SOCIAL SCIENCES & HUMANITIES Assessing Determinants of Technical Efficiency in Livestock Production: A Case Study from Shaanxi, China

Abstract

The demand for livestock products is rising, and China is actively encouraging farmers to increase their livestock production to meet this growing demand. At Shaanxi Province's livestock industry's current production output and growth rate, it appears unfeasible to meet the government's production target for 2025. Inefficiencies within livestock production can significantly impede the development of this industry. Therefore, this research employs the Data Envelopment Analysis (DEA) technique, considering Constant Returns to Scale (CRS) and Variable Returns to Scale (VRS) assumptions, to assess the technical efficiency of the livestock industry in Shaanxi Province. The data utilised are secondary data from 2010 to 2019. The findings reveal that the Shaanxi livestock industry has an average technical efficiency of 0.84 (CRS) and 0.92 (VRS), suggesting that there is room for further production growth with the current inputs, breeding scales and technology. Although dairy cows, cattle and goats have achieved full technical efficiency. Technical and scale inefficiencies still exist in hog and layer farming practices, which can be improved to increase production. Notably, hog farming demonstrated the lowest technical efficiency, scoring 0.68. The results of factors affecting inefficiency suggest that increasing spending on disease prevention and raising the selling price can both improve technical efficiency. Additionally, reducing death loss has the potential to improve technical efficiency. Thus, the government is expected to promote farm consolidation and expansion while actively advocating for establishing livestock production cooperatives.

Full text

Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Journal homepage: http://www.pertanika.upm.edu.my/
© Universiti Putra Malaysia Press
SOCIAL SCIENCES & HUMANITIES
Article history:
Received: 08 July 2023
Accepted: 18 March 2024
Published: 27 September 2024
ARTICLE INFO
DOI: https://doi.org/10.47836/pjssh.32.3.17
E-mail addresses:
li_yujie@student.usm.my (Yujie Li)
eslim@usm.my (Ee Shiang Lim)
limgheethean@usm.my (Ghee-Thean Lim)
* Corresponding author
ISSN: 0128-7702
e-ISSN: 2231-8534
Assessing Determinants of Technical Eciency in Livestock
Production: A Case Study from Shaanxi, China
Yujie Li, Ee Shiang Lim and Ghee-Thean Lim*
School of Social Sciences, Universiti Sains Malaysia, 11800 Gelugor, Penang, Malaysia
ABSTRACT
The demand for livestock products is rising, and China is actively encouraging farmers to
increase their livestock production to meet this growing demand. At Shaanxi Province’s
livestock industry’s current production output and growth rate, it appears unfeasible to meet
the government’s production target for 2025. Ineciencies within livestock production
can significantly impede the development of this industry. Therefore, this research
employs the Data Envelopment Analysis (DEA) technique, considering Constant Returns
to Scale (CRS) and Variable Returns to Scale (VRS) assumptions, to assess the technical
eciency of the livestock industry in Shaanxi Province. The data utilised are secondary
data from 2010 to 2019. The ndings reveal that the Shaanxi livestock industry has an
average technical eciency of 0.84 (CRS) and 0.92 (VRS), suggesting that there is room
for further production growth with the current inputs, breeding scales and technology.
Although dairy cows, cattle and goats have achieved full technical eciency. Technical and
scale ineciencies still exist in hog and layer farming practices, which can be improved
to increase production. Notably, hog farming demonstrated the lowest technical eciency,
scoring 0.68. The results of factors aecting ineciency suggest that increasing spending
on disease prevention and raising the selling price can both improve technical eciency.
Additionally, reducing death loss has the
potential to improve technical eciency.
Thus, the government is expected to
promote farm consolidation and expansion
while actively advocating for establishing
livestock production cooperatives.
Keywords: Data Envelopment Analysis (DEA),
ineciency, livestock industry, technical eciency,
Tobit regression

1166 Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Yujie Li, Ee Shiang Lim and Ghee-Thean Lim
INTRODUCTION
The global demand and production of
livestock products are increasing rapidly,
especially in China, due to population
growth, rising income, and changes in
lifestyle and dietary habits (Food and
Agriculture Organization, 2021). In 2022,
the National Bureau of Statistics of China
reported that China’s livestock industry
produced 52.959 million tons of pork,
6.975 million tons of beef, 5.141 million
tons of mutton, 24.825 million tons of
poultry, 34.088 million tons of eggs, and
36.827 million tons of milk (National
Bureau of Statistics of China, 2023).
China is a major producer of livestock
products, ranking first in the world in
producing pork, mutton, broiler and
eggs and third in dairy production (Food
and Agriculture Organization, 2022a).
Despite the increasing annual production
of livestock products, China remains the
world’s top importer of livestock products
(Food and Agriculture Organization,
2022b). Empirical research shows that
China’s current pork, beef, and mutton
production is inadequate to meet domestic
demand (Shi et al., 2015). In 2020, the
proportion of imported meat in China’s
total meat production reached 12.7%,
equivalent to 9.91 million tons, indicating
a signicant increase over the previous ve
years (AskCI Consulting, 2022). Amid the
COVID-19 pandemic, China intensied
restrictions on imported food products
after detecting the virus in frozen food
(Cadell, 2020). It has caused challenges
not only for food suppliers and supply
chains but also for China’s dependence
on imported meat and dairy products. The
China Government’s 14th Five-Year Plan
(2021–2025) includes a target to increase
livestock production, with an expected
growth of 15% in meat production (Patton,
2022; Shaanxi Provincial Department of
Agriculture, 2022). If all producers operate
at full technical eciency, the production
of the livestock industry would more easily
align with the output targets expected by
the government, consequently reducing the
demand for imported products.
The livestock industry plays a signicant
role in Shaanxi. Because it contributes to
80% of China’s goat milk production, and
its egg production is crucial to meeting the
demand of surrounding provinces (National
Bureau of Statistics of China, 2021).
Moreover, pork and milk provide the most
protein for Shaanxi residents. However,
this industry confronts obstacles such as
outbreaks of diseases, including African
swine fever for hog (Wang, Zhao, et al.,
2021) and avian influenza for layer. All
these obstacles may reduce the production
enthusiasm of farmers and lead to a scaling
down of farming operations.
As of 2021, Shaanxi’s livestock industry
ranked 20th out of 31 provinces in mainland
China, contributing 21.3% of the agricultural
production value. The province produced
1.274 million tons of meat, 634,000 tons
of eggs, and 1.619 million tons of dairy
products. The most recent “14th Five-
Year Plan for Livestock and Veterinary
Development in Shaanxi” (Shaanxi
Provincial Department of Agriculture,

