TECHNICAL EFFICIENCY OF CITRUS PRODUCTION IN SARGODHA DISTRICT, PUNJAB

Abstract

This study attempts to measure the Technical Efficiency (TE) of citrus farmers in Sargodha district based on primary data collected from 162 selected farms for the years 2005-6 and 2006-7. Cobb-Douglas production function was fitted to the data in order to find factors affecting citrus productivity which revealed that less amount of fertilizer use and heavy seasonal pruning were major determinants of low productivity. Nonparametric Data Envelopment Analysis was performed which concluded that farmers in 2006-7 were technically less efficient by more than half as compared to last year, which resulted in overall decrease in the citrus productivity in 2006-7 nearly one third of production of previous year. Malmquist Indices reveled a drastic decrease in the technical change, scale efficiency and total factor productivity change over time. The 'inverse farm size productivity relation' was also observed on citrus orchard possessing farms.

Full text

68
TECHNICAL EFFICIENCY OF CITRUS PRODUCTION IN SARGODHA DISTRICT, PUNJAB
Shahid Iqbal, Maqbool, H. Sial and Zakir Hussain*
Faculty of Management and Administrative Sciences, University of Sargodha, Pakistan
*
Faculty of Agriculture, University of Sargodha, Pakistan
ABSTRACT
This study attempts to measure the Technical Efficiency (TE) of citrus farmers in Sargodha district based on
primary data collected from 162 selected farms for the years 2005-6 and 2006-7. Cobb-Douglas production
function was fitted to the data in order to find factors affecting citrus productivity which revealed that less
amount of fertilizer use and heavy seasonal pruning were major determinants of low productivity.
Nonparametric Data Envelopment Analysis was performed which concluded that farmers in 2006-7 were
technically less efficient by more than half as compared to last year, which resulted in overall decrease in the
citrus productivity in 2006-7 nearly one third of production of previous year. Malmquist Indices reveled a
drastic decrease in the technical change, scale efficiency and total factor productivity change over time. The
‘inverse farm size productivity relation’ was also observed on citrus orchard possessing farms.
INTRODUCTION
Pakistan as a nation produces about 3 to 4 percent
of the world citrus fruits but sells out only about 0.8
percent of its harvest abroad. Pakistan is the sixth
largest producer of Kinnow (mandarin) and oranges
in the world, with 2.1 million tons. According to an
estimate nearly 95 percent of the total Kinnow
produced all over the world is grown in Pakistan
(Sharif & Waqar, 2005). Pakistan annually produces
about 1.70 million tons of citrus fruit predominantly
kinnow mandarin on the area around 185,000
hectares. Citrus fruit production share is about 40
percent of the total fruits produced in Pakistan.
Citrus fruit is grown in all four provinces of Pakistan.
Punjab produces over 95 percent of the crop and 70
percent of citrus grown in Punjab is under Kinnow
because of its greater population, favorable growing
conditions and adequate water (Nawaz, 2007). The
share of only citrus fruits in total value of fruit’ export
is about one-third. Owing to its increasing demand in
the domestic and international markets, production
of kinnow mandarin has not increased up to the
mark to meet the growing needs. Most of the
production is consumed locally in raw form without
any value addition; while exports range from
100,000 to 150,000 tons except the year 2005-06
witnessing all times high export of 200,000 tons and
a foreign exchange of $45 million was earned
through the export (Iqbal, 2006).
Historically, Kinnow is the parent cross of two
varieties made between “King mandarin” and “Willow
leaf mandarin” varieties developed by H.B Frost in
1915. It was done at the Citrus Research Centre of
the University of California Riverside US, but was
later released in 1935 (Khan, 2006). Kinnow was
introduced in Pakistan; before partition in Punjab
Agricultural College Lyallpur in the year 1943-44. In
the district of Sargodha it was introduced in early
50s in two villages; however oranges already existed
in the area (Warraich, 2006). In the past, it was
only marketed locally, but many firms has started its
export through its polishing and making value added
products such as juice concentrate, marmalade etc.
in the recent years. This has led to its increased
importance and attracted the attention of
researchers to analyze its productivity in the area.
The production of citrus was consistent from 1994 to
1998. The fruit yield during 1994-95 was 10 tons per
hectare and after five years in 1999-2000 it fell down
to about 9 tons per hectare [PAS 2001-6], which was
quite low when compared with developed countries
like United States, Australia etc., where average
yield was more than 30 tons per hectare (FAO
1998). Pakistan is even far below in yield/hectare,
even below from neighboring countries, India and
Iran. In recent years Pakistan’s average kinnow
yield is about 9.2 tons per hectare, while of all citrus
fruits its yield is 9.4 tons which reflects on poor
production performance when compared with
developed countries producing about more than 30
tons per hectare (Khan, 2006).
