The Economic Impacts of Bioenergy Crop Production on U.S. Agriculture

Marie E. Walsh, Daniel G. de la Torre Ugarte, Hosein Shapouri, Stephen P. Slinsky

Authors are Economist at Oak Ridge National Laboratory, Research Assistant Professor at the University of Tennessee, Economist at the U.S. Department of Agriculture, and Research Associate at the University of Tennessee.

1. Introduction

In response to the oil embargoes of the 1970s, programs to develop alternative energy sources were begun in the U.S. Energy security still remains a concern, but other issues such as global climate change, air and water quality, and economic development have added new immediacy to the development of alternative energy systems. Among the alternatives is biomass energy. In addition to research on conversion technologies, the U.S. Department of Energy (DOE) established, in 1978, the Bioenergy Feedstock Development Program (BFDP) at Oak Ridge National Laboratory. The BFDP is developing new crops and cropping systems that can be used as dedicated bioenergy feedstocks. The program has screened numerous potential crop candidates to identify promising species and research frequently involves collaborative efforts with university and USDA researchers. Recent efforts are focusing on three crops--switchgrass, hybrid poplar, and willow.

Switchgrass (Panicum virgatum), is a perennial warm season grass. Its native range includes the U.S. east of the Rocky Mountains and extends into Mexico and Canada. It is a dominant species of the remnant tall grass prairies in the U.S. Switchgrass is genetically diverse and includes both lowland and upland varieties. It can be planted, managed and harvested like traditional hay crops and uses existing agricultural equipment. Poplar (Populus spp.) is widely distributed throughout the U.S. and includes aspen and cottonwood species. Poplars being developed for commercial use are crosses between two or more Populus species which provide hybrid vigor to the offspring. Hybrid poplars can be established and managed with existing agricultural equipment and can be harvested with existing forestry equipment. Willow (Salix spp.) can be produced throughout the eastern U.S. Those being developed for energy are hybrid shrubs that are produced using a close-spaced, coppice system. Planting and harvesting of willows utilize specially designed machinery which are commercially available in Europe.

The production of switchgrass, hybrid poplar, and willow utilize agricultural management practices similar to those used in the production of traditional agricultural crops and forest plantations. Production is expected to occur on agricultural croplands, with bioenergy crops competing with traditional crops. Large-scale production of bioenergy crops could have important implications for the agricultural sector in terms of crop prices and farm income. To address these issues, the DOE Office of Transportation Technologies and USDA, in collaboration with the University of Tennessee Agricultural Policy Analysis Center (UT-APAC), have jointly undertaken a study to evaluate the potential economic feasibility and ramifications of bioenergy crop production in the U.S. using an agricultural sector model (POLYSYS) that has been modified to include switchgrass, hybrid poplar, and willow. The analysis seeks to: (1) estimate the farmgate price needed to make bioenergy crops economically competitive with alternative agricultural uses for cropland, (2) determine the regional distribution of bioenergy crop production, (3) estimate the potential impact of bioenergy crop production on traditional crop prices and quantities, (4) estimate the potential impact of bioenergy crop production on net farm income, and (5) evaluate the economic potential of a modified Conservation Reserve Program to serve as a source of bioenergy crops.

2. Methodology

2.1. The POLYSYS Model

POLYSYS, is an agricultural policy simulation model of the U.S. agricultural sector that includes national demand (simultaneous block), regional supply (linear programming), livestock, and aggregate income modules [1]. The model includes cropland planted to the eight major crops (corn, grain sorghum, oats, barley, wheat, soybeans, cotton, and rice), alfalfa and other hay, edible oils and meals sectors, and a livestock sector. Crop supplies are allocated among 305 Agricultural Statistical Districts (ASDs), while corresponding crop prices, exports, domestic demand, and stocks are calculated at the national level. Livestock production and net farm income related variables are estimated at the national level. POLYSYS is anchored to a baseline projection for the major crops and estimated results are calculated as deviations from the baseline. POLYSYS provides estimates of changes in supply, demand, price and farm income variables over a ten-year time period.

