...

PDF

by user

on
Category: Documents
2

views

Report

Comments

Description

Transcript

PDF
Have Hog Producers with Production Contracts Maintained an Economic Advantage of
Independent Hog Producers in Recent Years?
Nehring, Richard F.*, David Banker*, and Erik O’ Donoghue*
May 15, 2003
Abstract
Preliminary estimates of technical efficiency based on USDA data for 1997 through 2001 indicate that
independent operations were significantly more efficient than contract operations. Preliminary estimates
also indicate that both types of operations exhibited increasing returns to scale with contract operations
appearing to exhibit significantly higher returns to scale than independent operations, but that larger
contract and independent operations exhibit roughly comparable returns to scale. Our estimates of
excess nutrients that derive from both commercial fertilizer and manure, comparing the performance of
production contract operations and independent operations indicate that, in general, levels of excess
nutrients per acre of land are significantly higher on contract operations than independent operations.
The results suggest that adjusting the performance measures to include excess nutrients as a “bad
output” would tend to favor independent producers over contract operations compared to performance
measures that ignore pollution.
*Economists, Resource Economics Division, Economic Research Service, U.S.
Contact author: Richard Nehring, 104 Crossing Pointe Ct., Frederick, Md 21702.
Paper Presented at the Annual Meetings of the AAEA, July 2003, Montreal, Quebec.
Disclaimer: The views expressed here are those of the authors and may not be attributed to ERS or USDA.
1
Introduction
The growing importance of production contracts in hog production suggests large economic
benefits and/or reduced risk accrue to farmers participating in production contracts (see Figure 1).
The rapid growth in contracting has led to heightened concerns about the impacts of increasing
concentration on farm structure and the health of the rural economy. It has also led to efforts by
various levels of government to regulate contract production. Recent research suggests that
production contracts in hog production are associated with a substantial increase in productivity
representing a technological improvement over independent production as described in Key and
McBride (2002) for hog production in 1998. This implies that efforts to regulate contracting
operations may have large economic costs. However, recent consolidation trends in the hog sector
may have altered the relative economic performance of independent versus contract operators.
Preliminary estimates of technical efficiency based on USDA data for 1998 through 2001
indicate that while independent operators were much less technically efficient than contract
operations in 1998 and 1999, they were more efficient or nearly as efficient as contract operations
in 2000 and 2001. Using a Cobb-Douglas specification we estimate that the mean technical
efficiency score1 for contract operations in 1998 is 0.83 compared to 0.67 for independent
operations. Similarly, in 1999 we find that contract operations achieved mean technical efficiency
scores of 0.85 compared to 0.71 for independent operations. But we also find a significant
narrowing in the advantage of contract farmers over independent operations in Iowa, the major hog
producer. In contrast to the 1998 and 1999 results, we find that independent hog operations
outperformed contract operations with technical efficiency scores of 0.68 compared to 0.63 in
2000, while in 2001, we again find that contractors are more efficient with a score of 0.72
1 Technical efficiency represents the ratio of current to maximum possible or “best practice” production. Technical efficiency scores
2
compared to 0.64, but not by nearly as wide a margin as in 1998 and 1999. The technical
efficiency comparisons of independent operations with contract operations suggest a possible
narrowing of the competitive advantage of contract farms over independent farms between 1998
and 2001. This may be because of the exit from agricultural production of many independent
operations, who were likely much less efficient than the survivors. For example, between 1997
and 2001 the number of operations raising hogs in Iowa dropped from 18,000 to 10,500, while the
proportion of hog numbers on operations with fewer than 1000 hogs dropped from 37 percent to 15
percent (USDA 2002, 1998). Since independents represent the majority of operations reporting it is
clear that the scale of production on surviving independents increased sharply between 1997 and
2001.
These results suggest that the Key and McBride study may overemphasize the role of technological
improvements of contracted production over independent production in the rapidly changing hog sector.
Several major questions then arise. How important is the role of risk reduction in contracting hog
production ? How important is the role of scale in contracting hog production? Has the pace of contract
production abated in regions where remaining independent operations are technically and scale efficient?
Finally, does regulating contract production in this changed economic environment involve large
economic costs, particularly if the impacts of pollution are included—i.e does contracted production and
independent production involve the same pollution risks ?
In this study, we do two things: 1) develop farm level estimates of excess nutrients that derive from
both commercial fertilizer and manure, comparing the performance of production contract operations and
independent operations, and 2) calculate farm-level efficiency scores (performance measures of economic
activity) and assess economies of scale for farms involved in production contracts and independent farms
using a multi-output, multi-input model. We construct a panel data set of farms for 1997 and 2001 based
on pseudo cohorts and derive measures of efficiency and returns to scale. The analysis uses five years
(1997-2001) of USDA’s Agricultural Resource Management Survey (ARMS) data, incorporating both
may range from zero to one.