REVIEW OF LITERATURE
Dhehibi, et al. (2007) estimated farm level average
technical inefficiency 13.77 of citrus production by
fitting a stochastic frontier production function in
Tunisia. The authors found that the coefficients of
the share of productive trees, the agricultural
training, irrigation operations and the experience of
farmer estimated for technically efficient frontier had
the positive effect on technical efficiency of farmers.
The literature showed that Data envelopment
analysis or DEA is a linear programming technique
developed in the works of Charnes, Cooper and
Rhodes (1978) and Banker, Charnes, and Cooper
(1984). Various researchers have applied these
Int. J. Agric. Appl. Sci. Vol. 1, No.2, 2009

69
methods. The researchers are: Cinca et al. (2002);
Chapelle, and Plane(2005); Neto et al. (2006); Coelli
and Rao (2003);. Kuosmanen and Post
(2000);Camanho and Dyson (2006); Dimitrios and
Katerina (2006); Kuosmanen and Mika (2007);
Agnarsson(2000), Saleem (2005); Domah (2002).
Hypotheses to be tested
For studying the technical efficiency of citrus
farmers, we test the following hypotheses:
H1: The relationship between Yield and Number
of irrigations.
H2: The relationship between Yield and Number
of plants per field area.
H3: The relationship between Yield and Total
number of nutrients inputs in N P Ks.
H4: The relationship between Yield and Total
Area of the farm.
H5: The relationship between Yield and Other
crop dummy.
H6: The relationship between Yield and Number
of missing plants in the field.
H7: The relationship between Yield and Number
of leveling.
H8: The relationship between Yield and Number
of hoeings.
H10: The relationship between Yield and Age of
the plant.
H11: The relationship between Yield and
Aggregate circumference of the logs of
trees.
H12: The relationship between Yield and Distance
between the plants.
H13: All the farmers spatially efficient.
H14: All the farmers intertemporily efficient
Research Method
The empirical production analysis is key to economic
research. It involves three different questions: (1)
whether observed firm performance is consistent
with optimizing path which we call efficiency (in
allocation if inputs, in technology, technical change
etc.) or whether it is significantly deviated from
optimizing path which we call inefficiency, (2)
modeling the technology used by the firm, and
testing statistically the economies of scale and the
degree of technical progress, (3) one step ahead
forecast based on observed past performance. To
some extent, this is done in this study by using
Cobb-Douglas production function, the DEA and
Malmquist index technique which does not require
pricing data.
The degree farm’s economic behavior consistent
with optimizing path can be measured if we are able
to define efficiency in production and productivity of
a farm. Efficiency is defined as the measure of the
degree to which producers accomplish success in
allocating available inputs in such an optimum way
so that overall productivity or output they produce of
the farm is maximized. Efficiency may refer to either
technical or allocative and if both together we call
them economic efficiency. Productivity is the ratio of
all outputs to all inputs (Coelli et al., 2005).
Therefore, efficiency and productivity both measures
the performance of the farm. Efficiency estimation
involves the estimation of frontier functions of the
best performers and then measuring the
inefficiencies of the farms relative to the efficient
frontiers. Both, parametric and non-parametric
methods such as DEA are used to estimate frontier
function.
Hypotheses Testing
Citrus orchards have economic life up to 30 years.
There are underlying argo-economic factors, number
of irrigations, number of plants per field area, total
number of nutrients inputs in N.P.K., total Area of
the farm, number of ploughings, other crop dummy,
number of missing plants in the field, Number of
leveling, number of hoeings, age of the plant,
aggregate circumference of logs and distance
between the plants that determine the productivity
and efficiency. To know the significance of these
factors, a Cobb-Douglas production function is fitted
to the data using the method of Pooled Least
Squares (PLS) by pooling the data for the year
2005-6 and for the year 2006-7.
4.2 DEA is distribution free method based on
mathematical linear programming. This method
saves the researcher from highly restrictive
assumptions which are required when using a
parametric frontier. The present study employed
DEAP 2.1 software (Coelli, 1996) to analyze the
data for aforesaid DEA framework and DEA
Malmquist TFP indices mentioned below.
4.3 Malmquist indices are designed to measure the
Total Factor Productivity (TFP) change. Panel data
allow to measure TFP change indices using DEA.
These indices are decomposed into two parts: the
technical efficiency change and technical change.
Both of the changes occur because a Decision
Making Unit (DMU) i.e. firm shift from a previously
technical efficient frontier to a new more efficient
frontier. Now, as productivity is measured across a
large number of isoquants, each efficient at some
given point in time, the linear optimization programs
must be adjusted in such a way so that it account for
the change over time. Malmquist TFP index does
not require any price data. (Coelli,1996). With input-
oriented DEA, the linear programming model is
Shahid Iqbal, Maqbool, H. Sial and Zakir Hussain

70
designed so as to reckon how much the level inputs
of a firm could shrink if used efficiently(as is used by
efficient or/benchmark farmers within the sample) in
order to get the same level of output. In this study
both fixed and variable inputs have been used.
