International Journal of FoodandAgriculturalEconomics

ISSN 2147-8988
Vol. 2 No. 2 pp. 169-178

AN ESTIMATION OF TECHNICAL EFFICIENCY OF GARLIC

PRODUCTION IN KHYBER PAKHTUNKHWA PAKISTAN

Nabeel Hussain
The University of Agriculture, Peshawar-Pakistan
Email:hussain_nabeel@yahoo.com

Shahid Ali
The University of Agriculture, Peshawar-Pakistan

Naveed Miraj
The University of Agriculture, Peshawar-Pakistan

Muhammad Sajjad
Agriculture Livestock & Cooperation Department, Khyber Pakhtunkhwa
Peshawar-Pakistan
Abstract

This study was conducted to estimate the technical efficiency of farmers in garlic
production in Khyber Pakhtunkhwa province, Pakistan. Data was randomly collected from
110 farmers using multistage sampling technique. Maximum likelihood estimation technique
was used to estimate Cob-Douglas frontier production function. The analysis revealed that
the estimated mean technical efficiency was 77 percent indicating that total output can be
further increased with efficient use of resources and technology. The estimated gamma value
was found to be 0.93 which shows 93% variation in garlic output due to inefficiency factors.
The analysis further revealed that seed rate, tractor hours, fertilizer, FYM and weedicides
were positive and statistically significant production factors. The results also show that age
and education were statistically significant inefficiency factors, age having positive and
education having negative relationship with the output of garlic. This study suggests that in
order to increase the production of garlic by taking advantage of their high efficiency level,
the government should invest in the research and development aspects for introducing good
quality seeds to increase garlic productivity and should organize training programs to
educate farmers about garlic production.

Key words: Cobb Douglas, garlic, Khyber Pakhtunkhwa, MLE, stochastic frontier,
technical efficiency

1. Introduction

Garlic (Allium sativum, Linn.) is a perennial herb belonging to the family Alliaceae and
is widely used around the world as a condiment or for seasoning in different cuisines. It is
grown throughout Pakistan and consumed by most of the people because of its pungent,
spicy flavor that mellows and sweetens considerably with cooking. Research shows that
garlic may help guard against heart diseases and cancer. In addition, garlic reduces
cholesterol and blood pressure levels due to the presence of certain antioxidants.
Tissue culture techniques have an extremely important role to play in the conservation of
certain plants of economic importance. Of particular significance are those crops which are