2.2 Cropland Categories

In the contiguous 48 states in the U.S., 174.6 million hectares (431.4 million acres) are identified as cropland [2] with 103 million hectares (254.5 million acres) planted to the eight major crops [3], 11 million hectares (27.2 million acres) planted to alfalfa, and 13.4 million hectares (33.2 million acres) planted to other hay crops [4]. Another 7.7 million hectares (19 million acres) are idled, 24.4 million hectares (60.3 million acres) are planted to pasture, and 12.1 million hectares (29.8 million acres) are enrolled in the Conservation Reserve Program (CRP) [5]. POLYSYS includes all of these hectares except the 2% used for other uses such as for fruits, vegetables, and other minor crops. For this analysis, bioenergy crop production is limited to geographic regions where they can be produced under rainfed conditions and where sufficient research has been conducted to provide yield and management data for which experts have reasonable confidence. Bioenergy crops can be grown in other regions of the U.S. than used in this analysis, but data regarding appropriate varieities, management practices, and expected yields are lacking. These restrictions result in 149 million hectares (368 million acres) of cropland suitable for the production of at least one of the bioenergy crops.

The analysis includes the potential to produce bioenergy crops on Conservation Reserve Program lands. The CRP was begun in the 1985 Farm Bill and sets asides environmentally sensitive hectares under 10-15 year contracts. Hectares are planted to conservation crops such as perennial grasses and trees and farmers receive an annual rental payment for the land. Harvest is prohibited except under emergency conditions. Modifications to the CRP program have been suggested as a means to introduce bioenergy crops to agriculture and to reduce their price to end users. For this analysis, geographic limitations reduce the number of hectares from the enrolled 12.1 million hectares to 9.76 million hectares (29.8 million to 24.1 million acres). These hectares are further restricted by removal of lands that are most environmentally sensitive including (1) hectares enrolled in buffer strips to protect water quality, (2) hectares classified as wetlands, (3) hectares critical to watershed management, and (4) critical habitat hectares in Wildlife Conservation Priority Areas. These restrictions remove an additional 2.9 million hectares (7.2 million acres). Thus for this analysis, 6.8 million hectares (16.9 million acres) of CRP lands are identified as being potentially suitable for bioenergy crop production.

2.3 Crop Production Costs

Enterprise budgets for traditional crops and bioenergy crops are developed using the APAC Budgeting System (ABS) [6]. ABS generates consistent enterprise and rotation budgets for each of the 305 ASDs contained in POLYSYS. ABS generates cost of production based on operation schedules (i.e., field-level activities) using internal cost and technical databases derived from easily updated sources. Costs include variable input costs, fixed machinery costs, and labor. The method used is consistent with those used by USDA and recommended by the American Association of Agricultural Economics. To ensure consistency with the estimated costs of producing traditional crops, the ABS is used to estimated all costs associated with producing bioenergy crops with the exception of the harvesting costs of hybrid poplar and willow. These costs are estimated using BIOCOST, a budget generator model developed by Oak Ridge National Laboratory to estimate the cost of producing bioenergy crops [7]. BIOCOST uses methods similar to ABS.

Because bioenergy crops are not currently produced commercially on a large-scale, yields and management practices were based on research data and expert opinion. Experts from the U.S. Departments of Energy and Agriculture participated in a workshop where recommendations were made regarding yields, suitable geographic regions, and management practices. Yields of bioenergy crops on idled and pasture hectares are assumed to be 85 percent of those that can be obtained on hectares in crop production.

Bioenergy crop management practices used on CRP hectares were determined in subsequent discussions with USDA. Two management practices were decided on for use in the analysisone to achieve high levels of biomass production, and one to achieve high levels of wildlife diversity. The wildlife diversity management strategy utilizes fewer fertilizer and chemical inputs than does the production management strategy and places significantly greater constraints on the harvest of switchgrass (i.e., restricting harvest to alternating halves of a field each year to provide variations in physical structure as compared to annual harvest of the whole field in the production management strategy). Yields on CRP hectares are adjusted by an index of traditional crop yields obtained on CRP hectares prior to being enrolled in the CRP program.

2.4 Production Decision Issues

Due to the multi-year characteristics of bioenergy crops, a net present value approach is used to decide which crops are produced. To evaluate whether a bioenergy crop should be planted in a particular region, its net present value profit is calculated and compared to the corresponding net present value profits of each of the traditional crops. Given that the three bioenergy crops examined in this study have production cycles of different time lengths, a common planning horizon is used to insure comparability of discounted revenue streams. A planning horizon of 40 years is set because, although not a minimum common denominator, it spans a long enough time period to consider insignificant, the stream of net returns beyond this period. A real discount rate of 6.5 percent is used.