The Malmquist TFP index calculates the TFP
change occurred between two panel data sets (in
periods t and t+1) by computing the ratio of the
distances of each data point relative to a same
technology. The output orientated Malmquist TFP
index is computed from output distance function
(Coelli et al. 2005, p291).
21
1111
1
1
1101,
]
),(
),(
),(
),(
[),,,(
0
0
0
0
++++
+
+
+++
×==
Tt
t
tt
t
stt
t
tt
t
tttttt
xyD
xyD
xyD
xyD
xyxymTFP
When TFP index is computed one, it implies that
productivity over time has not changed, when it is
computed less than one it implies that productivity
has decreased and when it computed greater than
one it means that productivity increase over time.
Using output distance function, the Malmquist index
for productivity change is divided into efficiency
change; technical change and scale change (Coelli
et al. 2005, p58). To decompose the Malmquist TPF
index into aforesaid measures of change we can
compute the technical efficiency change over time
(from period t to t+1):
),(
),(
/
11
1
1
0
0
++
+
+
==∆
tt
t
tt
t
itit
xyD
xyD
TETETE
and the technical change:
21
11
11
11
]
),(
),(
),(
),(
[
0
0
0
0
++
++
++
×=∆
tt
t
tt
t
tt
t
tt
t
xyD
xyD
xyD
xyD
T
The Malmquist TFP index is thus the product of the
technical efficiency change and the technical
change, i.e.
∆
×
∆
=
T
TE
TFP
.
One can further decompose the Malmquist
productivity index. The output-orientated Malmquist
productivity index is further subdivided into four
factors: namely, Technical Change (
∆
T
)
, Technical
Efficiency Change (
∆
TE
)
, scale effects (SEC) and
an Output Mix Effect (OME i.e.
OME
SEC
TE
T
TFP
⋅
⋅
∆
⋅
∆
=
). Alternately the
Malmquist TFP index can be subdivided into
technical efficiency change, and technical change
which is further subdivided into input bias and output
bias. As the study uses two successive years’ data,
the above mentioned measures of technical
efficiency were calculated using DEAP 2.1 software.
Scope of the research
Sargodha district produces the largest amount of
citrus in Pakistan. Nearly 40 percent of the total area
under citrus (mainly kinnow) was in Sargodha district
with almost half of the total production [PAS 2001-6].
That is why this study was carried out in Sargodha
region, where largest amount of citrus is grown for
the past many decades. In future, the demand for
citrus fruits is expected to grow exponentially
induced by various factors like population growth,
consumer preferences and increase in per capita
income, but the area under citrus cannot chase the
growing demand. So there is requirement heavily
depend on increase in the productivity but the extent
of research, extension and technology transfer
efforts from various institutions to increase its
productivity are negligible. Keeping this view in mind
a study on static and dynamic citrus productivity
determinants and a measure of technical efficiency
was necessary. The study employs new methods
developed in production economics and intends to
make these investigations for citrus productivity in
the selected area.
Data Availability and Variables
Data Sources: The analyses for this study used
primary data which was collected from 162 randomly
selected farmers growing citrus in 27 villages
(sample six from each village) of Sargodha district of
Punjab Pakistan for the years 2005-6 and 2006-7.
The data was collected for two points in time, which
made calculation of Malmquist is Index possible.
Selected Variables used in DEA in the Study:
The variables are:
TIMCODE: It takes value 1 if the observed
inputs/outputs vector belongs to year 2005-6 and
value 2 if the observed inputs/outputs vector belongs
to year 2006-7.
TORA: is fixed input and consists of the total
operational holding area by the farmer to know
whether a larger farmer uses his inputs optimally or
mismanages use to large field under control.
FLDA: is the area under the orchard in the total
operational held area by the farmer.
NPLNT: represents the no. of plants in the orchard
or tree density of the orchard.
DIST: represents the distance between the plants as
single length measure as the tree is planted in
squares shape.
AGGCIRM: represents the aggregate circumference
of logs of the selected plant used as the proxy to
capture the effect of the past care taken of the
orchard. It is different from the variable AGPLT: the
age of the plant having a low value of correlation
Int. J. Agric. Appl. Sci. Vol. 1, No.2, 2009

71
coefficient of 0.60 with
NAGAY: represent the number of missing plants cut
due to some disease, overage or mishap.
AGPLT: represents the age of the age of the plant to
know whether aged plants give more commercial
bearing than that of the younger plants.
NOI: represents the number of irrigations.
NOP: represents the number of ploughings in the
orchard field.
NOL: represents the number of levelings in the
orchard field.