169
An Estimation of Technical Efficiency of Garlic Production…

normally vegetative propagated: garlic falls into this category (Gizawy & Lloyd, 1986). If
Pakistan can boost the cultivation of garlic, the country can also export it with other
condiments and spices (Ahmad, 2010). The rice, garlic, and mint; rice, potato, onion, and
mint; and a rice, chamomile, and mint cropping sequences increased the total equivalent
yield of oil by 276.3, 284.8, and 230 percent, respectively, as compared with a rice and
wheat cropping sequence. The rice, garlic, and mint cropping sequence increased net return,
production efficiency, and benefit as compared with other cropping sequences (Ram &
Kumar, 1996).
According to the United Nation’s Food and Agriculture Organization (FAO) estimates
world production of garlic is 22.23 million metric ton (MMT) approximately. Asia is the
largest garlic producing continent in the world and it contributes more than 80% to the total
world garlic production. China is the leading garlic producing country in 2010, which
produced 18.56 MMT of garlic accounting for over 77% of world output followed by India,
South Korea, Egypt, Russia, Myanmar, Ethiopia, USA, Bangladesh and Ukraine
respectively (FAO, 2010).
According to commodity wise imports garlic is the 19th most important import
commodity of Pakistan after palm oil, rapeseed, sugar refined, cotton lint, cake of soybeans,
chick peas, sunflower seed, dry onion, dry peas, jute, tea, wheat, arecanuts, flour of wheat,
tomatoes, residuals of fatty subs, lentils and fatty acids (FAO, 2010).
Pakistan ranked among the top 10 garlic importing countries of the world from 2001-
2010 with the exception of year 2001 and 2002 in which Pakistan is ranked as 13th and 14th
largest importer of garlic. From 2001-2010 Pakistan ranked only twice among the top 20
garlic exporting countries of the world, in the year 2003 and 2008. In the year 2003 and
2008 Pakistan was net importer as well as net exporter of garlic (FAO, 2010).
Forty four percent (44%) of the total garlic production of Pakistan is contributed by
Punjab followed by Khyber Pakhtunkhwa, Balochistan and Sindh with thirty five per cent
(35%) twelve per cent (13%) and eight per cent (8%) contribution, respectively. Concerning
the increase in area and production of garlic it has progressed well (GoP, 2010-11).
Increase in domestic production of garlic can be achieved by enhancing productivity of
garlic crop. Productivity can be accelerated by introducing new technology or by
improvement in efficiency or both. In Pakistan the adoption rate of new technology in
Pakistan is very slow, therefore improvement in efficiency is an appropriate option to
increase the agriculture productivity in short run (Javed et al. 2008). Measurement of the
efficiency of agricultural production is an important issue in developing countries.
Efficiency was introduced by Farrell (1957), who proposed that the two components,
technical and allocative efficiencies, combine together to give a measure of economic
efficiency. The term technical efficiency of a farm is its ability to produce the largest
possible potential output from existing set of inputs and existing technology while allocative
efficiency refers to the ability of a firm to produce at a given level of output using the least
cost combination of inputs. An effective economic development strategy depends on
enhancing productivity and output growth in agricultural sector (Bravo-Ureta & Pinheiro,
1997). Enhanced productivity increases return to the producers as well as to labor and
enables larger consumption of goods and services per person (Productivity Commission,
2013).
This study will provide help to farmers to identify factors that affect garlic growers’s
technical efficiency and determining the opportunity for increasing output. The findings of
this study will also be beneficial for policy makers to form sound programs related to
expand garlic production potential more effectively. This study, therefore, is an attempt to
assess technical efficiency of various resources used in the production process of garlic in
Khyber Pakhtunkhwa.

170
 N. Hussain, S. Ali, N. Miraj and M. Sajjad

2. Materials and Methods

2.1 Data Collection and Sample Size

This study was carried out in district Swabi of Khyber Pakhtunkhwa, Pakistan.
Multistage sampling technique was used for the sample selection. In the first stage of this
sampling technique district Swabi was purposively selected. In the second stage out of 17
villages 03 major garlic producing villages namely Jalbai, Jalsai and ToorDheer were
randomly selected. Based on the information taken from the local patwari of the three
villages there were 422 farmers in Jalbai, 476 in Jalsai and 510 in ToorDher. In the third
and final stage of multistage sampling technique, 110 garlic producers were selected through
proportional allocation sampling technique from each randomly selected village.
For measuring relationship between output and input relationship, mean technical
efficiency and technical inefficiency in garlic production data was analyzed by using the ML
estimates of the stochastic frontier model.

2.2 Model Specification

Stochastic frontier production model was developed by Aigner et al. (1977). Meeusen
and Broeck (1977) composed the error model. Their work was based upon the measure of
technical efficiency of Farrell in 1957.
Following Bravo-Ureta and Rieger(1991) the stochastic production function is defined
as:

Yi = f (Xi; β) + єi i = 1, 2, 3,..., n (1)

Where;Yi represents output of garlic for the ith farmer in Kgs/ha, f(X; ß) is a suitable
function such as Cobb-Douglas production function, Xi are the inputs used in production of
onion in units/ha, βi are the coefficients to be estimated, єi is a composed error term that
captures the error term and inefficiency component (vi,ui).The vi are random errors
associated with measurement errors in the yields of garlic reported or the combined effects
of input variables not included in the production function. The ui is assuming independent
and is obtaining by truncation (at zero) of the normal distribution with mean µ, and variance
σ2u. So, the specified empirical model of the Cobb-Douglas production function for the
garlic growers was given as follows:

lnYield= β0 + β1lnSeed + β2lnTrctrHrs + β3lnLabor + β4lnFert+ β5lnFYM + β6lnIrrig +

