Technical and Economic Efficiencies in Poultry Production in Imo State, Nigeria

This study was carried out to estimate the technical and economic efficiencies of poultry farmers in Imo State, Nigeria. The data was collected with semi-structured questionnaire from 140 randomly selected poultry farmers. A stochastic frontier production function was estimated by using the maximum likelihood estimation technique to obtain the technical and economic efficiencies of poultry farmers. The mean technical efficiency of poultry farmers was 75 percent, while their mean economic efficiency was 21 percent. The generalized likelihood test indicated that, the poultry farmers are not fully technically and economically efficient in resource use. There is 79% allowance to increase economic efficiency of poultry farmers by improvement in technical efficiency. Research Article American Journal of Experimental Agriculture, 3(4): 927-938, 2013 928


INTRODUCTION
Poultry farming is an important agribusiness enterprise that has a great potential for providing additional income to our farming community and educated unemployed persons of the rural areas through creating self employment opportunities [1] Poultry production is one of the important sub-sectors in the Nigerian economy [2].
In addition to its contribution to the Gross Domestic Product (GDP) and provision of employment opportunities, poultry production is a major source of protein in the country. Poultry meat is found to be one of the most nutritious and most complete food known to man and it provides a means by which rapid transformation of animal protein intake can be achieved [3]. Poultry products compare favourably with economically produced animal protein [4]. Despite the nutritive value of poultry meat its production in the country is grossly inadequate as reflected in the wide gap between demand and supply of the product. This could be attributed to numerous problems that poultry farmers in Nigeria are facing.
These problems include low capital base, inefficient management, technical inefficiency, economic inefficiency, diseases and parasites and poor housing [5], high cost of feeds, poor quality of day old chicks (DOC), inadequate extension and training facilities [6]. The poultry production capacity of farms has to increase rapidly to be able to meet up with the increasing demand, and for this to be achieved, the present level of technical and economic efficiency must be improved upon. Presently, there are dearth of studies on efficiency of poultry farms [7,8]. Particularly, there are limited information about the performance of poultry farms in terms of technical and economic efficiencies [9].
Efficiency is defined as how effectively a production unit uses variable resources for the purpose of profit maximization given the best production technology available [10]. However, a number of empirical studies on poultry production in Nigeria have focused more on production constraints [5,6], economic analysis [7,11], and profitability [3,9]. For other studies that attempted to ascertain the resource use efficiency of poultry farmers in Nigeria, the ordinary least squares (OLS) estimation techniques were used. These include works by [8,12,9,13]. The use of the ordinary least squares (OLS) estimating technique makes it difficult to determine farm level efficiency as it provides only an average function [14,15] though it provides consistent estimates of the parameters except the intercept (10). To overcome this shortcoming of the OLS, the stochastic frontier function was developed and has been used by several researchers to estimate efficiency in agricultural production. Its beauty lies in its ability to test and quantify the inefficiency of individual farmers in a sample because it allows for statistical noise rather than attributing all deviation to efficiency difference. It is also straight forward to implement and interpret [16]. A situation that is not possible with other partial measures of efficiency such as the OLS [17]. Most of the previous studies that used the stochastic frontier production function in agriculture of developing countries limited their work to the determinants of technical efficiency. Studies such as those of [18,19,20,21,22,23,24,10,17]. [25,26,27] estimated technical efficiency of crop farmers in developing countries, including Nigeria. There is a paucity of published empirical works carried out in Nigeria generally and Imo State specifically that have made use of the stochastic frontier production function to estimate technical and economic efficiency simultaneously in poultry production. This study is therefore, intended to use the stochastic frontier function to provide estimates of technical and economic efficiency in poultry production in Imo State, Nigeria.
This study was conducted in the year 2012 in the three agricultural zones of Imo State. The State lies within latitudes 5 0 40 1 and 7 0 05 1 North and longitude 5 0 35 1 and 8 0 30 1 East. It had a population of about 3.92 million people in 2006 (NPC, 2006). Imo State is divided into three agricultural zones of Owerri, Orlu and Okigwe, and further subdivided into 27 Local Government Areas (LGAs). Farming is the major occupation of the people.
The weather and environmental conditions of the state favour the production of poultry and other livestock such as sheep, goat, rabbit, pigs, etc. A pre-survey was initially carried out in each zone to identify poultry farmers through the assistance of Imo State Agricultural Development Programme (ADP) extension agents and officials of All Farmers Association of Nigeria (AFAN). The number of poultry farmers identified varied among the LGAs in each agricultural zone. Two LGAs from each zone that had the highest number of poultry farmers were purposively selected, making a total of six LGAs. The sampling frame was the list of poultry farmers in each selected LGA.
Proportionate sampling followed by random sampling techniques were employed in each LGA to selected the sample size of 140 poultry farmers. The study used mainly primary data which were collected with the aid of semi-structured questionnaire. Data were collected on variables such as socioeconomic characteristics of farmers, resource inputs and output in poultry production. Data were analyzed using descriptive statistics such as mean, frequency distribution and percentages, as well as stochastic frontier production function.