1167Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Assessing Technical Efficiency in Livestock Production
2022) anticipated a rise in production, with
an expected production of 1.8 million tons of
meat, 0.8 million tons of eggs and 3 million
tons of dairy products by 2025. Maintaining
the current breeding practices will make
it challenging for the livestock industry
to achieve this government objective in
production.
The output of the livestock industry
in Shaanxi increased annually, but the
growth rate has shown a decreasing trend
year on year (Figure 1). In comparison to
the national average growth rate, Shaanxi
exhibits a lower growth rate (National
Bureau of Statistics of China, 2021). It is
possibly due to technical inefficiencies,
which consequently impact the production
of Shaanxi’s livestock industry (Terry et
al., 2021). Thus, it is important to know
whether ineciency exists in the Shaanxi
livestock industry and identify the factors
aecting it. Such an understanding would
enable producers in the livestock industry to
take necessary steps to improve eciency
and enhance the overall performance of the
livestock industry.
Many studies in the eld of livestock
do not distinguish between animal species
or focus solely on specic animal species
(Kuhn et al., 2020; Wang, Han, et al., 2021;
Zhou et al., 2015), whereas this study adopts
a more comprehensive approach. This
study not only individually examines the
primary livestock species within Shaanxi’s
livestock industry but also conducts a
comparative analysis of technical eciency
across different scales of production.
Moreover, a signicant portion of research
on factors aecting ineciency focuses on
emphasising the factors related to livestock
farmers, such as farmers’ gender or their
experience in the livestock industry (Tian
et al., 2015; Wang, Han, et al., 2021). In
contrast, this paper extends the analysis
to explore how medical and epidemic
prevention, death loss, and selling prices
aect ineciency in the Shaanxi livestock
industry.
Figure 1. Shaanxi livestock growth rate
Source: National Bureau of Statistics of China (2021)
-0.1
0
0.1
0.2
0.3
0.4
0.5
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
Shaanxi livestock growth rate
Trend of growth rate

1168 Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Yujie Li, Ee Shiang Lim and Ghee-Thean Lim
In light of the abovementioned
considerations, this study seeks to accomplish
two objectives: (1) to evaluate the level of
technical efficiency and (2) to identify
factors that affect technical inefficiency
in the Shaanxi livestock industry. The
ndings of this study will provide valuable
insights for policymakers on how to increase
livestock production and modify relevant
policies to promote sustainable development
in the industry.
LITERATURE REVIEW
Production efficiency refers to a firm’s
ability to achieve the maximum output
with a given set of inputs and contemporary
technology (Farrell, 1957). However,
efficiency cannot be directly observed.
Therefore, appropriate methods are needed
to measure it. There are two main techniques
for measuring technical efficiency: non-
parametric Data Envelopment Analysis
(DEA) and parametric Stochastic Frontier
Analysis (SFA). The choice of technique can
inuence the technical eciency results, and
there is no consensus on which technique is
most appropriate for agricultural technical
efficiency (Heshmati et al., 1995). DEA
cannot require a specic functional form to
be imposed on the data and can easily be
adapted to multiple outputs. Additionally,
DEA is deterministic and attributes all
deviations from the frontier to ineciency,
making it sensitive to measurement errors
and other statistical noise in the data.
Unlike SFA, DEA is more inclusive of
small samples (Zhu, 2009). SFA represents
a parametric approach, also known as an
econometric approach, which involves
tting an assumed structure of the observed
data (Aigner et al., 1977; Meeusen & van
Den Broeck, 1977). The main advantage of
SFA is its ability to handle random noise.
However, SFA requires a specic functional
form to be imposed on the underlying
technology and a distributional assumption
to be imposed on the inefficiency term.
Upon comparison, the DEA method has
been selected as the preferred technique for
measuring eciency in this study.
DEA is a non-parametric technique to
measure the eciency of rms by comparing
their production set with a production
frontier. Based on the fundamentals of
efficiency, DEA was developed into two
assumptions: CCR (Charnes, Cooper and
Rhodes) and BCC (Banker, Charnes and
Cooper). The CCR assumption is also
known as the CRS (constant returns to scale)
assumption. Assuming that the Decision-
Making Units (DMUs) are operating using
the given inputs and technology, the results
generated by DEA would indicate the
technical efficiency (TE) score. When
the score equals 1, the DMUs function
eciently and operate at optimal production
levels. Conversely, a score lower than 1
suggests that the DMUs are inefficient
(Zhang et al., 2017). Banker et al. (1984)
introduced the BBC assumption (also
known as VRS, variable returns to scale) as
an improvement over the CRS assumption,
which envelops the data points more
tightly than the CRS assumption. Technical
eciency (TE) is the result of CRS DEA,
while pure technical eciency (PTE) is the