β7lnWeed + єi (2)

Where;
Yield = Yield of garlic in kg per hectare
Seed = Seed rate used in kg per hectare
TrctrHrs = Total tractor hours used per hectare
Fert = Chemical fertilizers (Urea and DAP) in kg per hectare
Labor = Total labor man days per hectare
FYM= Farm yard manure used in kg per hectare
Irrig = Number of irrigations per season
Weed= Volume of weedicides used for one hectare
βI = Coefficients to be estimated
єi = Composed error term

171
An Estimation of Technical Efficiency of Garlic Production…

єi = vi + ui
vi = Natural error term
ui =Technical inefficiency error term
ln = Natural logarithm

The explanatory variables (seed rate, tractor hours, labor, chemical fertilizers (Urea,
DAP), farm yard manure, number of irrigations and volume of weedicides) have been
incorporated for estimation of elasticities of production function and technical
efficiency/inefficiency factors in the model. These explanatory variables have the major
contribution to the cost of production therefore these were incorporated in the production
function.
This research work is being followed by keeping in view the research work done by
(Adewumi & Adebayo, 2008) and Wakili (2006) who applied the seed rate, Kibaara (2005)
applied the labors and farm yard manure, Obwona (2006), Balbalola et al (2009) and
Maganga (2012) applied the fertilizers.
The inefficiency model based on Battese and Coelli (1995) was specified as follows:

µi = g(Zi : σi ) (3)

µi =σo + σ1AGE + σ2EXP + σ3EDU + σ4FARM SIZE + ωi (4)

Where;
µi =Technical inefficiency error term
σi = Coefficients to be estimated
AGE = Age of the garlic growers in years
EXP = Farming experience of the garlic growers in years
EDU = Education of the garlic growers in years
FARM SIZE = Area under garlic for the farmers in hectare
ωi = Random error term
Technical efficiency for individual farmer can be defined as the ratio between observed
output and corresponding frontier output, which can be expressed as follows:

TEi = Y0b / Yfr = f (β, X) + (vi + ui ) / f (β, X) + (vi ) (5)

Where Y0b is the observed output produced by the individual farmer and Yfr is the
frontier output, the maximum output that a farmer can produced from the given resources.
TE takes the value between 0 and 1.

2.3 Model Adequacy Tests

2.3.1 Tests for Detection of Heteroscedasticity Problem

The important assumption of the homoscedasticity of the classical linear regression

model is that, the variance of each disturbance term µi, appearing in the population
regression function are homoscedastic and symbolically it can be written as:

E(µi ²) = σ² i = 1,2,3…..,n (6)

If the above mentioned assumption is seriously violated then a problem of

heteroscedasticity will appear, which means that variance of the error term will no more
remain same and may result in overestimating the goodness of fit. Heteroscedasticity can

172
 N. Hussain, S. Ali, N. Miraj and M. Sajjad

cause ordinary least square estimates of the variance of the coefficients to be biased, leading
to type I or type II error which means that OLS is not BLUE (Best Linear Unbiased
Estimator). Heteroscedasticity mostly occurs in cross sectional data as ours, as compared to
the time series data (Gujarati & Porter, 2009).
The presence of heteroscedasticity can be tested by several methods but we use Breusch-
Pagan-Godfrey test and Goldfeld-Quandt test to find the heteroscedasticity in our data as
follows:

2.3.2 Goldfeld-Quandt Test

Procedure of Goldfeld-Quandt test is given as follows:

1. Arrange the data in ascending order according to the values of Xi, beginning with the
lowest value of X.
2. After omitting central observation ‘c’ the remaining data was divided into two groups
each of (n – c)/2 observations.
3. Run separate OLS regressions for each (n – c)/2 observations and obtain the residual sum
of squares (RSS) for each regression i.e. RSS1 for smaller values (the small variance group)
and RSS2 for larger values (the large variance group). Each RSS has (n – c – 2k)/2 df, where
k is the number of parameters including the intercept term.
4. Compute the ratio: λ = RSS2/df ÷ RSS1/df, If we take µi to be normally distributed and if
there is a valid assumption for homoscedasticity then “λ” of the above equation follow the
F-distribution with numerator and denominator df each of (n – c – 2k)/2 respectively. If the
computed value of λ (=F) is greater than the F-tabulated value at the chosen level of
significance, then we will reject the hypothesis of homoscedasticity other wise not (Gujarati
& Porter, 2009).

2.3.3. Breusch-Pagan-Godfrey (BPG) Test

Procedure of Breusch-Pagan-Godfrey test is given as follows:

1. Run the regression of model (2) by OLS and obtain the error terms µ1, µ2,
µ3,……………., µn.
2. Obtain σ² = Σ µi²/n. (7)
3. Construct “pi” variable by the following equation.
pi = µi²/ σ² (8)
4. Run the regression “pi” on the Z’s as fallows.
pi = α0 + α1Z1 + α2Z2 + α3Z3 + ………….+ αnZn + vi (9)
Where, vi is the error term for the above regression.
5. Obtain the explained sum of squares (ESS) from the regression eq. (9) and compute Θ, as
Θ = 1/2(ESS). Consider the normal distribution of µi that if there is homoscedasticity and if
the sample size n increases, then Θ ~ X²m-1, which shows that Θ follows the chi-square
distribution with (m-1) degrees of freedom. Therefore if the computed value of Θ (=X²)
exceeds the critical chi-square value at the chosen level of significance, then one can reject
the hypothesis of homoscedasticity otherwise accept it (Gujarati & Porter, 2009).

2.4 Test for Multicollinearity Problem

The basic assumption of our study is that explanatory variables should not be correlated
with each other and if this assumption is violated then it means there is a problem of
multicollenearity. Multicollinearity naturally occurs in time series data by showing “perfect”
or exact linear relationship among some or all the explanatory variables of the regression

173
An Estimation of Technical Efficiency of Garlic Production…

model. Multicollinearity problem can be detected by performing the correlation matrix test
(Gujarati & Porter, 2009).

3. Results and Discussion

3.1 Result of Goldfeld Quandt Test for Heteroscedasticity

The estimate value of λ is 1.66 which follows F distribution (Gujarati & Porter, 2009).
The critical F 0.05 (45, 45) value is 1.69. Since the estimated value does not exceed the critical
value, therefore, we accept the null hypothesis of homoscedasticity.

3.2 Result of Breusch-Pagan-Godfrey Test for Heteroscedasticity

The estimated value of Θ is 0.874 which follows chi-square distribution. Tabulated χ²0.05
(7)value is 14.0671. As our estimated value does not fall in the critical region, so we can not
reject the null hypothesis of homoscedasticity. This result reinforces that the model (2) is not
plagued with the problem of heteroscedsticity.

3.3 Result of Correlation Matrix

The above correlation matrix between explanatory variables shows that there is no linear
relationship among all the explanatory variables because none of the value is greater than
0.80 which suggests that the basic assumption of multicollinearity has not been violated so
multicollinearity problem does not exist in the model (Table1).

Table 1. Result of Correlation Matrix

Variables lnSeed lnTracthrs lnLAB lnFERT lnFYM lnIRRI lnWEED
lnSEED 1.000
lnTRAC 0.489 1.000
lnLAB 0.220 -0.129 1.000
lnFERT 0.678 0.569 0.210 1.000
lnFYM 0.509 0.406 0.076 0.464 1.000
lnIRRI -0.071 0.004 0.147 -0.069 0.093 1.000
lnWEED 0.175 0.470 -0.102 0.406 0.269 0.071 1.000
Note: Yield= Yield of garlic in kg per hectare

3.4 Major Determinants of Garlic Yield

To achieve higher and impressive level of garlic production, it is important to adopt

better practices of crop management. Seed rate, tractor hours, labor days, fertilizers,
farmyard manure, number of irrigations, weedicides and age, education, experience and
farm size under garlic of the respondents were the different factors which affect the garlic
yield.