NOH: represents the number of hoeings in the
orchard field.
TOTAL: represents the total of nutrients input of
fertilizers in NPK
YIELD: (KG/ACRE) represent the yield in kilograms
of the fruit per hectare.
The Results of Hypotheses Testing
Summary of descriptive statistics is provided in
Tables 1 to 4. Table-1 showed the descriptive
statistics of major variables used the production
process of citrus. The critical variables effecting yield
are: number of plants; average aggregate
circumference of the main branches of the citrus
plant; fertilizers and its prices, and, number of
missing trees.
Summary Results of Hypotheses (No. 1to12)
Testing.
To know the significance of these factors, a Cobb-
Douglas production function was fitted to the data
using the method of Pooled Least Squares (PLS) by
pooling the data for the year 2005-6 and for the year
2006-7.
ln yield=lnb
0
+b
1
ln(Number of irrigations) +b
2
ln (No.
of plants per field area) +b
3
ln (Total no.of nutrients
inputs in N P K) + b
4
ln(Total Area of the farm) +
b
5
ln(No.of ploughings) +b6ln(Other crop dummy) +b
7
ln (No. of missing plants in the field) +b
8
ln (No.of
leveling) +b
9
ln (No.of hoeings) +b
10
ln (Age of the
plant)
Correlation matrix showed no signs of
multicolinearity. The R
2
was 0.54, and goodness of
fit was fairly high for such cross-sectional pooled
data with adjusted R
2
as 0.53. The number of
irrigations showed negative sign, though non-
significant, was perhaps because of shift in mode of
irrigation, mostly joint irrigation to wheat plus citrus
and Berseem plus Citrus, citrus consuming only 40
percent of the of the water. The number of plants
was significantly affecting the yield that increasing
the number of plants per orchard can increase the
production. The number of missing plants
(LNNAGAY) showed negative sign though not
significant which augmented the evidence that if the
occurrence of missing plants be removed which
results in increase in the number of plants which can
lead to significant increase in production. Distance
between the plants was also highly significant
implying that increasing the distance between the
trees will enhance the yield in the existing setup of
the orchard. The nutrients inputs in N P K have
positive sign but non-significant. The magnitude of
nutrients inputs in N P K was dropped drastically
due to sudden increase in the prices of fertilizers in
those months of the year 2006-7 depriving of its role
as significant contributor to yield. This result was
consistent with earlier studies which found the
contribution of FYM, fertilizer and plant protection
positive and significant but these inputs were also
found underutilized.(Sharif & Burhan, 2005),
Distance between the trees is positively correlated
with other crops and negatively correlated with
number of plants per field area in Table-2 but the
coefficients of distance between the trees and
number of plants per field area is positively
significantly affecting the yield in our estimated
production function. Also, the other crop dummy
coefficient has negative impact on the yield leading
towards a conundrum which needs extra explanation
in global perspective. Some recent experiments in
the developed countries the distance between the
trees inversely affect the yield. The reason is simple
as that when the distances between the trees are
small then there is no other crop within citrus
orchards, which results in larger the citrus yields
(Nawaz et al., 2007).
The farm area was non-significant contributor to
yield but with negative sign confirming the inverse
relationship between size and average productivity
supporting the hypothesis of Sen (1962). Number of
ploughings, Number of Leveling, Number of Hoeings
and Age of the Plant all had positive sign and
significantly affecting the yield barring Number of
Leveling which was not significant. Besides Age of
the Plant, Measurement of Aggregate
Circumference of logs of the selected plant, used as
proxy of past care health and growth of the citrus
tree was highly significant. Another important
variable was Other Crop grown in the orchard field,
which resulted in negative sign as expected and
significant at 10 percent implying that crop
Shahid Iqbal, Maqbool, H. Sial and Zakir Hussain

72
negatively affected the yield.
Summary Results of Hypotheses Testing (#. 13).
In the DEA, the non-optimizing behavior of a firm is
difficult to interpret as inefficiencies may occur due
to specification error, data problems or truly non-
optimizing behavior. The results of input DEA
analysis showed that the overall mean technical
efficiency (T.E.), Variable Returns to Scale (VRS)
mean technical efficiency (T.E.) and scale efficiency
were 63 percent; 97 percent and 64 percent
respectively for the year 2005-6. The output related
VRS mean technical efficiency (T.E.) and scale
efficiency were 80 percent each respectively for the
year 2005-6. In the panel data of the year 2006-7,
the input DEA analysis showed that the overall
mean technical efficiency (T.E.), VRS mean
technical efficiency (T.E.) and scale efficiency were
33 percent 91 percent and 36 percent respectively.