1169Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Assessing Technical Efficiency in Livestock Production
result of VRS DEA. The dierence between
these two is the scale eciency (SE), which
represents the ratio of the actual output of
a DMU to its optimal output at the ecient
scale of operation. If there is a dierence
between the TE and PTE scores, it indicates
that the rm is operating at a suboptimal
size. Scale eciency equals one only when
the scores of TE and PTE are equal (Färe &
Lovell, 1978).
Eciency is widely used in the livestock
industry, with DEA commonly employed.
Such studies have identied ineciencies
in several countries in the dairy industry,
including Australia (Fraser & Cordina,
1999), Sweden (Hansson & Öhlmér, 2008),
and Estonia (Luik-Lindsaar et al., 2019).
Some studies have utilised both CRS DEA
and VRS DEA for efficiency analysis.
Notably, the VRS efficiency scores are
relatively higher than the CRS eciency
scores, as VRS can capture the eciency
from scale benets. For example, previous
studies have found that in Hawaii’s pig
farming industry, the VRS eciency score
was 0.726, whereas the CRS score was
0.644 (Sharma et al., 1997). Similarly,
Lansink and Reinhard (2004) reported
a VRS score of 0.90 and a CRS score
of 0.89 in the Netherlands. Mugera and
Featherstone (2008) found VRS and CRS
scores of 0.41 and 0.33 in the Philippines.
Besides, Galluzzo (2019) investigated dairy
farms in Iceland and reported a CRS score
of 0.881 and a VRS score of 0.946. İlkikat
et al. (2020) studied hair goat farms in
Turkey and found a CRS score of 0.67 and
a VRS score of 0.76. These ndings suggest
that the DEA technique can be applied to
various livestock categories, regardless of
geographic location. Also, scale ineciency
is prevalent in most livestock farming,
making it important to analyse CRS and
VRS to enhance farming scale.
Most previous studies on efficiency
in the livestock industry in China have
mostly concentrated on the hog industry
(Kuhn et al., 2020; Somwaru et al., 2003;
Tian et al., 2015; Wang, Zhao, et al., 2021;
Yang et al., 2008; Zhou et al., 2015). For
instance, Yang et al. (2008) surveyed 39 hog
farmers to assess their technical eciency
in Taiwan between 2003 and 2004. The
results demonstrated that farms could
increase their output by an average of 52.8%
while maintaining the same input levels.
In an earlier study, Somwaru et al. (2003),
who used a non-parametric technique,
discovered that Shaanxi experienced
technical inefficiency in the livestock
industry with technical eciency scores of
0.75, higher than the national average of
0.24 in 1996. Similarly, several studies, such
as Tian et al. (2015), Zhou et al. (2015), and
Wang, Zhao, et al. (2021) used parametric
methods to examine the eciency of the
hog industry in China and found it to be
inecient. Moreover, these studies revealed
that the technical eciency of the northern
provinces, including Shaanxi, is lower than
that of other provinces.
The variance in technical efficiency
among farms can be utilised to discover
the factors that affect inefficiency. The
standard approach entails conducting a
regression analysis of efficiency scores

1170 Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Yujie Li, Ee Shiang Lim and Ghee-Thean Lim
against a series of explanatory variables
(Lansink & Reinhard, 2004). In economics,
determining the impact of exogenous
factors on production involves converting
the technical eciency score into technical
ineciency, which is obtained by subtracting
the technical eciency score from 1 (Coelli
et al., 2005; Farrell, 1957). Then, factors
affecting inefficiency can be identified
through Tobit regression analysis applied to
truncated data (Dogan et al., 2018; Liu et al.,
2021; Zhang et al., 2017). Several studies
examining efficiency in the hog farming
industry across dierent countries, including
China (Tian et al., 2015), the Philippines
(Mugera & Featherstone, 2008), and Turkey
(İkikat et al., 2020), found that a higher
level of education among farmers leads to
increased eciency in hog farming, likely
due to improved knowledge of scientic
breeding techniques. Similarly, Wang, Han,
et al. (2021) conducted a survey involving
449 herders within the Inner Mongolia
grassland area of China in 2017. Their
ndings revealed that the existing policies
have facilitated the expansion of livestock
farming scales, leading to the departure
of inefficient farmers from the industry.
Additionally, Jo et al. (2021) surveyed farms
in Heilongjiang Province, determining that
a reduction in death losses contributes to
enhanced technical efficiency within the
livestock industry.
This study’s selection of factors
aecting ineciency draws upon theoretical
frameworks and previous empirical
investigations. Within the Keynesian theory,
governmental financial aid to farmers
for epidemic prevention is perceived to
safeguard their operations and stimulate
heightened production. It aligns with the
objectives outlined in the 10th to the 14th
Shaanxi Five-Year Development Plans
(spanning from 2001 to 2025) within
the livestock industry, which accentuate
the importance of bolstering epidemic
control measures to enable farmers to
achieve enhanced incomes (Crop Farming
Management Oce of Shaanxi Province,
2022; Shaanxi Government, 2008,
2022; Shaanxi Provincial Department
of Agriculture, 2018; Shaanxi Statistics
Bureau, 2015). Moreover, according to
risk management theory, death loss has
adverse impacts on production outcomes
by introducing uncertainty and potential
disruptions to agricultural operations,
thereby diminishing efficiency. The
law of supply implies that as prices of
agricultural products escalate, farmers
exhibit a heightened inclination to expand
production and augment supply to the
market. Conversely, lower prices might lead
to diminished production or market exit,
resulting in reduced quantities supplied.
METHODOLOGY
Variable Selection and Description
To obtain precise results of technical
efficiency, the number of input variables
should be neither too few nor too many
(Reinhard et al., 1999). Based on livestock
features and plenty of readings, such as
Sharma et al. (1997), Fraser and Cordina
(1999), Galluzzo (2019) and Soh et al. (2021),
this study selected one output variable, three