3.5 Factors of Technical Efficiency

The effect of regressors on garlic production is discussed below in detail:

174
 N. Hussain, S. Ali, N. Miraj and M. Sajjad

The coefficient for the seed rate is positive which indicates a positive relationship
between seed rate and garlic production by implying that one percent change in seed rate
will increase garlic yield by 0.22 percent and is statistically significant showing the same
results as estimated by (Adewumi & Adebayo, 2008) and Wakili (2006). The positive
coefficient for tractor hours shows that one percent increase in tractor hours will increase the
garlic production by 0.24 percent and is statistically significant. The coefficient for labors is
negative and statistically insignificant, which shows a negative impact on the production of
garlic and these results are the same as the results of Kibaara (2005). The coefficient of
fertilizers is positive and statistically significant which indicates that one percent change in
fertilizers will increase garlic production by 0.29 percent and the results are in accordance
with the results of Obwona (2006), Balbalola et al (2009) and Maganga (2012). The
coefficient of farm yard manure is also positive and statistically significant which implies
that one percent change in farm yard manure will increase yield of garlic by 0.09 percent
and the same results were estimated by Kibaara (2005). Irrigation is insignificant and its
coefficient is also negative. Weedicide is statistically significant and its coefficient is also
positive which shows that one percent change in weedicide will increase garlic production
by 0.02 percent. It was concluded from maximum likelihood estimates of Frontier 4.1 that
seed rate, tractor hours, fertilizers, farm yard manure and weedicides are statistically
significant 1%, 5% and 10% level of significance with a positive coefficients. Labors and
irrigation are statistically insignificant and shows negative relationship with garlic yield. The
mean technical efficiency is found to be 0.78 indicating that there is a considerable room to
improve the garlic production in the study area through the efficient use of available
resources and technology.

3.6 The factors of technical inefficiency

The technical inefficiency factors are those factors which show their expected influence
on the technical efficiency of the garlic growers. The inefficiency factors were estimated by
using the estimated (σ) coefficients of the inefficiency effects. The inefficiency effects were
specified as those related to age, education, farm size and farming experience under garlic.

Age: The coefficient of age is positive and significant, which shows a significant positive
relationship with technical inefficiency. This result is the same as that found by Msuya and
Ashimogo (2005).

Farming experience: The coefficient of farming experience is negative as well as

insignificant showing that the relationship between farming experience and technical
inefficiency is negative. Msuya and Ashimogo (2005) also found the same result.

Education: Education plays an important role in adoption of better technology and

achieving high output. The coefficient for education is negative and significant which shows
negative relationship with technical inefficiency and positive effect on garlic yield.

Farm size: The coefficient of farm size is statistically insignificant and the negative
coefficient of farm size implies that farm size of farmers has positive but insignificant
impact on yield of garlic.
So only age and education were statistically significant and age coefficients were
positive while coefficients of education were negative. Farming experience and farm size is
statistically insignificant and their coefficients were negative.

175
An Estimation of Technical Efficiency of Garlic Production…

3.7 Variance Parameters

Maximum likelihood estimates was used to estimate the gamma value with the
estimation of mean technical efficiency and the value of parameter estimates for the
inefficiency effects model.
The theory says that the true value of gamma should be greater than zero but less than
one and its value can be calculated through the estimated values of variance parameters δ2
and δu. The maximum likelihood estimates (MLE) of the stochastic frontier production
function estimates a positive coefficient of variance parameter (σ 2) which is significant at
10% level and shows goodness of the distributional assumption of the composite error term.
The value of gamma (γ) was calculated by the formula σ 2u/σ2, which is 0.9355 and
significant at 1% level indicating 93.55% variation in garlic yield due to inefficiency factors.