3
1)whole farm data, including income and operator characteristics, and 2) hog production practices and
cost data. Finally, we infer the relative risk of water pollution based on these findings, recognizing that
pollution risk may vary by climate and soil type. This study focuses on farms producing hogs for
slaughter from independent and contracting operations. And, we focus only on production contracts, and
ignore marketing contracts.
Model
We use a multi-input distance function and stochastic frontier and inefficiency procedures to estimate
efficiency scores, assess factors influencing efficiency, and estimate returns to scale. The input distance
function permits a multi-input, multi-output technology without requiring observations on output and
input prices as described by Coelli and Perelman (1996, 2000). In contrast to a cost or profit function, the
input distance function does not require a system of equations in the estimation procedure. The input
distance vector considers how much the inputs may be proportionally contracted with outputs held fixed.
In this sense it implies cost minimization. The appropriate functional form is ideally flexible, easy to
calculate, and permits the imposition of homogeneity. Following Coelli and Perelman, we use stochastic
production frontier (SPF) measurement to econometrically estimate the input distance function
DI(X,Y,R), after implementing theoretically required regularity conditions, making a functional form
assumption, and specifying a stochastic structure allowing for both a “white noise” error and a one-sided
error representing deviations from the production frontier. Writing the distance function accordingly,
assuming it can be approximated by a translog functional form to limit a priori restrictions on the
relationships among arguments of the function, we obtain:
(1a)
ln DIit/X1,it = α0 + Σ k αk ln X*kit +0.5 Σ k Σ l βkl ln X*kit ln X*lit
+ Σ m αm ln Ymit + 0.5 Σ m Σ n αmn ln Ymit ln Ynit + Σ k Σ m ϕkm ln Ykit ln X*mit , or
(1b) -ln X1,it = α0 + Σ k αk ln X*kit + 0.5Σ k Σ l βkl ln X*kit ln X*lit
+Σ mαm ln Y mit + 0.5Σ m Σ n αmn ln Y mit ln Ynit + Σ k Σ m ϕ km ln Ykit ln X*mit - ln DIit ,
4
where i denotes the ith farm, m,n the outputs, k,l the inputs, and t, time period. More precisely, X*lit
represents the lth input divided by land so that the specification is essentially specified on a per land basis,
which seems reasonable as we often interpret farm production and productivity per unit of land. This
functional relationship, which embodies a full set of interactions among the X and Y arguments of the
distance function, can be more compactly written as -ln X 1,it = TL(X/X1,Y,t) = TL(X*,Y,t). We append a
symmetric error term, v to equation (1b) to account for noise, and also change the notation “- ln DIit” to
“u”. The resulting -ln X1,it = TL(X*,Y) + v - u function (with the sub-scripts suppressed for notational
simplicity) may be estimated by maximum likelihood (ML) methods, to impute the TE measures as the
distance from the frontier. In addition to land the Xit represent expenditures on five other inputs: labor,
fuel, fertilizer and other chemicals, miscellaneous operating expenses, and capital services. Our outputs
are revenue from corn, soybeans, other crops, and livestock ( i.e. hogs and other livestock).
It is assumed that the inefficiency effects are independently distributed and Ui arises by truncation (at
zero) of the normal distribution with mean µ, and variance s 2 , where µI is defined by
(2) µi= δ0 + δ 1 pmark + δ2 acres + δ 3 age + δ4 education + δ 5 rent + δ 6 debt + δ7 biocorn
+ δ8 biosoybeans + δ 9 off-farm + δ 10 excessn + δ 11 excessp
+ δ12 cohortsmall + δ13 cohortlarge
where pmark represents the proportion of operations that sell slaughter hogs under a production contract,
acres is a continuous variable representing acres per farm, age represents the age of the operator, education
represents the education score for the operator (where 1=less than high school, 2=high school diploma,
3=some college, 4=BA or BS degree, and 5=graduate school), rent represents the ratio of acres rented to
total acres operated, debt represents the debt/asset ratio, biocorn represents the proportion of corn acres in
GMO corn, biosoy represents the proportion of soybean acres in GMO soybeans, off-farm represents the
ratio of off-farm earnings to farm earnings, excessn represents the amount of excess nitrogen per acre
operated after accounting for all nitrogen credits and uptake of nitrogen by crops, excessp represents the
amount of excess phosphorous per acre operated after accounting for all phosphorous credits and after
5
uptake of phosphorous by crops, cohortsmall represents the dummy for small and medium commercial and
residential farms, and cohortlarge represents a dummy for very large family farms and nonfamily farms.
All continuous variables (that is, all of the inefficiency effects except for the cohort dummies and pmark)
are in logs.