1171Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Assessing Technical Efficiency in Livestock Production
input variables and three factors that are
likely to have an impact on ineciency. The
description of the variables is as follows:
(1) Output: The output value
measurement is based on each animal’s
farm price. In the case of hogs, cattle, and
goats, the primary output is determined
by the live weight price of each animal
at the time of sale. In the case of layer,
the output value is the farm price of total
eggs from 100 layers.
(2) Labour: In livestock production,
labour input is quantied as the number
of employees’ working days required
per animal. However, when evaluating
layers specically, the measurement unit
employed is the working days required
for the management of 100 layers.
These working days are calculated
based on 8 hours per day. For instance,
if one goat needs six workdays of
labour, an employee would spend 48
hours on one goat.
(3) Young: This variable is measured
as the average purchase price of each
young animal. The expenses of young
animals vary across dierent livestock
categories. The expenses for cattle, hogs,
and goats consist of the acquisition cost
of each young animal. While the layer
is small livestock, the expenses are the
purchase price of 100 chicks.
(4) Feed: The weight of feed is the total
weight of grains, beans, fodder and
additives consumed by each animal or
100 layers.
(5) MEP: Medical and epidemic
prevention (MEP) expenses include
immunising livestock, preventing
epidemics, testing, quarantine,
eradicating infectious diseases and
government-enforced controlling
measures. MEP measures the individual
expenditures for each animal, while
for layers, it quantifies the expenses
of 100 layers. As it protects farmers’
output, this variable is expected to
have a negative impact on technical
ineciency.
(6) DL: Livestock death loss (DL)
occurs when livestock dies due to
various causes, such as disease, disaster,
nutritional deciencies and inadequate
management. Calculating the death
loss per animal involves dividing the
number of animal deaths by the initial
number of animals and multiplying the
result by the farm price per animal.
The loss of livestock can lead to
reduced production, increased costs
and decreased protability for farmers.
(7) SP: The selling price (SP) represents
the average selling price at which 50
kg of livestock products are sold in
the Shaanxi agricultural wholesale
market. This factor is expected to have a
negative eect on technical ineciency
since high selling prices can encourage
farmers to increase their breeding
activities.
Data
A two-stage analysis was used in this study
to examine the technical eciency of the
ten categories of animals in the Shaanxi
livestock industry (Liu et al., 2021). In the

1172 Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Yujie Li, Ee Shiang Lim and Ghee-Thean Lim
first stage, Data Envelopment Analysis
(DEA) is employed to calculate the technical
efficiency score of each category in the
Shaanxi livestock industry. In the second
stage, Tobit regression is used to estimate
the factors that aect ineciency in each
category. Variables such as inputs (including
labour, young animals and feed), output, and
factors aecting ineciency (medical and
epidemic prevention, death loss and selling
price) are selected. The data utilised in this
study constitute secondary data sourced
from the China Agricultural Product Cost‒
Benefit Compilation (Price Department
of the National Development and Reform
Commission & Price Cost Research
Centre of the National Development and
Reform Commission, 2020). This dataset,
characterised as panel data, encompasses
ten categories of animals within the Shaanxi
livestock industry from 2010 to 2019. A
linear interpolation method was applied to
generate missing data points to minimise the
impact of missing data. This study’s price
data is anchored to 2010, with adjustments
made using the Shaanxi Livestock Consumer
Price Index (CPI) or the Shaanxi Young
Animal CPI. The CPI data are sourced from
the Shaanxi Statistical Yearbook. The data
source also classifies livestock into four
categories based on their breeding scale:
backyard, small-, medium- and large-scale.
Backyard farming refers to a practice that
involves individual households or small
farmers raising livestock in a small pen or
backyard. This study utilises ten categories
of DMUs: backyard and medium-scale
dairy cows, backyard, small-, medium-, and
large-scale hogs, small- and medium-scale
layers, and backyard goats and cattle. The
classication standard of the breeding scales
for each category is shown in Table 1.
Models
Essentially, the DEA technique is built
on the technical assumptions of Constant
Returns to Scale (CRS), which assumes that
the DMUs are operating at an optimal scale.
Banker et al. (1984) extended this method
to incorporate technologies with Variable
Returns to Scale (VRS), which would
envelop the data points more tightly than
the CRS assumption. Prior to computing
technical eciency, it is essential to select
the orientation of minimising the inputs
or maximising the outputs based on the
variable (input or output) that the manager
needs to control the most. Substantively,
Table 1
Standard classication of breeding scale based on quantity
Livestock Backyard Small-scale Medium-scale Large-scale
Dairy Cow Q≤30 -50<Q≤500 -
Hog Q≤30 30<Q≤100 100<Q≤1000 Q>1000
Layer - 300<Q≤1000 1000<Q≤10000
Goat Q≤100 ---
Cattle Q≤50 ---
Notes. Q = Quantity, signifying the number of animals bred on the farms
Source: Price Department of the National Development and Reform Commission (2019)

1173Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Assessing Technical Efficiency in Livestock Production
the Shaanxi Provincial Government has
been encouraging the enhancement of
livestock production. Thus, this study adopts
an output-oriented approach. Following
Coelli et al. (2005), the output-oriented
VRS assumption in Equation 1 and the CRS
assumption in Equation 2 for measuring
technical eciency are given:

,
,
subject to
−

+ ≥ 0, (1)


−  ≥ 0,

′×1
= 1
 ≥ 0,
subject to

,
,
subject to
−

+ ≥ 0, (1)


−  ≥ 0,

′×1
= 1
 ≥ 0,
(1)

,
,
subject to
−

+ ≥ 0, (2)


−  ≥ 0,
 ≥ 0,
subject to
, ,
subject to
−+ ≥ 0, (2)
−  ≥ 0,
 ≥ 0,
(2)
where

× 1
1≤  ≤ ∞
1



()
′×1= 1
represents a

× 1
1≤  ≤ ∞
1



()
′×1= 1
vector of
constants (weights). The parameter

× 1
1≤  ≤ ∞
1



(

)

′×1
= 1
signifies the ratio between the distance
from the optimum point to the origin of the
coordinate axes and the distance from the
observed point to the origin of the coordinate
axes. This ratio is equivalent to calculating
the inverse of technical eciency, subject to
the constraint that

× 1
1≤  ≤ ∞
1



(

)

′×1
= 1
. The vectors
X and Q are the observed inputs and outputs,
respectively. The value of

× 1
1≤  ≤ ∞
1



(

)