The outputs VRS mean technical efficiency (T.E.)
and scale efficiency were 60 percent and 54 percent
respectively. The result indicated that input related
technical efficiency was dropped to nearly less than
one half than that of the previous year, VRS mean
technical efficiency (T.E.) and scale efficiency were
dropped by 6 percent and 18 percent respectively
from the year 2005-6 to 2006-7. The output oriented
technical efficiency was dropped to nearly less than
one half than that of the previous year, VRS mean
technical efficiency (T.E.) and scale efficiency were
dropped by 20 percent and 26 percent respectively
from the year 2005-6 to 2006-7. This decrease in
relative efficiencies, in 2006-7 than that of previous
year, resulted in a downfall of productivity of the
citrus fruit in the area.
Summary Results of Hypotheses (No. 14)
Testing.
Computer output of summery results of software
DEAP 2.1 is summarized as below in Table 4. Input
and Output oriented Malmquist DEA five indices
were estimated for each firms (farmers) in each
year. All indices were relative to the previous year.
Hence the output begins with year 2. These are
averaged out over all firms (farmers) to summarized
form of results. These are:
Table-4: Input and Output oriented Malmquist DEA
indices in Sargodha, Punjab
Parameters of
efficiency
Input oriented
Malmquist DEA
(%)
Output oriented
Malmquist DEA
(%)
Technical efficiency change 37 33
Technical efficiency change 40 41
Pure technical efficiency change 93 58
Scale efficiency change 31 64
Total factor productivity 15 15
It was evident from these results that the technical
efficiency dropped from 63 percent in 2005-6 to 33
percent in 2006-7. Input oriented scale efficiency
was 64 percent whereas output oriented scale
efficiency was 79.7 percent which were all less than
hundred. All the values for five indices of changes
namely technical efficiency change, technical
efficiency change, pure technical efficiency change,
scale efficiency change and total factor productivity
are less than hundred and are quite low showing
poor technological skills and methods of production
and poor efficiencies in all five categories measured
through Malmquist Index.
Summery and Conclusions
The study mainly concluded that technical efficiency
was about 63 percent in 2005-6 and 33 percent in
year 2006-7. Sargodha citrus growers might have
increased their production by as much as 37
percent in 2005-6 and 63 % in 2006-7 ( or 50
percent on yearly average basis) through more
efficient use of production inputs
The immediate implications are that mostly citrus
growers are small farmers and lack in knowledge or
resources to maintain and manage their citrus crops
while large farmer also have no edge in production.
The reasons for such a low yield included poor
agronomic practices like low use of fertilizers,
excessive tree cuts (due to some horticulture
disease which needs further horticulture research)
resulting in low tree density in the orchards and
possibly shifting mode of irrigation from canal water
to tube well water for the orchard area. The overall
view which is that as permanent source of income,
citrus orchard is less taken care off as regard to the
current crop.
Excessive tree cuts need further investigation of
pest and viral disease phenomenon in the area.
which lead to a poor farmer with no information on
pest related and viral, air borne and soil borne
diseases lead our ignorant citrus growers’ to ultimate
decision to cut the tree which result in persistent low
tree density and loss of productivity
Conventional citrus plantations in Pakistan are
spaced at 22 ft × 22 ft to 20 ft × 22 ft giving about 90
to 100 trees per acre, or 225 to 250 trees per
hectare of low tree density, respectively. In
developed countries, the tendency since 1990 has
changed towards closer planting distances (
“hedgerow” plantings patterns on an 11 ft × 22 ft
spacing that result in a high-density citrus orchard
over 500 trees per hectare) which resulted more fruit
yields per hectare(Nawaz et al., 2007).
Int. J. Agric. Appl. Sci. Vol. 1, No.2, 2009

73
Though coefficient Age of the Plant of in our
production function is significantly affecting the yield
but its contribution in productivity is very low. Poor
yields are not only due to lower tree density but also
due to longer commercial bearing age as compared
to advanced countries. Citrus trees in Pakistan take
up to 9 years to reach commercial bearing, whereas,
in advance countries citrus trees start producing
commercial yield at sixth year. This issue needs
further exploration by agronomy scientists.
In Pakistan there are no linkages between
researcher and growers. Growers do not have crop
management protocols for basic agricultural
practices such as irrigation, fertilization, pruning and
treatment of disease which result in poor productivity
and fruit quality. Excessive tree cuts (no. of nagay =
missing plants) need further investigation of pest
and viral disease phenomenon in the area. which
lead to a poor farmer with no information on pest
related and viral, air borne and soil borne diseases
that lead our ignorant citrus growers’ to ultimate
decision to cut the tree, Which result in persistent
low tree density and loss of productivity.