The maximum-likelihood estimates for the parameters of the stochastic frontier model defined by
equations (1b) and (2) were estimated using FRONTIER Version 4.1 (Coelli). . For the SPF model -u thus
represents inefficiency; the efficiency scores generated by FRONTIER essentially measure exp(-U) =
DI(X*,Y,R). This is therefore our measure of technical efficiency.
The expected signs on the coefficients for acres, rent, bicorn, and biosoy, are negative, signifying that
these variables are likely to be negatively related to inefficiency and positively related to efficiency.
Similarly, the expected signs on the coefficients for age, debt, and off-farm are likely to be negative. The
expected coefficients for education, excessn, excessp are ambiguous. The coefficients on excessn and
excessp are ambiguous because it is unclear whether larger operations with relatively more livestock, and
hence more excessn and excessp are likely to be more technically efficient on average than large grain
farms with relatively little excessn and excessp. The coefficient for off-farm is likely to be positive
because we have not included off-farm income as an output. Nor have we included off-farm hours worked
as part of the wage bill. Thus, in our model, off-farm is likely to be positively related to inefficiency and
hence, negatively related to efficiency, because time spent in off-farm employment negatively influences
the quality and availability of on-farm employment. The expected sign on pmark is ambiguous given the
evidence for 1998-2001 on the comparison of technical efficiency scores for independent and contract
operations cited above. Further our analysis incorporates data for 1997, for which we have not calculated a
comparison of the technical efficiency scores of independent operations compared to contracting
operations.
The SPF-based scale economy measure may also be computed from the estimated model via
derivatives or scale elasticities: -,DIY = -, m∂ln DI(X,Y,t)/∂ln Ym = ,X1Y for M outputs Ym . This measure
is based on evaluation of (scale) expansion from a given input composition base.
6
Nutrient Balance Use
We develop farm-level estimates of excess nitrogen (phosphorus) from commercial fertilizer and
manure sources for hog producing states, including Southern, Eastern, and Western states as well as those
in the Corn Belt. At the national level we see that between 1994/95 and 2001 the share of the value of
production under contract for hogs doubled to 60 percent as shown in figure 1. In Figure 2 we see that the
majority of specialized hog Agricultural Statistics Districts in the United States are located in the Corn
Belt, North Carolina, and Oklahoma.
In addition to hogs, cattle, dairy, and poultry are major sources of manure in these states. Using hogs
as an example, in corn producing states we see that hog output per farm, measured as value of production
adjusted for inflation, increased dramatically between 1995 and 2000 (USDA Costs and Returns data). In
the states intensively surveyed ( Illinois, Indiana, Iowa, and Minnesota, each with 50 or more observations
in each time period) hog output per farm increased dramatically—276 percent in Illinois, 202 percent in
Iowa, and 185 percent in Minnesota. Only Indiana showed no appreciable growth in hog output per farm.
In the less intensively surveyed states, (Michigan, Nebraska, Ohio, South Dakota, and Wisconsin) the data
also suggest large increases in output per farm. In the thinly surveyed states (Kansas and Missouri) there
was little increase in output per farm. Changes in concentration in other species were mixed during 19962001. USDA data indicate close to a 200 percent increase in cattle output per farm in Kansas and South
Dakota but only small increases in dairy output per farm in the key dairy states of Michigan, Minnesota,
and Wisconsin. Poultry output per farm increased nearly 200 percent for the major corn producing states,
but concentrations by state cannot be identified from the available USDA data.
Excess nitrogen (phosphorus) is defined as the difference between the amount of nitrogen (phosphorus)
applied from all sources (chemical fertilizers plus soybean, legume and/or manure credits) and the amount
of nitrogen (phosphorus) removed during the crop production process. To calculate excess nitrogen and
phosphorous at the farm level, we employ well-known nutrient balancing techniques.
7
Data
Our approach uses U.S. farm level data from the 1997-2001 Agricultural Resources Management Study
(ARMS) surveys. ARMS is an annual survey covering farms in the 48 contiguous states, conducted by the
National Agricultural Statistics Service and the Economic Research Service. All hog-producing states
represented in ARMs phase III surveys were selected. In order to allow inferences to the state and
regional level we use weighted observations. In general, observations in each of the years analyzed
included more than twenty hog- producing states. IL, IN, IA, MN, NC, NE, and OH were considered as
individual states. Observations in Michigan and Wisconsin were considered as one eastern Lake state.
Observations in North Dakota and South Dakota were considered as one upper Northern Plains state.
Observations in DE, ME, MD, NH, NJ, NY, PA, RI, VA, and WV were considered as one eastern state.
Observations in AL, AR, FL, GA, KY, LA, MO, SC, TN, and TX were considered as one southern state,
and observations in CO, ID, KS, OK, OR, WA, and WY were considered as one western state. Hence, we
used 12 “states” or regions in the anlaysis.