′×1
= 1
refers to the
level of technical eciency (TE) ranging
from 0 to 1 of the i-th decision-making unit,
with a value of 1 indicating that the farm is
technically ecient and on the production
frontier (Farrell, 1957). The VRS DEA
assumption involves three constraints. In
the first constraint, the observed output
(qi) of the i-th farm is multiplied by

× 1
1≤ ≤ ∞
1



(

)

′×1
= 1
and
compared to the maximum output vector of
the theoretically ecient farm (

× 1
1≤ ≤ ∞
1



(

)

′×1
= 1

× 1
1≤  ≤ ∞
1



()
′×1= 1
). With
the same quantity of inputs, the constraint
indicates that the theoretically efficient
farm produces more or the same volume
output than the actual output produced
by the i-th farm. The second constraint
illustrates that the observed input (xi) in the
i-th farm is more than or equal to the input
(Xλ) of the theoretically ecient farm. The
third constraint of

′×1
= 1
signies the
inefficiency of a farm evaluated against
other farms of similar size. Such constraint
enables the evaluation of farm eciency
in terms of technical and scale eciencies
(Mohd Idris et al., 2013).
If the technical eciency score turns out
to be

× 1
1≤  ≤ ∞
1



(

)

′×1
= 1
= 1, this farm is technically ecient.
Then, the output of this farm is as much as
the production of the technically ecient
farm using the same volume of inputs. If
the eciency score turns out to be

× 1
1≤  ≤ ∞
1



(

)

′×1
= 1
< 1,
the farm is technically inecient. It means
that the farm’s output can be increased
to the level of

× 1
1≤  ≤ ∞
1



(

)

′×1
= 1

× 1
1≤  ≤ ∞
1



()
′×1= 1
. Notably, the linear
programming problem needs to be solved I
times to obtain a value for each sample farm.
Hence, a value of

× 1
1≤  ≤ ∞
1



(

)

′×1
= 1
is calculated for each
farm. As shown in Equation 2, the VRS DEA
assumption can be transformed into the CRS
DEA assumption by removing the constraint
of

′×1
= 1
(Siafakas et al., 2019).

1174 Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Yujie Li, Ee Shiang Lim and Ghee-Thean Lim
Scale efficiency refers to the extent
to which a DMU is operating at the most
productive scale size. If a DMU is operating
below its optimal scale size, it may be able
to increase its eciency by adjusting its
practice scale. Dividing the CRS eciency
score by the VRS eciency score allows
the capture of the impact of the scale eect.
Scale eciency can be expressed as:
 =



(3)
After obtaining the technical eciency
scores through DEA in the first stage,
the efficiency scores are converted into
inefficiency scores. Subsequently, a
regression analysis is conducted to examine
the relationship between ineciency scores
and other exogenous variables in the second
stage. The technical ineciency scores are
derived by subtracting the TECRS or TEVRS
scores obtained from the rst stage from 1
(Coelli et al., 2005; Farrell, 1957). The Tobit
regression handles the truncated ineciency
estimates at 0 and 1 (Greene, 1994). The
explanatory variables in the regression
model, such as medical and epidemic
prevention, death loss and selling price,
reect the factors that aect ineciency (the
explained variable). The following equation
expresses the Tobit regression:

,
=
0
+
1

,
+
2

,
+
3

,
+
,

,
=
0
+
1

,
+
2

,
+
3

,
+
,
(4)
Where Ineffi,t refers to the technical
inefficiency score of each category of
livestock (i-th) ranging between 0 and 1 for t
periods. It is obtained from the reciprocal of
results of the CRS or VRS DEA assumption
subtracted by one; β0 is the intercept; β1 to
β3 are coecients estimated for individual
independent variables; MEPi,t denotes
medical and epidemic prevention expenses;
DLi,t indicates livestock death loss during
farming; SPi,t is the average selling price in
Shaanxi wholesale markets; and εi,t refers
to the error term.
RESULTS AND DISCUSSION
Technical Eciency
Figure 2 indicates that there have been
fluctuations and downward trends in the
technical eciency scores of the Shaanxi
livestock industry in recent years. The
average technical eciency scores over the
decade reached 0.84 (CRS) and 0.92 (VRS);
the scale efficiency score was 0.91. The
technical eciency scores under the VRS
assumption, which range from 0.70 to 1, are
higher than those under the CRS assumption
(0.54–1). Additional details regarding these
ndings are presented in Table 2. Geometric
means are calculated for each category to
compare the technical eciencies across
various categories between 2010 and
2019. Dairy cow, goat, and cattle farms
are operating at full technical efficiency
(1.00) during the study period. However,
ineciencies are mainly observed in hog
and layer farming practices, indicating that
enhancing the performance of these two
species could lead to an overall improvement
in the technical eciency of the livestock
industry in Shaanxi.

1175Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Assessing Technical Efficiency in Livestock Production
Figure 2. Technical eciency score in the Shaanxi livestock industry
Source: Authors’ work
0.7
0.75
0.8
0.85
0.9
0.95
1
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
VRS efficiency score
CRS efficiency score
Table 2
Technical eciency and ineciency in the Shaanxi livestock industry from 2010 to 2019
Products Technical Ineciency SE Technical Eciency
VRS CRS VRS CRS
Dairy Cow 1
Backyard 0 0 1 1 1
Medium-scale 0 0 1 1 1
Hog 0.68
Backyard 0.30 0.46 0.77 0.70 0.54
Small-scale 0.17 0.41 0.71 0.83 0.59
Medium-scale 0.13 0.32 0.78 0.87 0.68
Large-scale 0 0.04 0.96 1 0.96
Layer 0.89
Small-scale 0.17 0.21 0.96 0.83 0.79
Medium-scale 0 0 1 1 1
Goat 00111
Cattle 00111
Mean for all livestock 0.08 0.16 0.91 0.92 0.84
Notes. CRS: Constant Returns to Scale, SE: Scale Eciency, VRS: Variable Returns to Scale
Source: Authors’ work
From Table 2, it is evident that within
the Shaanxi livestock industry, hog farms
demonstrate the lowest eciency scores
(0.68), lower than those of cows (1.00), goats
(1.00), and layers (0.89). Moreover, across
all scales of hog farming and small-scale
layer farming, the technical eciency score
obtained via VRS analysis is consistently
higher than that obtained through CRS
analysis, indicating the potential for