These developments require pass practical
technological demonstrations extension and transfer
stages and finally adoption by farmers, a big task,
definitely the governments can do. Until the field
demonstrations and extensions research knowledge
from empirical evidence does not shows higher
overall productivity for citrus farmer with high density
with no crops planned in the orchards while when
planting i.e. substituting more citrus tree in place of
other crops, Pakistani citrus farmer will continue to
rely on one third of the average of international citrus
yields and will remain fully unaware of the new
developments that close distance orchard
plantations are more productive. This is important
implication with regard to over all field productivity,
which need be explored on empirical basis allowing
for farmers self fodder and grains requirements.
The coefficient of total area of the farms containing
citrus orchards in our estimated production function
is negatively affecting the yield though not significant
implying the detection of inverse form size
productivity relationship. This result need further
exploration and requires rigorous evidence and
leads to implication of land reforms that will increase
the overall productivity of in the country.
Kinnow is the prime citrus produce exported abroad
and fetch reasonable price due to its special flavor
and exquisite unique taste. Pakistani kinnow has a
great demand in international market but an average
number of 12.2 seeds in kinnow 11.2 in musambi,
9.5 in feutral, and 8.8 in succari per fruit is a big
obstacle for exports to developed countries where
seedless cultivars are preferred (Khan, 2006).
The storage life of kinnow can be extended up to 30
days by using wax technology that favours citrus
exports. Citrus exports of Pakistan are hampered
due to hurdles like, poor logistics/distribution and
transportation system, lack of storage and
processing, expensive production inputs, lack of
information and access to market on international
trade, lack of credit facilities and current marketing
systems which relies upon harvest contractors and
middlemen buying the crop in advance (Mahmood,
2006; Azam, 2006; Sherif et al. 2007). If these
obstacles are removed and exports of kinnow
flourish fetching good price to citrus farmer, the
technical efficiency of the farmer will increase.
Table 1: Descriptive statistics of selected variables in Sargodha Region, Punjab
VARIABLE DESCRIPTION Units and codes AVERAGE STANDARD DEVIATIONS
Crop year TIMCODE 2006 2007 FOR BOTH 2006 2007 FOR BOTH
Farmers (number) COUNT 162.0 162.0 324.0 162.0 162.0 324.0
Total area TORA 17.1 17.5 17.3 16.9 16.7 16.7
No. of plants /kanal NPLFA 10.9 9.7 10.3 2.3 1.8 2.2
Aggregate circumference AGGCIRM 135.9 114.4 125.1 60.0 53.2 57.6
Ploughings (number) NOP 7.4 7.1 7.2 2.1 2.0 2.1
Levelings (number) NOL 2.4 2.9 2.7 1.4 1.1 1.3
Irrigations (number) NOI 8.9 8.1 8.5 3.0 2.5 2.8
Hoeings (number) NOH 1.6 1.5 1.5 0.7 0.8 0.7
NPK total TOTAL 144.3 116.9 130.6 59.2 60.8 61.5
Pest attack dummy DUMPEST 0.7 0.5 0.6 0.4 0.5 0.5
Pest sprays dummy PESTSPR 0.8 0.9 0.8 0.4 0.4 0.4
Pest sprays # NOS 1.4 1.2 1.3 0.9 0.7 0.8
Yield kg/acre YLDkg/ac 10577.3 2463.5 6520.4 6659.1 2774.9 6515.4
Droppage of fruit DROP 26.0 9.9 18.0 31.3 11.8 24.9
No of plants missing in rows NONAGAY 5.0 7.9 6.5 9.5 10.6 8.0
Shahid Iqbal, Maqbool, H. Sial and Zakir Hussain

74
Table 2: Correlations matrix between selected independent variables in Sargodha, Punjab
LNNOI
LNNPLFA LNTOTNPK LNTORA LNNOP LNCRPDM LNNAGAY LNLOL LNNOH LNAGPNT LNAGCRM LNDST
LNNOI 1.00 -0.13 -0.16 0.10 0.25 0.30 0.13 -0.31 0.01 0.50 0.32 0.23
LNNPLFA -0.13 1.00 -0.06 -0.02 -0.42 0.04 -0.06 -0.01 -0.22 0.03 0.17 -0.29
LNTOTNPK -0.16 -0.06 1.00 0.14 -0.04 -0.21 0.04 0.09 0.13 0.17 0.03 0.03
LNTORA 0.10 -0.02 0.14 1.00 -0.12 -0.35 -0.05 -0.09 -0.01 0.27 -0.02 -0.07
LNNOP 0.25 -0.42 -0.04 -0.12 1.00 0.24 0.09 -0.21 0.21 0.00 -0.08 0.08
LNCRPDM 0.30 0.04 -0.21 -0.35 0.24 1.00 0.08 -0.06 -0.31 0.06 0.13 0.17
LNNAGAY 0.13 -0.06 0.04 -0.05 0.09 0.08 1.00 -0.05 -0.13 0.10 0.02 0.20
LNLOL -0.31 -0.01 0.09 -0.09 -0.21 -0.06 -0.05 1.00 -0.06 -0.25 -0.11 0.01
LNNOH 0.01 -0.22 0.13 -0.01 0.21 -0.31 -0.13 -0.06 1.00 -0.03 0.00 -0.01
LNAGPNT 0.50 0.03 0.17 0.27 0.00 0.06 0.10 -0.25 -0.03 1.00 0.60 0.25
LNAGCRM 0.32 0.17 0.03 -0.02 -0.08 0.13 0.02 -0.11 0.00 0.60 1.00 0.24
LNDST 0.23 -0.29 0.03 -0.07 0.08 0.17 0.20 0.01 -0.01 0.25 0.24 1.00
Table3: Regression Results relating Yield of citrus with selected independent variables in Sargodha Region, Punjab
Dependent. Var.: LNYLD = Log(Yield in Kg/Acre ) Method: Pooled L S
Included obs.: 162 Total panel obs. 324
Variable Coeff. S. E. t-Stat. Prob.