Production decisions on CRP, idle, and pasture hectares require additional considerations to those made on cropland planted to traditional crops. On CRP hectares, it is assumed that existing contracts, upon expiration, can either be renewed under the same conditions that were in effect upon initial enrollment in the program, or the hectares can be planted to bioenergy crops under a modified contract. This assumption greatly simplifies the analysis by not requiring development of a model to determine re-enrollment decisions for hectares currently under contract, or to determine hectares that may be newly enrolled during CRP sign-ups that occur during the period of analysis. In exchange for being allowed to produce and harvest bioenergy crops on CRP hectares, 25 percent of the current rental rate is forfeited. Thus, the decision to allocate CRP hectares to bioenergy crop production is reduced to a comparison of the NPV of producing bioenergy crops with the NPV of the foregone CRP rental payments.

Cropland hectares that are idle or in pasture are assumed to be so for economic reasonsthat is, given prices, costs of production, and yields, the most economic use of the land is either to not plant a crop or to dedicate it to pasture. In order to return these hectares to production of the major crops, or to plant bioenergy crops, the net present value returns of these crops must be higher than the most profitable crop under the baseline assumptions. Furthermore to account for possible inertia to keep the land in its current use and/or the value of pasture land in livestock operations, a premium of 10 percent above the baseline net present returns for idled hectares and 15 percent for pasture hectares is required.

2.5 Land Allocation Rules

To avoid corner solutions in the regional linear programming models, POLYSYS contains embedded flexibility constraints that limit the hectares that a given crop can lose or gain each year. To accommodate the addition of bioenergy crops, these allocation rules were modified. The extent to which hectares can be increased or decreased relative to the baseline is a function of whether the NPV returns for traditional crops is positive, negative, or a mixture for three years. Larger hectare changes are permitted for crops that have negative NPV returns than for those with positive NPV returns. Additionally, the amount of idle and pasture hectares that can be reallocated is limited as are hectares for crops that comprise more than 20% of all planted hectares in an ASD. Once hectares are allocated to a bioenergy crop, the hectares remain allocated to the bioenergy crop for the duration of its productive life cycle due to the assumption of some type of contractual arrangement between farmers and bioenergy crop users which limits, for the duration of the bioenergy crop production cycle, the conversion of hectares back to traditional crop production.

2.6 Price Expectations

To improve the estimation of expected prices and net present value returns for each crop, a rational expectation hypothesis is implemented. Rational expectations are premised on the hypothesis that in a given year, farmers will anticipate price changes resulting from significant shifts of hectares from traditional crops to bioenergy crops and will incorporate these price changes into their planting decisions. These changes in expected prices are estimated in POLYSYS through an iterative process. POLYSYS first solves the supply module using a net present value return. After the cropland hectares are allocated and the effective supply computed, equilibrium demand quantities and market prices are estimated through interaction of the supply and demand modules. POLYSYS next utilizes these estimated market prices as the new expected prices to calculate the NPV returns of all crops. The POLYSYS supply module solves for the second time in the same time period, and the new land allocation and effective supply estimates become the simulated values for that time period. Finally, these values interact with the demand module to solve for market-clearing demands and output prices for each crop. This two-stage procedure is repeated until the changes in prices between two iterations for the traditional crops is less than a predetermined convergence factor.

3. RESULTS

Multiple scenarios have been run, but to facilitate the discussion of the economic feasibility and potential economic impacts of bioenergy crop production on the agricultural sector, the results of just two scenarios are presented. These two scenarios are chosen as they serve to place boundaries around the discussion.

The first scenario assumes a bioenergy farmgate price of $US 33/dry Mg, 34.91/dry Mg, and 36.19/dry Mg ($US 30/dry ton, $31.74/dry ton, and $32.90/dry ton) for switchgrass, willow, and hybrid poplar respectively. Because the energy density of the three bioenergy crops differs slightly, an equivalent energy price in $/GJ ($/Mbtu) results in slightly different $/dry Mg ($/dt) prices. This scenario also assumes that the wildlife management practices are employed on CRP hectares and that farmers receive 75 percent of their rental rate in exchange for the right to harvest and sell bioenergy crops.

The second scenario assumes a bioenergy farmgate price of $US 44/dry Mg, 46.55 dry Mg, and 48.26/dry Mg ($40/dt, $42.32/dt, and $43.87/dt) for switchgrass, willow, and hybrid poplar respectively. This scenario assumes that the production management practices are employed on CRP hectares and that farmers receive 75 percent of their rental rate in exchange for the right to harvest and sell bioenergy crops.