ANALYTICAL FRAMEWORK
The stochastic frontier production function was specified as; Where, Y is output of poultry, Xi is actual input vector, β is vector of production function parameters, V is random error term with zero mean, and U is non-negative one sided error term.
Given that the functional form for this study is self dual, i.e. Cobb-Douglas, the corresponding cost frontier can be derived and written as, Where, C is minimum cost associated with the production of poultry, P is vector of input prices, Y is output of poultry, γ is vector of parameters.
Using Sheppard's Lemma we can obtain This is the system of minimum cost input demand equations [28,16,15]. Substituting a farm's input prices and quantity of output in equation (3) yields the economically efficient input vector Xi. With observed levels of output given, the corresponding technically and economically efficient costs of production will be equal to Xi, P and Xi, respectively, while the actual operating input combination of the farm is Xi, P.
The three cost measures can then be used to compute the technical efficiency (TE) and economic efficiency (EE) indices as follows; Note however that, efficient production is represented by an index of 1.0 while the lower values indicate a greater degree of inefficiency. Using the method by [15,14] which was based on the work of [30], efficiency can then be measured using the adjusted output as shown in equation [6].
Where, U can be estimated as; Where, f(.) and F*(.) are standard normal density and cumulative distribution functions respectively, δ = δ/ δv, εi = V -U and δ 2 = δu + δv 2 Y* is the observed output adjusted for statistical noise. When εiδ and δ estimates, are replaced in equations (6) and (7), it will provide estimates for U and V.
In this study, a Cobb-Douglas function was fitted to the stochastic production frontier of the poultry farmers using the Maximum Likelihood method. This functional form has been used in many empirical studies particularly those relating to developing countries' agriculture [16,14]. It has been useful in estimating economic efficiency because it's self-dual. However, [14] had opined that functional form has limited effect on empirical efficiency measurement. The production function model is specified as follows; Where, Y is output of poultry (kg), X1 is quantity of feed (kg), X2 is flock size (number of birds), X3 is labour input (mandays), X4 is cost of drugs and medication (N), X5 is capital (N), X6 is cost of management (N), and X7 is other inputs (N), Ln is natural logarithm, δ0 -δ7 are coefficients to be estimated, εi is composed error term which is also defined as V-U. the use of single equation model is justified by the assumption that farmers maximize expected profits as it is often assumed in similar studies [15,14]. It is expected a priori that the coefficients of X1, X2,X3, X4, X5, X6, X7, will be positive. The cost frontier function is also specified thus; LnC = lna0 +a1lnPx1 + a2lnPx2 + a3lnPx3+a4lnPx4 +a5lnPx5 +a6ln Px6+a7lnPx7 +a8Y* +εi……. (11) Where, C is cost of production per poultry farmer, P 1 is cost of feed, P2 is cost of birds (N/bird), p3 is cost of labour (N/manday), p4 is cost of drugs and medication (N/bird), p5 is cost capital (N), p6 is cost of management (N/manday), p7 is other costs (N/bird), Y* is output of matured bird in kg adjusted for statistical noise, a 1 -a7 are parameters to be estimated, a0 is the y -intercept, and εi is the composed error term. It is expected a priori that the coefficients of Px1, Px2, Px3, Px4, Px5, Px6 and Px7 will be positive.