1176 Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Yujie Li, Ee Shiang Lim and Ghee-Thean Lim
efficiency improvement through optimal
scale management. Additionally, the results
of the CRS and VRS assumptions provide
evidence that the technical eciency score
increases with an expansion in the breeding
scale.
In the case of layer, small-scale breeding
demonstrates a technical eciency score of
0.79, indicating a potential 21% increase in
output. Meanwhile, medium-scale breeding
operates at full technical efficiency. The
ndings are consistent with Zhong et al.
(2021) and conrm the suitability of the
government’s ve-year plan for livestock,
which promotes farmers’ increasing
breeding scales in the layer sector.
The technical eciency score of a hog
(0.68) in Shaanxi is higher than the 0.58
reported by Tian et al. (2015) in Shaanxi
and 11 other provinces but lower than the
0.75 measured by Somwaru et al. (2003) and
0.84 measured by Zhou et al. (2015). These
ndings are inconsistent with prior research
and possibly attributable to divergent data
sources where backyard farming is relatively
more prominent in the surveyed data.
In Shaanxi, there has been a signicant
decrease in the proportion of households
involved in backyard hog farming, which
declined from 89% in 2010 to 43% in 2020
(National Bureau of Statistics of China,
2010, 2021). This shift has resulted in a
corresponding increase in concentrated hog
farming, encompassing small-, medium-,
and large-scale farms, which has risen
from 11% to 57% over the span of a decade
(National Bureau of Statistics of China,
2010, 2021). The results indicate that hog
farming across different scales does not
demonstrate optimal technical eciency.
Yet, larger-scale operations tend to exhibit
higher levels of technical efficiency.
Backyard hog farms typically remain small-
scale, with many still employing traditional
labour-intensive feeding methods, such as
using crop straw and swill for feed (Xiao
et al., 2012). These practices contribute
to low levels of technical proficiency
and production efficiency. Due to the
recent sharp rise in feed prices, backyard
farmers have been compelled to reduce feed
quantity, leading to an increased reliance on
swill feeding. Regrettably, this tendency is
anticipated to have adverse eects on hog
health, including elevated mortality rates as
well as lower farming eciency (Xiao et al.,
2012). Therefore, it is recommended that
larger-scale breeding operations be adopted
in the hog farming industry.
Factors Aecting Technical Ineciency
The results of the Tobit regression analysis
used to identify the factors affecting
inefficiency in the Shaanxi livestock
industry of each category are presented
in Table 3. The expenses of medical and
epidemic prevention have a negative impact
on the overall technical ineciency score,
which implies that increasing prevention
expenses might lead to a decrease in the
ineciency score, subsequently increasing
the technical efficiency of Shaanxi’s
livestock industry. As discovered by Yan et
al. (2023), compared to backyard farmers,
large-scale hog farmers are more proactive
in terms of biosecurity construction, aiming

1177Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Assessing Technical Efficiency in Livestock Production
to reduce losses and enhance production
eciency. Medium- and large-scale farms
exhibit enhanced financial capabilities,
providing farmers with greater capital to
invest in improved disease-preventive
measures. In turn, it facilitates increased
specialisation and reduces the susceptibility
to disease outbreaks. However, increasing
investment in disease prevention necessitates
transitioning farms into medium- and large-
scale operations. Therefore, policies that
support the farm consolidation of backyard
and small-scale farms into larger ones
and facilitate the growth of medium-scale
farms are necessary to increase production
and technical efficiency (Crop Farming
Management Oce of Shaanxi Province,
2022). Consolidating and expanding farms
would allow larger farms to allocate more
resources to preventing outbreaks like avian
inuenza (Wang, Zhou, et al., 2021).
Additionally, it is observed that death
loss positively correlates with technical
ineciency, consistent with ndings by Jo et
al. (2021). Death loss results in a decrease in
ineciency scores, subsequently enhancing
the technical efficiency of the livestock
industry in Shaanxi, particularly in hog and
layer farming. The breeding environment
for hogs and layers is typically less sanitary,
which can more easily result in death loss.
An elevated death loss is associated with a
decline in technical eciency, resulting in
decreased output production and diminished
farmer income. This trend prompts the
departure of numerous backyard and small-
scale farmers from the livestock industry
(Wang, Han, et al., 2021).
On the other hand, the average selling
price of livestock products shows a negative
impact on technical inefficiency in the
Shaanxi livestock industry, suggesting
that a higher selling price might lead to
a decrease in the inefficiency score and
improve the technical eciency of Shaanxi’s
livestock industry. The market selling price
and revenue of livestock products are key
factors determining the duration of livestock
feeding days. Farmers can adjust the length
of feeding days to manage their production
costs and maintain their income during
price fluctuations. When farmers face an
increase in the prices of livestock products,
they may choose to shorten the duration of
feeding days. Farmers may increase feeding
days when food prices fall to maintain their
Table 3
Results of the Tobit regression model on factors that aect technical ineciency
Factors Estimate Std. Error Prob.
VRS CRS VRS CRS VRS CRS
Intercept 0.3678 0.8534 0.1857 0.1636 0.0477** 0.0000***
MEP -0.0090 -0.0102 0.0021 0.0018 0.0000*** 0.0000***
DL 0.0181 0.0136 0.0064 0.0054 0.0044*** 0.0111**
SP -0.0006 -0.0009 0.0002 0.0002 0.0068*** 0.0000***
Note. *** and ** denote signicance at the 1 per cent and 5 per cent levels, respectively
Source: Author’s work