Both scenarios use the 1999 USDA baseline for the eight major crops in the analysis and the 1999 Food and Agricultural Policy Research Institute baseline for alfalfa and other hay. CRP baseline hectares and location are those prevailing on October 1, 1998. Idle and pasture baseline hectares and location are obtained from the 1997 Census of Agriculture and are assumed to be representative of, and consistent with, the baseline information for the major crops. Bioenergy crop production is assumed to begin in the year 2000, and the largest shift in land is assumed to occur in that year. Annual impacts for the period 2000-2008 are available for each scenario, but to facilitate the presentation of the results, only the results from the year 2008 are presented. Results from 2008 are presented because it is assumed that most of the initial shock resulting from the allocation of land to bioenergy crop production will have occurred by this time, and that the agricultural production sector will have settled into a new long-run equilibria.

3.1 Land Use Impacts

Estimated total national hectares allocated to bioenergy crops under the two scenarios are presented in Table 1. Under the first scenario, 7 million hectares (17.4 million acres) of cropland are planted to bioenergy crops with 4.2 million hectares (10.4 million acres) coming from hectares that are planted to traditional crops. An estimated 2.5 million CRP hectares (6.2 million acres), 0.09 million idled hectares (0.2 million acres), and 0.2 million pasture hectares (0.5 million acres) are converted to bioenergy crop production. For scenario two, an estimated 17 million hectares (41.9 million acres) are planted to bioenergy crops with 9.5 million of those hectares (23.4 million acres) coming from hectares planted to traditional crops and 5.2 million hectares (12.9 million acres) from converted CRP hectares. An estimated 0.85 million idled hectares (2.1 million acres) and 1.4 million pasture hectares (3.5 million acres) are also converted to bioenergy crop production.

Table 2 presents the shift in hectares from the major crops that result from the introduction of bioenergy crops. Total cropland hectares planted increase from the baseline 131.7 million (325.4 million acres) to 135.2 million (333.9 million acres) under scenario 1 and to 140.4 million hectares (346.8 million acres) under scenario 2. Increased hectares come predominantly from CRP hectares in scenario 1 and from CRP, idled, and pasture hectares in scenario 2. Many traditional crops lose hectares as a result of bioenergy crop production, but gain hectares from production on idled and pasture hectares resulting in a smaller net impact. This effect occurs because the shift in hectares resulting from the introduction of bioenergy crops results in lower production of traditional crops and thus higher traditional crop prices. The higher prices provide sufficient incentive to return some idled and pasture hectares to traditional crop production.

Bioenergy crops compete for hectares not only with traditional crops, but with each other as well, with hectares allocated to the most profitable crop given assumed yields, production costs, and the land allocation rules described previously. Given these assumptions, switchgrass is relatively more profitable than poplars and willows in nearly all ASDs and poplars are relatively more profitable than willows. As a result, switchgrass dominates the other two bioenergy crops and nearly all (i.e. 99%) of the acres in traditional crops, idled, or in pasture that are shifted to bioenergy crop production are shifted to switchgrass production under both price scenarios. The change in management practices on CRP hectares to the wildlife scenario make poplars relatively more profitable than switchgrass in some ASDs. Thus under the wildlife CRP scenario (scenario 1), hectares planted to bioenergy crops are split between poplars and switchgrass. Under the CRP production scenario (scenario 2), CRP hectares are allocated to switchgrass. Willows, being the least profitable of the three bioenergy crops, is out-competed by the other two bioenergy crops in this analysis.

Total national production in the year 2008 under scenario 1 is an estimated 60.9 million dry Mg (67 million dry tons) annually. While most of the poplar is not harvested until after 2008, its contribution can be annualize to provide an equivalent annual amount of 10.9 million dry Mg (12 million dry tons) for a total of 71.6 million dry Mg (78.8 million dry tons) of feedstock. Under scenario 2, an estimated 170.9 million dry Mg (188 million dry tons) could be produced annually. This production would be generated exclusively from switchgrass, as the results indicate that in this scenario no other bioenergy crops would have significant hectares. Figure 1 depicts the regional distribution of bioenergy crop production (in acres) for the second scenario.