Four outputs are included in the model estimation. The crop outputs consist of corn, soybeans, and
other crops, measured as the total value of production of each. Livestock production is measured as the
total value of livestock production. For the variable inputs, labor costs are the annual per-farm
expenditures on labor; energy is expenditures on gasoline, diesel fuel and other fuels; fertilizer is
expenditures on fertilizer, lime and other chemicals; and materials is expenditures on seed, feed and
miscellaneous operating expenses. Capital machinery is measured as the annualized flow of capital
services from assets (excluding land). Our land variable is an annualized flow of services from land and is
constructed as an annuity based on a 20-year life and 10 percent rate of interest.
To support empirical production studies using panel data, the temporal pattern of a given farm’s
production behavior must be established. In the absence of genuine panel data, repeated cross-sections of
data across farm typologies may be used to construct a pseudo panel data (see Deaton, Heshmati and
Kumbhakar, Verbeek and Nijman) The pseudo panels are created by grouping the individual observations
into a number of homogeneous cohorts, demarcated on the basis of their common observable time8
invariant characteristics, such as quality of land as determined by geographic location and size of farm as
determined by the gross value of sales. The subsequent economic analysis then uses the cohort means
rather than the individual farm-level observations.
Farm-level data were assigned to cohorts by typology, (and sub typology), by gross value of sales,
by state, and by year for the hog-producing states, generally following ERS farm typology groups (as they
are divided by gross value of sales) described in Table 1. Cohort 1 is represented by hog farms with gross
value of sales of less than $100,000. Cohort 2 is represented by hog farms with gross value of sales of
$100,000 to $249,999. The largest cohort, cohort 7, represents hog farms with gross value of sales of
greater than $1,000,000. Altogether, we form seven cohorts, which are delineated by gross value of sales
as shown in Table 2. The resulting panel data set consists of 7 cohorts for each of 12 states, for 19972001, measured as the weighted mean values of the variables to be analyzed. In total we have 420 annual
(cohort) observations (84 per year, a balanced panel), summarizing the activities of 517 farms in 1997,
1954 in 1998, 530 in 1999, 342 in 2000 and 326 in 2001. To translate these nominal values into real terms
for the panel data, all variables are deflated by the estimated increase or decrease in cost of production in
1998, 1999, 2000, and 2001 compared to 1997 (in terms of agricultural prices).
A summary of the sample data used in the output distance function estimations is presented in Table
3 for 2000. The average farm size varies from 95 acres in the limited resource typology to 8,796 acres on
the very large family farm typology. Excess nitrogen at close to 40 pounds per acre operated and excess
phosphorous at close to 30 pounds per acre operated are highest nonfamily farms. The average age of
farmers is highest in retirement and low sales typologies, and lower in the residential and higher sales farm
typologies. The farmer education average of 2.45 is between a high school diploma (2) and some college
(3), and tends to be slightly greater in the high sales typologies.
Input distance function results
The maximum-likelihood (ML) estimates of the parameters of the output distance stochastic
production frontier are presented in Table 4. Given the pseudo-cohort nature of the data, cohort dummies
are added to take account of cohort-specific effects (Heshmati and Kumbhakar). Close to 80 percent of
9
the coefficients of the model are significant at the 10 percent level or better. The estimate of the variance
parameter, ? (where ? =FU 2/( FV2 + F U2), is also significantly different from zero, which implies that the
inefficiency effects are significant in determining the level and variability of output of farmers in the corn
states analyzed.
Turning to the factors influencing efficiency, we find that the coefficient on pmark is positive and
significant, indicating that the variable representing the proportion of contract production is positively
associated with technical inefficiency and, therefore, negatively associated with technical efficiency.
Similarly, we find that the coefficient on acres operated is negative and significant, indicating that the size
effect is negatively associated with technical inefficiency and, therefore, positively associated with
technical efficiency, confirming our hypothesis. Among the other factors influencing efficiency we find
that the coefficients on acres rented, education, and biosoy are significant and also positively influence the
efficiency frontier. In contrast, we find that the coefficients on age, biocorn and excess nitrogen are
significant but negatively influence the efficiency frontier. And, we find that the ratio of off-farm earnings
to farm earnings to be significantly related to technical inefficiency. This is surprising given the focus on
farms producing corn.
Using the coefficients found in Table 4, an increase in farm size of 10 percent would increase the
efficiency of production on the corn farms analyzed by 4.6 percent. Similarly, an increase in rented land of
10 percent would increase efficiency by about 2.4 percent.