Socio-Economic Characteristics of Farmers
In The larger household sizes could be an advantage to the poultry farmers in the area of provision of household labour. The mean farming experience indicates that the poultry farmers are experienced enough in poultry production to understand the rudiments of poultry farming. The mean farm size implies that most of the poultry farmer are operating at small scale. The poultry farmers received poor extension contact which could lead to low adoption of poultry production technologies. The mean age of the farmers indicates that they are at their active stage of life to under take the level of operations involving in poultry production.

Estimation of Stochastic Production and Cost Frontier Functions
The estimates of the stochastic production frontier function (Table 3) indicate that, all the coefficients carried the expected positive signs. The coefficients of feed (X1), flock size (X2), labour (X3), and capital (X5) were significant at 1% level, while the coefficients of drugs and medication (X4), management (X6), and other inputs (X7) were significant at 5% level. The sum of the elasticities was 1.958, indicating that, the poultry farmers were operating in the region of increasing returns to scale. The gamma (γ) was 0.783 which was high enough and significant at 1% level. It gives an indication that the unexplained variations in output are the major sources of random errors. It also shows that about 78 percent of the variations in output of poultry farmers are caused by technical inefficiency. It also confirms the presence of the one-sided error component in the model and hence, the use of the OLS in estimating the function, becomes inadequate in representing the data. The sigma square (δ2) estimate was 0.531 and significant at 1%, and therefore, assures us of the goodness of fit and correctness of the distributional assumptions of the composite error. The generalized likelihood test gave a value of 22.73 which indicates that the farmers are not fully technically efficient. [16,10,17,14], obtained similar results in their different studies. In the cost frontier function (Table 4), all the variables carried the expected positive signs. The coefficients of feed cost (Px1), cost of birds (Px2), labour cost (Px3), and output (Y*) adjusted for statistical noise were significant at 1%, while the coefficients of cost of management (Px6), was significant at 5%. The coefficients of cost of drugs and medication (Px4), cost of capital (Px5), and cost of other inputs (Px7) were not significant even at 5% level.
The gamma (γ) estimate was 0.98 and was significant at 1% level indicating that 98% of the variation in output were caused by economic inefficiency. The sigma square (δ2) was 0.65 and was significant at 1% level, and indicated the goodness of fit and correctness of the specified assumptions of the distribution of the compound error term [15]. The generalized likelihood test gave a value of 67.25 which indicates that the farmers are not fully such economically efficient.

Estimates of Technical and Economic Efficiency in Poultry Production
The estimates of Technical Efficiency of poultry farmers are presented in Table 5 The high mean technical efficiency is suggestive of the fact that only a small fraction of the output of the output is attributed to resource wastage. The technical efficiency distribution agrees with that obtained by [10,17,14]. The Economic Efficiency estimates are presented in Table 6. It shows a range from 0.14 to 0.56. The mean economic efficiency in poultry production in the study area was 21%. The estimates also show that for the average poultry farmer to attain the level of the most economically efficiency farmer in the sample, he/she would experience a cost saving of 62.5 percent (i.e,1 -21/56) in poultry production.
However, the least economically efficient poultry farmer will experience efficiency gain of about 25 percent (i.e., 1 -14/56) in poultry production to attain the level of the most economically efficient farmer in the sample. The results are also lower than that obtained [16,14].

CONCLUSION
This study estimated technical and economic efficiencies of poultry farmers in Imo State, Nigeria. The study found that poultry farmers in the state are not fully technically and economically efficient in their use of resources and therefore, there is enough allowance to increase their efficiencies if some important policy variables are addressed. It has also shown that the major efficiency problem of the poultry farmers is not so much of technical. Therefore, the farmers with their current resource base and technology, if technical efficiency is improved can substantially improve economic efficiency, which is the product of technical and allocative efficiencies.