Table 2. Maximum Likelihood Estimates of the Stochastic Frontier Production

Function for Garlic in District Swabi (Dependent variable = Yield of garlic in kg/ha)
Variables Parameters Coefficients Standard error T ratios
Constant Β0 5.5744 0.6518 8.5515
Seed rate Β1 0.2249* 0.0693 3.2409
Tractor Hours Β2 0.2423* 0.0637 3.8048
Labor Β3 -0.0327 0.0518 -0.6307
Fertilizer Β4 0.2876* 0.0610 4.7130
FYM Β5 0.0961** 0.0464 2.0687
Irrigation Β6 -0.0694 0.0909 -0.7631
Weedicide Β7 0.0211*** 0.0132 1.5987
Inefficiency Effect Model
Constant σ0 0.6936 0.4532 1.5301
Age σ1 0.0116*** 0.0069 1.6812
Experience σ2 -0.0335 0.0278 -1.2050
Education σ3 -0.0392*** 0.0214 -1.8318
Farm Size σ4 -2.7184 2.6322 -1.0327
Variance Parameters
Sigma square σ2 0.3642 0.4208
Gamma Γ 0.9355 0.0725 12.9038
Mean efficiency X 0.7752
Note:*, **, *** are significant 1%, 5% and 10% level respectively.
It was concluded from the ML estimations of stochastic frontier Cobb-Douglas
production function, that seed rate, tractor hours, fertilizers, FYM and weedicides were
statistically significant at 1%, 5% and 10% level of significance with positive coefficients.
The inefficiency model shows that area and education under garlic crop are statistically
significant at 10% level of significance respectively, with age having positive and education
having negative coefficient.
The ML estimations of Frontier 4.1 estimates, the value of gamma (γ) is 0.9355and
significant at 1% level of significance, respectively. It shows that 93.55% variation is due to
inefficiency factors, respectively in the yield garlic. So, it indicates that the random error
component of inefficiency, significantly contribute to the analysis (Table 2).

176
 N. Hussain, S. Ali, N. Miraj and M. Sajjad

Table 3. Frequency Distribution of Technical Efficiency of Garlic Growers

Technical efficiency Frequency Percentage
<0.50 5 5
0.51 - 0.60 11 10
0.61 - 0.70 14 13
0.71 - 0.80 19 17
0.81 - 0.90 34 31
>0.90 27 25
Sample size 110
Minimum 0.21
Maximum 0.95
Mean 0.77

3.8 Frequency Distribution of Technical Efficiency of Garlic Growers

Table 3 shows the estimated technical efficiency’s frequency distribution for the garlic
growers. The minimum and maximum values for estimated technical efficiencies are 0.2109
and 0.9563 with a mean efficiency 0.7752 respectively. So these results indicate that by
using the available inputs the yield of garlic can be improved.

4. Conclusion and Recommendations

Present study applied Cobb Douglas stochastic frontier production function for the
estimation of technical efficiency of garlic production in Khyber Pakhtunkhwa province of
Pakistan. The result shows that the mean technical efficiency was 93.55 percent which
shows that farmers in the study area were highly efficient. The results of socio economic
characters such as age, education, farm size and experience shows that education is the only
factor which significantly affect farm technical inefficiency. Based upon these findings it is
recommended that government should introduce latest mechanical technology to
reduce/replace number of labors used in garlic production. Due to high prices of oil and
shortage of electricity in the country the farmers were facing problems to irrigate the farms,
so development of canal irrigation system to solve the irrigation problem of the famers will
boost up garlic production.