C -5.38
***
1.42 -3.80 0.00
LNNOI= Log (Number of irrigations). -0.16 0.13 -1.26 0.21
LNNPLFA= Log (Number of plants per field area). 1.55
***
0.24 6.36 0.00
LNTOTNPK= Log (Total number of nutrients inputs in N P K). 0.001 0.06 0.01 0.99
LNTORA= Log (Total Area of the farm). -0.04 0.04 -1.04 0.30
LNNOP= Log (Number of ploughings). 0.58
***
0.13 4.62 0.00
LNCRPDM= Log (Other crop dummy). -0.02
*
0.01 -1.70 0.09
LNNAGAY= Log (Number of missing plants in the field). -0.01 0.01 -1.13 0.26
LNNOL= Log(Number of leveling) 0.04 0.03 1.40 0.16
LNNOH= Log (Number of hoeings). 0.07
***
0.02 2.85 0.01
LNAGPNT= Log (Age of the plant). 0.31
***
0.12 2.68 0.01
LNAGCRM= Log (aggregate circumference of logs ). 0.97
***
0.10 10.14 0.00
LNDST= Log (Distance between the plants). 2.41
***
0.56 4.27 0.00
R
2
0.54
Adj. R
2
0.53
F-stat. 30.73
S.E. of reg. 0.59
D.-Wat. Stat 1.69
Prob.(F-stat.) 0.00
*** show that the coefficient is significantly different from zero at 0.01 probability level
** show that the coefficient is significantly different from zero at 0.05 probability level
* show that the coefficient is significantly different from zero at 0.10 probability level
REFERENCES
Agnarsson, Sveinn, “Productivity in the Icelandic
Fish Processing Industry 1985-1995: A
omparison of Methods”, Institute of Economic
Studies, University of Iceland, Aragata 14,
101Reykjavik, Iceland, Institute of Economic
Studies, working paper series, W00:14
November 2000.
Angelidis Dimitrios and Lyroudi Katerina (2006),
Efficiency in the Italian Banking Industry: Data
Envelopment Analysis and Neural Networks,
International Research Journal of Finance and
Economics, ISSN 1450-2887 Issue 5 (2006) ©
Euro Journals Publishing, Inc. 2006.
Azam, Haji.”Citrus Growing & Management
Practices, Industry & Supply Chain, Extension
Services Citrus Growing in Pakistan: a Growers
Perspective/Economic Analysis”, Presented in
Workshop on Citrus research, development and
extension in Pakistan and Australia: Islamabad,
Pakistan, 24
th
, 25
th
July, 2006.
Banker, R.D., Charnes, A. and Cooper, W.W. “Some
Models for Estimating Technical and Scale
Inefficiencies in Data Envelopment Analysis”,
Management Science, 30, 1078-1092, 1984.
Camanho A. S. and Dyson R. G., “Data
envelopment analysis and Malmquist indices for
measuring group performance”, J Prod Anal
(2006) 26:35–49 DOI 10.1007/s11123-006-
0004-8©Springer LLC 2006.
Chapelle Karine and Patrick Plane, “Productive
Efficiency in the Ivorian Manufacturing
Sector:An Exploratory Study Using a Data
Envelopment Analysis Approach”, the
Developing Economies, XLIII-4 : 450–71
December, 2005.
Charnes, A., W.W. Cooper and E. Rhodes,
“Measuring the Efficiency of Decision Making
Units”, European Journal of Operations
Research, 2, 429-444, 1978.
Int. J. Agric. Appl. Sci. Vol. 1, No.2, 2009

75
Cinca, S., Molinero C. M., and García F. C., “behind
DEA efficiency in financial institutions”,
Discussion Papers in Accounting and Finance,
University of Southampton, UK Number AF02-
7March 2002ISSN1356-3548, 2002.
