input costs b benefits of the rotation (for example lower the incidence of soil parasites). I o average cost per unit of output. se 42 imply at economies of scales do not exist on peanut farms because either some critical data or cluded in the model is technology since the data is not available from the survey used in e analysis. In conclusion, southeast Alabama peanut producers will have to look at ain in business. As stated prior, these plement a rotation that included a bioenergy crop, or, if further plies that economies of size exist, they could possible increase peanut acres. average variable, average fixed, or average total cost level. The outcome does not th variables were possibly excluded from the model. One possible variable that is not in th alternative production practices in order to rem producers could im research im