Refrences

Adewumi, M. O., & Adebayo, F. A. (2008). Profitability and Technical Efficiency of Sweet
Potato Production in Nigeria.Journal of Rural Development 31: 105-20.
Ahmad, M. S. (2010). Improving Garlic Production.Daily News Paper, Dawn.
Aigner, D. K., Lovell, C. K., & Schmidt, P. (1977). Formulation and Estimation of
Stochastic Frontier Production Function Models.Journal of Econometrics 6: 21-37.
Babalola, D. A., Ajani, O. I. Y., Omonona, B. T., Oni, O. A., & Awoyinka, Y. A. (2009).
Technical Efficiency Differential In Industrial Sugarcane Production:The Case Of
Jigawa State, Nigeria. Journal of Life and Physical Sciences 3: 56-59.
Battese, G. E., & Coelli, T. J. (1995). A Model for Technical Inefficiency Effects in a
Stochastic Frontier Production Function for Panel Data.Empirical Economics 20: 325-
32.

177
An Estimation of Technical Efficiency of Garlic Production…

Bravo-Ureta, B. E., & Pinheiro, A. E. (1997). Technical, Economic and Allocative

Efficiency in Peasent Farming: Evidence from the Domain Republic. The Developing
Economies 35: 48-67.
Bravo-Ureta, B. E., & Rieger, L. (2008).Alternative Production Frontier Methodologies and
Dairy Farm Efficiency. Journal of Agricultural Economics 41:215-26.
Farrell, M. J. (1957). The Measurement of Productive Efficiency.Journal of the Royal
Statistical Society (A, general)120: 253-81.
FAO. (2010). Food and Agriculture Organization, 2010.www.fao.org
Gizawy, A. M. E., & Ford-Lloyd, B. V. (1996). An in Vitro Method for the Conservation
and Storage of Garlic (Allium sativum) Germplasm. Plant Cell, Tissue and Organ Culture
9: 147-50.
GoP. (2011). Agricultural Statistics of Pakistan 2010-11.Ministry of Food and
Agriculture.Pakistan Bureau of Statistics, Agriculture Statistics Section Islamabad.
Gujarati, D. N., & Porter, D. C. (2009). Basics Econometrics.5th ed. McGraw Hill Inc. New
York, USA.
Javed, M. I., Adil, S. A., Javed, M. S., & Hassan, S. (2008). Efficiency Analysis of Rice-
Wheat system in Punjab Pakistan.Pakistan Journal of Agricultural Sciences 45 (3).
Kibaara, B. W. (2005). Technical Efficiency in Kenyan’s Maize Production: An Estimation
of the Stochastic Frontier Approach.
Mangana, A. M. (2012). Technical Efficiency and its determinants in Irish Potato
Production: Evidence from Dedza District, Central Malawi. African Journal of
Agricultural Research 7: 1794-99.
Meeusen, W., & Van den Broeck., J. (1977). Efficiency Estimation from Cobb-Douglas
Production Functions with Composed Error. International Economics Review 18: 435-
44.
Miko, S. (1999).Response of garlic (Allium sativum L.) to levels of nitrogen, phosphorus
and irrigation internals. Ph.D. Thesis, P.G. School(pp. 187).
Msuya, E., & Ashimogo, G. (2005). Estimation of Technical Efficiency in Tanzanian
Sugarcane Production: A Case Study of Mtibwa Sugar Estate Outgrowers Scheme.
Economic and Development Papers, Mzumbe University(pp. 28-46).
Obwona, M. (2006). Determinants of Technical Efficiency Differentials Amongst Small and
Medium Scale Farmers in Uganda: A Case of Tobacco Growers. African Economic
Research Consortium.
Ram, M., & Kumar, S. (1996). The Production and Economic Potential of Cropping
Sequences with Medicinal and Aromatic Crops in a Subtropical Environment.Journal of
Herbs, Spices & Medicinal Plants 4: 23-29.
Wakili, A. M. (2006). Economic Analysis of Cowpea Production in Nigeria.Russian Journal
of Agricultural and Socio-Economic Sciences.

178