We find the mean technical efficiency score for all farmers is 0.800. This set of results implies that our
sample of farms could reduce their inputs by about 20 percent without compromising output if they could
achieve best management practices by producing on the frontier. Our preliminary estimate of returns to
scale is 0.65, i.e. hog farms, on average, exhibit increasing returns to scale. We also find that independent
operations exhibit significantly lower returns to scale than contract operations, 0.63 compared to 0.71,
indicating that contract operations are, on average, slightly more scale efficient than independent
operations; that is, independent operations are, on average relatively too small. The t-test for the
comparison of means of the two groups is 3.33. More interestingly, the returns to scale are roughly
10
comparable for the largest 50 percent of contracting operations (ranked by level of livestock output)
compared to the 50 percent of independent operations—0.77 versus 0.73. In contrast, the returns to scale
of the lowest 50 percent of contracting operations ranked in terms of livestock output are estimated at 0.67
compared to only 0.48 for the smallest 50 percent of independent operations.
Nutrient Balance Results
We find that excess nitrogen and phosphorous levels per acre operated appear to have remained fairly
constant during the period of analysis. Based on the USDA survey data analyzed, average excess nitrogen
(phosphorous) per acre operated hovered at close to 30 (20) pounds during 1997-2001. We also find that
excess nutrient levels are generally significantly higher on contract operations than on independent
operations. For example, in 1999, contract operations exhibited 48.32 pounds of excess nitrogen compared
to 26.32 pounds on independent operations. The t-test for the comparison of means for the two excess
nitrogen groups is 2.46. Similarly, contract operations in 1999 exhibited 42.03 pounds of excess
phosphorous compared to 17.84 pounds on independent operations. The t-test for the comparison of means
for the two excess phosphorous groups is 2.85. As shown in Table 5 contract operations exhibited
significantly more excess nitrogen per acre operated than independent operations in 2000 and 1999 and
significantly more excess phosphorous than independent operations in all years analyzed except 1997.
Summary and Conclusions
Preliminary estimates of technical efficiency based on USDA data for 1997 through 2001 indicate
that independent operations were significantly more efficient than contract operations. Preliminary
estimates also indicate that both types of operations exhibited increasing returns to scale with contract
operations appearing to exhibit significantly higher returns to scale than independent operations. The
returns scale results suggest that small independent operations, in particular, are too small to be
economically competitive. Our estimates of excess nutrients that derive from both commercial fertilizer
and manure, comparing the performance of production contract operations and independent operations,
indicate that, in general, levels of excess nutrients per acre of land are significantly higher on contract
operations than independent operations. The results suggest that adjusting the performance measures to
11
include excess nutrients as a “bad output” would tend to favor independent producers over contract
operations compared to performance measures that ignore pollution.
In future research it would be desirable to assess costs of production by type of operation. Our
preliminary results suggest feed costs, in particular may differ significantly by type of operation.
Finally, additional data available in 2002, likely to include many more hog observations than 2000 and
2001 because it is a census year, could strengthen the results.
12
References
Aigner, D.J., C.A.K. Lovell, and P. Schmidt, "Formulation and Estimation of Stochastic Frontier
Production Function Models." Journal of Econometrics. 6(1977):21-37.
Bagi, F.S. "Relationship Between Farm Size and Technical Efficiency in West Tennessee Agriculture."
Southern Journal of Agricultural Economics. 14(1982):139-44.
Ball, Eldon, Jean-Pierre Butault, and Richard Nehring, "United States Agriculture, 1960-96: A Multilateral
Comparison of Total Factor Productivity,@ E RS Staff paper, United States Department of Agriculture,
AGES 00-03. Washington DC, 2000.
Ball, Eldon, Jean-Pierre Butault, and Richard Nehring, "Levels of Farm Sector Productivity: An
International Comparison,@ Journal of Productivity Analysis. 15(2000):287-311.
Ball, Eldon, Jean-Pierre Butault, Richard Nehring, and Agapi Somwaru "Agricultural Productivity
Revisited,@ Amer. J. of Ag. Econ.. 79(1997):1045-1063.
Battese, G. E. and S. Broca, "Functional Forms of Stochastic Frontier Production Functions and Models
for Technical Inefficiency Effects: A Comparative Study for Wheat Farmers in Pakistan," Journal of
Productivity Analysis. 8(1997):395-414.
Byrnes, P., R. Fare, S. Grosskopf and S. Kraft, "Technical Efficiency and Size: the Case of Illinois Grain
Farms." European Review of Agricultural Economics. 14-4(1987):367-81.
Carter, Collin and Andrew J. Estrin, "Market Versus Structural Reforms in Rural China." Journal of
Comparative Economics. 29(2001):527-41.
Coelli, Tim, "A Guide to FRONTIER Version 4.1: A Computer Program for Stochastic Frontier
Production and Cost Function Estimation," mimeo, Department of Econometrics, University of New
England, Armidale, 1994.