1178 Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Yujie Li, Ee Shiang Lim and Ghee-Thean Lim
income. It illustrates the relationship between
higher prices and increased efficiency, as
well as the inverse association between
lower prices and declining eciency due
to prolonged feeding periods. It is not a
healthy practice for farmers and consumers.
Therefore, it is imperative for the government
to implement measures to stabilise prices,
enabling farmers to maintain a sustainable
livestock feeding base.
CONCLUSION
This study employs Data Envelopment
Analysis (DEA) to investigate technical
efficiency. It utilises Tobit regression to
examine the factors aecting ineciency
across various scales and species of farming
operations in the Shaanxi livestock industry
from 2010 to 2019.
The main ndings of this study are as
follows. Firstly, the livestock industry in
Shaanxi exhibits ineciencies in current
farming practices. Goat, dairy cow and beef
farming exhibit full technical efficiency.
However, both hog and layer farming
practices show the presence of technological
ineciency and scale ineciency. Notably,
the results indicate that the technical
eciency scores of hog and layer farming
increase as the breeding scale increases. This
study also employed the Tobit regression to
explore the potential eect between technical
ineciency and three inuencing factors.
The findings indicated that increasing
medical and epidemic prevention expenses,
reducing death loss and raising selling prices
are crucial improvements that can enhance
industry performance.
From a policy perspective, the ndings
of this study provide valuable insights
for policymakers. Firstly, the breeding of
dairy cows, goats and cattle in Shaanxi
has reached full technical efficiency,
suggesting that the government should
motivate more farmers to participate in
these farming activities. Secondly, the
technical efficiency of hog and layer
farming increases with the expansion of the
production scale. Therefore, the Shaanxi
government could actively promote the
establishment of livestock production
cooperatives and incentivise farmers
to participate by combining adjacent
breeding facilities. This approach would
enable farms to accumulate funds to
enhance epidemic prevention measures.
Additionally, reducing death losses could
enhance technical efficiency within the
livestock industry. Hence, the Shaanxi
provincial authority could implement
subsidies and encourage farmers to reduce
stocking density. Lastly, raising the selling
price of livestock products can enhance
the technical efficiency of the industry
and stimulate farmers to increase livestock
numbers and achieve higher profits.
However, it is important to note that this
higher selling price is not beneficial to
consumers. Therefore, it is advisable for
the Shaanxi Provincial Government to
establish market regulatory mechanisms,
such as price support and reserve systems,
to mitigate the eects of price volatility.
Implementing these measures can help
achieve the government’s goal of boosting
Shaanxi’s livestock production.

1179Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Assessing Technical Efficiency in Livestock Production
However, this study’s limitation is
its oversight of the livestock industry’s
environmental and sustainability impact.
Pursuing higher output through excessive
resource exploitation may adversely
aect Shaanxi’s ecosystems and natural
resources. Therefore, it is recommended
that future studies focus on examining
the environmental eciency of Shaanxi’s
livestock industry. By quantifying the
relationship between resource utilisation
and environmental impacts, measures can
be formulated to promote the sustainable
development of Shaanxi’s livestock
industry.
ACKNOWLEDGEMENTS
The authors thank all participants for their
valuable cooperation in this study. Special
thanks are extended to Universiti Sains
Malaysia (USM) for the support provided
throughout this study.
REFERENCES
Aigner, D. J., Lovell, C. A. K., & Schmidt, P. (1977).
Formulation and estimation of stochastic
frontier production function models. Journal
of Econometrics, 6(1), 21–37. https://doi.
org/10.1016/0304-4076(77)90052-5
AskCI Consulting. (2022, January 22). China’s
meat imports analysis 2021. AskCI.
https://www.askci.com/news/data/
maoyi/20220124/0843371733022.shtml
Banker, R. D., Charnes, A., & Cooper, W. E. (1984).
Some models for estimating technical and scale
ineciencies in Data Envelopment Analysis.
Management Science, 30(9), 1078–1092. https://
doi.org/10.1287/mnsc.30.9.1078
Cadell, C. (2020, November 15). China’s sweeping
COVID controls put frozen food importers
on ice. Reuters. https://www.reuters.com/
article/us-health-coronavirus-china-frozenfood-
idUSKBN27T0YK
Coelli, T. J., Rao, D. S. P., O’Donnell, C. J., & Battese,
G. E. (2005). An introduction to eciency and
productivity analysis (2nd ed.). Springer.
Crop Farming Management Office of Shaanxi
Province. (2022). Interpretation of the
development plan for fruit, vegetable, and
tea industries. Department of Agriculture
and Rural Affairs of Shaanxi Province.
http://nynct.shaanxi.gov.cn/www/
bm7831/20221206/9811351.html
Dogan, N., Kaygisiz, F., & Altinel, A. (2018).
Technical and economic efficiency of laying
hen farms in Konya, Turkey. Brazilian Journal
of Poultry Science, 20(2), 263–272. https://doi.
org/10.1590/1806-9061-2017-0649
Food and Agriculture Organization. (2021). World
food and agriculture statistical yearbook 2021.
https://www.fao.org/policy-support/tools-and-
publications/resources-details/zh/c/1458575/
Food and Agriculture Organization. (2022a).
FAOSTAT-Crop and livestock products. https://
www.fao.org/faostat/en/#data/QCL
Food and Agriculture Organization. (2022b). Trade
of agricultural commodities 2005–2021.
https://openknowledge.fao.org/server/api/
core/bitstreams/75eaf30a-b1b0-4e7b-8e21-
d8970ff5833b/content#:~:text=The%20
share%20of%20agricultural%20products,to%20
7.9%20percent%20in%202021.
Färe, R., & Lovell, C. a. K. (1978). Measuring the
technical eciency of production. Journal of
Economic Theory, 19(1), 150–162. https://doi.
org/10.1016/0022-0531(78)90060-1
Farrell, M. J. (1957). The measurement of productive
efficiency. Journal of the Royal Statistical