Figure 1. Bioenergy Crop Plantings on All Acres, Scenario 2, Year 2008

3.2 Price Impacts

The shift of cropland from traditional crops to bioenergy crops results in higher prices for traditional crops. The impact on traditional crop prices is a function of the hectares shifted to bioenergy crops, as well as the elasticity of supply and demand for each crop. Price impacts are presented in Table 3. Traditional crop prices increase by an estimated 4 to 9 percent under scenario 1 depending on crop, and by an estimated 9 to 14 percent under scenario 2. It should be noted that the higher estimated prices for the major crops, except wheat, are within the range of historical market prices experienced by these crops over the last five years. Because POLYSYS calculates the price as changes from the baseline, the level of final prices is highly influenced by the price level assumed in the baseline. It should also be noted that POLYSYS is not able to estimate price changes for alfalfa and other hay crops which are generally determined locally rather than at a national level.

3.3 Net Farm Income and Net Returns Impacts

The overall impact of bioenergy crop production on agriculture is summarized by changes in farm income. To account for the contribution of bioenergy crops to net farm income and net farm returns, the net present value resulting from the production of biomass is expressed in terms of an annuity and added to the farm income measures used. Due to the inability to determine price impacts for alfalfa and other hay, unaccounted losses in the livestock sector may occur. Under scenario 1, a gain of $US 2.6 billion is projected with bioenergy crops accounting for $US 500 million of that total and the rest resulting from higher traditional crop prices. Under scenario 2, a gain of $US 6.0 billion is projected with bioenergy crops accounting for $US 2.3 billion of that total and the rest resulting from higher traditional crop prices.

3.4 Energy Supply Implications

Under scenario 1, bioenergy crops could supply approximately 1.3 EJ (1.23 Quads) of primary energy. Given current conversion efficiencies, it could be used to produce 26.5 billion liters (7 billion gallons) of ethanol and displace 106 million barrels of oil if used as a transportation fuel in place of gasoline. Current annual ethanol production in the U.S is about 5.3 billion liters (1.4 billion gallons). Alternatively, that quantity of biomass could be used to produce 130.5 billion kilowatt hours of electricity (assuming a gasifier combined cycle technology) which is equivalent to about 2.65 percent of the electricity currently produced in the U.S. Under scenario 2, bioenergy crop production could supply approximately 3.07 EJ (2.91 Quads) of primary energy. Given current conversion efficiencies, it could be used to produce 63.2 billion liters (16.7 billion gallons) of ethanol and displace 253 million barrels of oil. Alternatively, that quantity of biomass could be used to produce 307 billion kilowatt hours of electricity which is equivalent to about 7.3 percent of the electricity currently produced in the U.S.

4. Summary

The analysis indicates that at farmgate price of less than $44/dry Mg (switchgrass), nearly 17 million hectares (41.9 million acres) of agricultural cropland in the U.S. could produce bioenergy crops at a profit greater than existing agricultural uses. Additionally, farm income could increase by nearly $US 6 billion as a result of bioenergy crop production. Traditional crop prices are estimated to increase between 9 and 14 percent over baseline price projections with estimated prices being within the market price range over the last five years. Total annual biomass production is estimated to be 171 million dry Mg (188 dry tons), equivalent to 3.07 EJ (2.91 Quads) of primary energy. This quantity of biomass could displace an estimated 253 million barrels of oil or supply an estimated 7.3 percent of the U.S. electricity needs.

5. References

1. De La Torre Ugarte, Daniel G. and Daryll E. Ray. Biomass and Bioenergy Applications of the POLYSYS Modeling Framework. Biomass and Bioenergy. Vol. 18(4):291-308. March 2000.
2. U.S. Department of Agriculture, National Agriculture Statistical Service, Census of Agriculture1997, July 1999.
3. U.S. Department of Agriculture. Agricultural Baseline Projections to 2008. World Agricultural Outlook Board, Office of the Chief Economist, U.S. Department of Agriculture. Prepared by the Interagency Agricultural Projections Committee. Staff Report No. WAOB-99-1, February 1999.
4. Food and Agricultural Policy Research Institute. The Rainbow Book: A Summary of the November 1998 FAPRI Baseline. University of Missouri, January 1999.
5. U.S. Department of Agriculture, Farm Service Agency. The Conservation Reserve Program, 18^th Signup. March 1999.
6. Slinsky, Stephen P. and Kelly H. Tiller. Application of an Alternative Methodological Approach for Budget Generators for Research. Selected paper, published abstract. Journal of Agricultural and Applied Economics. 31(2). August 1999 (Forthcoming).
7. Walsh, Marie.E. and D.A. Becker. BIOCOST: A Software Program to Estimate the Cost of Producing Bioenergy Crops. Proceedings of Bioenergy 96, Nashville, Tennessee, Sept. 15-19th, 1996. Pages 480 to 486.