Coelli, Tim, "A Guide to DEAP Version 2.0: A Data Envelopment Analysis (Computer) Program,"
mimeo, Department of Econometrics, University of New England, Armidale, 1996.
Coelli, Tim and George Battese, "Identification of Factors which Influence the Technical Inefficiency of
Indian Farmers." Australian Journal of Agricultural Economics. 40-2(1996):103-28.
Coelli, Tim, D. S. Prasada Rao, and George Battese, An Introduction to Efficiency and Productivity
Analysis. Kluwer Academic Publishers, 1998.
Coelli, Tim and Sergio Perelman, "Technical Efficiency of European Railways: A Distance Function
Approach." Applied Economics 32(2000):1967-1976.
13
Deaton A.,”Panel Data From Time Series Cross-Sections," Journal of Econometrics. 30(1985):109-126.
Dubman, Robert W. AVariance Estimation with USDA=s Farm Costs and Returns Surveys and Agricultural
Resource Management Study [email protected], U.S. Dept. of Agriculture, Economic Research Service, AGES 0001, Washington D.C., April 2000.
Färe, R., S. Grosskopf, M. Norris and Z. Zhang, “Productivity Frowth, Technical Progress, and Efficency
Changes in Industrialized Countries,” Amer. Econ. Review , 84(1994):66-83.
Farrell, M. J., “The Measurement of Productive Efficiency,” Journal of the Royal Statistical Society
Society, Series Series A, CXX, Part 3(1957):253-290.
Fixen, Paul E. and Ford B. West, “Nitrogen Fertilizers: Meeting Contemporary Challenges,” Ambio, A
Journal of the Human Environment, Vol. 31, No. 2 (2002): 169-176.
Heshmati Almas and Subal C. Kumbhakar, “Estimation of Technical Efficiency in Swedish Crop Farms:A
Pseudo Panel Data Approach,” J. of Agr. Econ. 74(1992):745-750.
Hjalmarsson, L., S. C. Kumbhakar and A. Heshmati. "DEA, DFA and SFA: A comparison," Journal of
Productivity Analysis. 7(1996):303-327.
Hoppe, Robert A., Janet Perry and David Banker, "ERS Farm Typology: Classifying a Diverse Ag
Sector", Agricultural Outlook, AGO-266, ERS, USDA, November 1999.
Jondrow, J. C.A.K. Lovell, I.S. Materov, and P. Schmidt. "On the Estimation of Technical Inefficiency in
the Stochastic Frontier Production Model." Journal of Econometrics. 19(1982):223-238.
Kaliragan, K, “Farm-Specific Technical Efficiencies and Development Policies,” Journal of Economic
Studies, ( 11(1984):3-13.
Kellog, Robert L., Charles H. Lander, David C. Moffitt, and Noel Gollehon, “Manure Nutrients Relative
to the Capacity of Cropland and Pastureland to Assimilate Nutrients: Spatial and Temporal Trends for the
United States,” USDA Publication nps00-0579, Washington, D.C., December 2002.
Key, Nigel, and William MacBride, “Production Contracts and Productivity in the U.S. Hog Sector,”
Paper presented at Annual AAEA meetings in Long Beach CA, July 2002.
Kott, Phillip S. AUsing the Delete-a-Group Jacknife Variance Estimator in NASS [email protected], U.S. Dept. of
Agriculture, National Agricultural Statistics Service, RD-97-xx, Washington, D.C., December 1997.
Kumbhakar, Subal C., Basudeb Biswas and Dee Von Bailey. " A Study of Economic Efficiency of Utah
Farmers: A System Approach." Review of Economics and Statistics. 71((Nov. 1989):595-604.
Lau L. J., Yotopoulos P. A. “A Test for relative efficiency and application to Indian agriculture,"
American Economic Review. 61(1):94-109.
McBride, William D. & Lee A. Christensen—Environmental and Economic Dimensions of the Hog
Manure Problem. Selected paper. American Agricultural Economics Association meeting. August 2000.
Tampa, FL.
14
McBride, William D. and Nigel Key. Economic and Structural Relationships in U.S. Hog Production.
Agricultural Economics Report 818. Economic Research Service, USDA. Febriaru 2003.
Morrison Paul, Catherine J., Warren E. Johnston, and Gerald A. G. Frengley, “Efficiency in New Zealand
Sheep and Beef Farming: The Impacts of Regulatory Reform.” The Review of Economics and
Statistics.82(2):325-337.
National Commission on Small Farms, "A Time to Act," A Report of the USDA National Commission on
Small Farms, January 1998.
Peterson, Willis, "Are Large Farms More Efficient?" Staff Paper P97-2, Department of Applied
Economics, College of Agricultural, Food and Environmental Sciences, University of Minnesota January
1999.