Coelli T.J., “A Guide to DEAP Version 2.1: A Data
Envelopment Analysis (Computer)”, CEPA
Working Paper 96/08, Department of
Econometrics, University of New England,
Armidale, NSW 2351, Australia, http://www.
une.edu.au/econometrics/cepa/wp.htm,1996.
Coelli, T.J., D.S. P. Rao, C.J. O’Donnell and G.E.
Battese (2005). “An introduction to Efficiency
and productivity analysis.” Springer, USA.
Coelli, Tim, S. Prasada Rao, “Total Factor
Productivity Growth in Agriculture: A Malmquist
Index Analysis of 93 Countries,1980-2000”,
Centre for Efficiency and Productivity Analysis
School of Economics University of Queensland,
Working Paper Series No. 02/2003.
Dhehibi, B., Lassad Lachaal, Mohammed Elloumi
and Emna B. Messaoud, “Measurement and
Sources of Technical Inefficiency in the
Tunisian Citrus Growing Sector,” Paper
prepared for presentation at the
Mediterranean Conference of Agro-Food Social
Scientists. 103
rd
EAAE Seminar ‘Adding Value
to the Agro-Food Supply Chain in the Future
Euro-Mediterranean Space’. Barcelona, Spain,
April 23
rd
- 25
th
, 2007.
Domah Preetum, “Technical efficiency in electricity
generation - the impact of smallness and
isolation of island economies”, Massachusetts
Institute of Technology Center for Energy and
Environmental Policy Research CMI Working
Paper 14, DAE WP 0232, Queens’ College,
Department of Applied Economics University of
Cambridge, UK,2002.
FAO “Report on fruit production in Pakistan”, Food
and Agriculture Organization, United Nation
Publications, July 1998.
Iqbal, Muhammad. “Marketing/Supply
Chain/Agribusiness Issues Regarding Citrus in
Pakistan”, Presented in Workshop on Citrus
research, development and extension in
Pakistan and Australia: Islamabad, Pakistan,
24
th
, 25
th
July, 2006.
Khan,Sardar Riaz A. “Citrus quality to meet global
emand”, http://www.pakissan.com/english/index.
shtml, 2006.
Kuosmanen Timo and Mika Kortelainen, “Stochastic
Nonparametric Envelopment of Data: Cross-
sectional frontier estimation subject to shape
constraints”, Working Paper Keskustelualoitteita
#46 Joensuun yliopisto, Taloustieteet, 2007.
Kuosmanen, T. and G.T. Pos., “Shadow Price
Approach to Productivity Measurement: A
Modified Malmquist Index”, Working Paper,
Erasmus University, P.O. Box 1738 3000 DR
Rotterdam, TheNetherlands, 2000.
Mahmood A. M. and A. D. Sheikh. “Citrus Export
System in Pakistan”,J. Agric. Res., 44(3),2006.
Nawaz Azher M. , Waqar Ahmed, &M. Mithal
Jiskani,“High density planting-an approach to
increase citrus yields” http://www.pakissan.com/
english/index.shtml, 2007.
Neto L. B., Marcos Pereira Estellita Lins., Eliane
Gonçalves Gomes, João Carlos Correia
Baptista Soares de Mello, Fabiano S. Oliveira,
“Integrating the regulatory and utility firm
perspectives, when measuring the
efficiency of electricity distribution”, European
Journal of Operational Research xxx (2006)
xxx–xxx www.elsevier.com/locate/ejor.
Pakistan Agricultural Statistics 2001,4,5,6, Statistics
division govt. of Pakistan. 2001,4,5,6.
Saleem, M., “Technical Efficiency in Electricity
Generation Sector of Pakistan The impact of
Private and Public Ownership A Stochastic
Production Frontier Analysis”, Asia Pacific
School of Economics and Governments, The
Australian National University, Canberra,
Australia, 2005.
Sen A.K., “An Aspect of Indian Agriculture”, the
Economic Weekly, Annual No. February, 1962).
Sharif M. & Waqar Akhtar. “Fruit Exports Prospects
in Pakistan” AR2005.DOC, www.parc.gov.pk/
data/SSD-Database/.2005.
Sharif, M. & Burhan Ahmad, “Marketing of Fruits
Under WTO Regime” Published in "The News"
Rawalpindi / Islamabad, Lahore and Karachi
August 29, 2005.
Sharif, M., Usman Mustafa & Umar Farooq, “
Managing Fresh Fruits Trade under WTO
Scenario: A Case Study of Citrus Export from
Pakistan” http://www.pakissan.com/english/index.
shtml,2007.
Warraich, Mohammad Mohsin Ch., “Orchard
Contracting System, Is There a Possibility for
Shift?” Presented in Workshop on Citrus
research, development and extension in
Pakistan and Australia: Islamabad, Pakistan,
2006.
Shahid Iqbal, Maqbool, H. Sial and Zakir Hussain