Ribaudo, Marc. Managing Manure-New Clean Water Act Regulations Create Imperative for Livestock
Producers. Amber Waves, Volume 1, No. 1. February 2003.
Ribaudo, M.O., N.R. Gollehon and J. Agapoff. “Land Application of Manure by Animal Feeding
Operations: Is More Land Needed?” Journal of Soil and Water Conservation, Vol. 59, No. 1 (2003): 3038.
Sharma, Khem, R., Pingsun Leung, and Halina M. Zaleski, “Technical, allocative and economic efficiencies in
swine production in Hawaii: a comparison of parametric and nonparametric approaches, Agricultural Economics,
20(1999):23-35
United States Department of Agriculture, “Agricultural Prices 1999 Summary,” National Agricultural
Statistics Service, Washington, DC, July 2000.
United States Department of Agriculture, “Agricultural Statistics 2002,” United States Government
Printing Office, Washington, DC, 2002.
United States Department of Agriculture, “Agricultural Statistics 1998,” United States Government
Printing Office, Washington, DC, 1998.
Westcott, Paul C. and C. Edwin Young, “U.S. Farm Program Benefits:Links to Planting Decisions and
Agricultural Markets,” Ag. Outlook, October 2000.
15
16
Table 1. The Farm Typology Groups
Small Family Farms (sales less than $250,000)
1. 1. Limited-resource. Any small farm with: gross sales less than $100,000, total farm
assets less $150,000, and total operator household income less than $20,000. Limitedresource farmers may report farming, a nonfarm occupation, or retirement as their major
occupation
2. 2. Retirement. Small farms whose operators report they are retired (excludes limitedresource farms operated by retired farmers).
3. 3. Residential/lifestyle. Small farms whose operators report a major occupation other
than farming (excludes limited-resource farms with operators reporting a nonfarm major
occupation).
4. 4. Farming occupation/lower-sales. Small farms with sales less than $100,000 whose
operators report farming as their major occupation (excludes limited-resource farms
whose operators report farming as their major occupation).
5. 5. Farming occupation/higher-sales. Small farms with sales between $100,000 and
$249,999 whose operators report farming as their major occupation.
Other Farms
6. Large family farms. Sales between $250,000 and $499,999.
7. Very large family farms. Sales of $500,000 or more
8.
8. Nonfamily farms. Farms organized as nonfamily corporations or cooperatives, as
well as farms operated by hired managers
Source: U.S. Department of Agriculture, Economic Research Service
17
Table 2. Group Definitions by Agricultural Statistics Districts Groupings
-----------------------------------------------------------------------------------------------------------------------------------Cohort
GV Sales
COH1
<100,000
COH2
100,000-249,999
COH3
250,000-324,999
COH4
325,000-499,999
COH5
500,000-749,999
COH6
750,000-999,999
COH7
>1,000,000
--------------------------------------------------------------------------------------------------------------------------------------------------------------
18
Table 3: Summary Statistics for Selected Variables in Hog States, 2000
-----------------------------------------------------------------------------------------------------------------------------------------------------------Farms
Corn Soybeans Excess Excess Livestock labor Acres Age Educ.
Type
(%)
Nitrogen Phos
--- dollars ------- ---# per acre------ --Dollars per farm-----------------------------------------------------------------------------------------------------------------------------------------------------------Limited
1.0
14,183 9,113
32.91 15.63 15,764 27,646 202 37.99 1.64
Resource
Retirement
1.9
0
Residental/ 21.8
lifestyle
2,988
Farming/
21.4
lower sales
9,784
0
0
10,363 16,796
156 67.67 1.00
3,476
23.87
15.91
26,937 16,755
95 47.27 2.67
9,857
11.93
7.70
32,310 30,750
460 53.60 2.16
27,570 26,880
24.38
16.68
87,756 33,945
633 48.02 2.28
43,794 44,837
21.82
19.72 239,585 37,884
800 46.57 2.84
Very Large 12.5
Family Farms
59,410 72,874
4.10
3.48
8,796 47.20 2.63
Nonfamily
Farms
39,397 78,875
37.71
30.55 602,123 52,393
Farming/
23.5
higher sales
Large
16.6
family farms
0.9
0
904,917 74,516
887 49.98 3.67
All Farms 100.0
26,772 7,170
25.41 18.90 189,617 34,994
718 52.05 2.47
------------------------------------------------------------------------------------------------------------------------------------------------------------
19
Table 4. Input Distance function Results
-----------------------------------------------------------------------------------------------Variable
Parameter t-test Description of variable
----------------------------------------------------α0
12.740 (11.48) constant
α XF
0.030
(0.15) fertilizer
α XL
-1.162
(4.71) labor
α XE
0.115
(1.48) fuel
α XM
0.162
(1.34) feed or miscellaneous
α XK
-0.614
(5.25) capital
α Y1
-0.450
(4.92) corn output
α Y2
-0.286
(0.28) soybean output
α Y3
0.174
(2.47) other crop output
α Y4
-0.707
(4.64) livestock output
βXF*XF
-0.032 (3.03) fert*fert
βXL*XL
0.054 (1.63) labor*labor
βXE*XE
0.001 (0.34) fuel*fuel
βXM*XM
-0.067 (2.71) feed*feed
βXK*XK
-0.081 (4.69) capital*capital
α Y1*YI
0.034 (4.86) corn*corn
α Y2*Y2
0.012 (3.93)
soybeans*soybeans
α Y3*Y3
0.047 (5.74) other crops*other crops
α Y4*Y4
0.047 (7.20)
livestock*livestock
α Y1*Y2
0.005 (1.49) corn*soybeans
α Y1*Y3
-0.006 (1.66)
corn*other crops
α Y1*Y4
0.026 (3.30)
corn*livestock
α Y2*Y3
0.001 (0.23)
soybeans*other crops
α Y2*Y4
-0.014 (1.80)
soybeans*livestock
α Y3*Y4
-0.020 (4.09)
other crops*livestock
βXF*XL
-0.028 (1.46)
fert*labor
βXF*XE
-0.016 (2.01)
fert*fuel
βXF*XM
-0.023 (1.15)
fert*feed
βXF*XK
0.046 (2.30)
fert*capital
ϕ XF*Y1
0.033 (3.18)
fert*corn
ϕ XF*Y2
-0.039 (3.23)
fert*soybeans
ϕ XF*Y3
0.019 (2.31)
fert*other crops
ϕ XF*Y4
-0.029 (2.06)
fert*livestock
βXL*XE
-0.040 (2.67)
labor*fuel
βXL*XM
-0.093 (1.99)
labor*feed
βXL*XK
0.093 (2.28)
labor*capital
ϕ XL*Y4
0.077 (3.73)
labor*livestock
βXE*XM
0.038 (3.14)
fuel*feed
βXE*XK
-0.001 (0.15)
fuel*capital
ϕ XE*Y4
-0.016 (2.26)
fuel*livestock
20
Table 4. Input Distance function Results (continued)
----------------------------------------------------------------------------------------------Variable
Parameter t-test Description of Variable
-----------------------------------------------------βXM*XK
0.094 (2.62)
feed*capital
ϕ XM*Y1
-0.028 (3.13)
feed*corn
ϕ XK*Y2
0.046 (4.63)
capital*soybeans
α 1997
-0.007 (0.21)
α 1998
0.106 (2.61)
α 1999
0.133 (3.08)
α 2000
0.210 (4.54)
α 2001
-0.211 (3.35)
δ0
0.527 (2.91)
δ contract
0.961 (9.26)
δ acres
-0.464 (6.02)
δ AGE
0.561 (3.43)
δ ED
-0.443 (1.76)
δ DEBT
0.049 (0.87)
δ RENT
-0.235 (4.57)
δ BIOCORN
0.129 (2.63)
δ BIOSOY
-0.076 (1.65)
δ OFF-FARM
0.108 (2.51)
δ XN
-0.117 (2.54)
δ XP
-0.008 (0.13)
δ cohort2
-0.013 (0.18)
δ cohort3
0.262 (1.87)
δ2
0.094 (9.21)
γ
0.736 (21.10)
Log-liklihood
239.159
21
Table 5. Nutrient Balance Comparisons 1997-2001
-----------------------------------------------------------------------------------------------------------------------------------------Year
Independent
Contracting
t-test
Operations
Operations
-----------------------------------------------------------------------------------------------------------------------------------------------------pounds per acre operated-----------------------------------------------------------------------------------------------------------------------------------------------------------2001
Excess N
22.79
28.52
0.52
Excess P
16.79
29.90
1.92
Obs
249
66
2000
Excess N
Excess P
Obs
24.31
17.87
276
37.59
30.16
66
1.77
3.04
1999
Excess N
Excess P
Obs
26.33
16.79
455
48.32
29.90
75
2.46
2.84
1998
Excess N
Excess P
Obs
25.56
18.67
1698
31.26
30.92
256
1.11
3.32
1997
Excess N
Excess P
Obs
34.30
25.20
438
28.36
22.24
79
-1.12
1.03
--------------------------------------------------------------------------------------------------------------------------------------------------
22
Figure 1.
Share of the Value of Production Under Contract
for Hogs
70
60
Percent
50
40
30
20
10
0
1994-1995
1996-1997
1998-2000
23
2001
Figure 2. Hog States Surveyed*
*ASD (Agricultural Statistics District)
*Total value of livestock-value of hog production/total value of livestock, where zero indicates 100 percent
specialization.
24
Fly UP