1180 Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Yujie Li, Ee Shiang Lim and Ghee-Thean Lim
Society. Series a (General), 120(3), 253–290.
https://doi.org/10.2307/2343100
Fraser, I., & Cordina, D. (1999). An application
of data envelopment analysis to irrigated
dairy farms in Northern Victoria, Australia.
Agricultural Systems, 59(3), 267–282. https://
doi.org/10.1016/s0308-521x(99)00009-8
Galluzzo, N. (2019). An analysis of technical
eciency in Icelandic dairy and sheep farms.
Studies in Agricultural Economics, 121(3),
144–150. https://doi.org/10.7896/j.1916
Greene, W. H. (1994). Econometric analysis. Journal
of the American Statistical Association, 89(428),
1567–1569. https://doi.org/10.2307/2291031
Hansson, H., & Öhlmér, B. (2008). The effect of
operational managerial practices on economic,
technical and allocative eciency at Swedish
dairy farms. Livestock Science, 11 8(1–2), 34–43.
https://doi.org/10.1016/j.livsci.2008.01.013
Heshmati, A., Kumbhakar, S. C., & Hjalmarsson,
L. (1995). Efficiency of the Swedish pork
industry: A farm level study using rotating
panel data 1976–1988. European Journal of
Operational Research, 80(3), 519–533. https://
doi.org/10.1016/0377-2217(94)00135-y
İkikat, T. E., Ağır, H., & Aydoğan, İ. (2020).
Evaluating technical efficiency of hair goat
farms in Turkey: The case of Mersin Province.
Tropical Animal Health and Production, 52(6),
3707–3712. https://doi.org/10.1007/s11250-
020-02407-2
Jo, H., Nasrullah, M., Jiang, B., Li, X., & Bao, J.
(2021). A survey of broiler farmers’ perceptions
of animal welfare and their technical eciency:
A case study in Northeast China. Journal
of Applied Animal Welfare Science, 25(3),
275–286. https://doi.org/10.1080/10888705.2
021.1912605
Kuhn, L., Baležentis, T., Hou, L., & Wang, D.
(2020). Technical and environmental eciency
of livestock farms in China: A slacks-based
DEA approach. China Economic Review, 62,
Article 101213. https://doi.org/10.1016/j.
chieco.2018.08.009
Lansink, A. O., & Reinhard, S. (2004). Investigating
technical eciency and potential technological
change in Dutch pig farming. Agricultural
Systems, 79(3), 353–367. https://doi.org/10.1016/
s0308-521x(03)00091-x
Liu, W., Shankar, S., & Li, L. (2021). Is specialization a
strategy to improve farm eciency in Northwest
China? Review of Development Economics,
25(3), 1695–1710. https://doi.org/10.1111/
rode.12782
Luik-Lindsaar, H., Põldaru, R., & Roots, J. (2019).
Estonian dairy farms’ technical eciency and
factors predicting it. Agronomy Research, 17(2),
593–607. https://doi.org/10.15159/ar.19.067
Meeusen, W., & Van Den Broeck, J. (1977). Eciency
estimation from Cobb-Douglas production
functions with composed error. International
Economic Review, 18(2), 435–444. https://doi.
org/10.2307/2525757
Mohd Idris, N. D., Chamhuri, S., & Talib, B.
(2013). Determinants of technical efficiency
on pineapple farming. American Journal of
Applied Sciences, 10(4), 426–432. https://doi.
org/10.3844/ajassp.2013.426.432
Mugera, A. W., & Featherstone, A. M. (2008). Backyard
hog production eciency: Evidence from the
Philippines. Asian Economic Journal. https://doi.
org/10.1111/j.1467-8381.2008.00278.x
National Bureau of Statistics of China. (2010). China
statistical yearbook 2010. https://www.stats.gov.
cn/sj/ndsj/2010/indexeh.htm
National Bureau of Statistics of China. (2021). China
statistical yearbook 2021. https://www.stats.gov.
cn/sj/ndsj/2021/indexeh.htm
National Bureau of Statistics of China. (2023).
Statistical communiqué of the People’s Republic
of China on the 2022 national economic and social

1181Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Assessing Technical Efficiency in Livestock Production
development. http://www.stats.gov.cn/english/
PressRelease/202302/t20230227_1918979.html
Patton, D. (2022, February 11). China targets 15%
increase in meat output by 2025. Nasdaq. https://
www.zawya.com/en/business/china-targets-15-
increase-in-meat-output-by-2025-hwsq83x2
Price Department of the National Development and
Reform Commission (2019). Compilation of
national agricultural product cost and benet
data-2019. China Statistics Press. https://
www.stats.gov.cn/zs/tjwh/tjkw/tjzl/202302/
t20230215_1907941.html
Reinhard, S., Lovell, C. A. K., & Thijssen, G.
(1999). Econometric estimation of technical
and environmental eciency: An application
to Dutch dairy farms. American Journal of
Agricultural Economics, 81(1), 44–60. https://
doi.org/10.2307/1244449
Shaanxi Government. (2008). Special plan for key
areas of rural economic development during the
10th Five-Year plan for national economic and
social development in Shaanxi province. http://
www.shaanxi.gov.cn/zfxxgk/fdzdgknr/ghxx/
swgh/zxgh_2451/200801/t20080101_1661543.
html
Shaanxi Government. (2022). Modern fruit industry
development plan for the 14th Five-Year plan in
Shaanxi province. http://www.shaanxi.gov.cn/
xw/ldx/bm/202210/t20221027_2257223_wap.
html
Shaanxi Provincial Department of Agriculture. (2018).
Modern agriculture development plan for the
13th Five-Year plan in Shaanxi province (2016-
2020). http://nynct.shaanxi.gov.cn/jgw/home/
index/detail/catId/738/articleId/5200.html
Shaanxi Provincial Department of Agriculture. (2022).
Shaanxi Provincial Department of Agriculture
and Rural Aairs on the issuance of the “14th
Five-Year plan” for the development of animal
husbandry and veterinary medicine in Shaanxi
province. http://nynct.shaanxi.gov.cn/www/
snynctwj/20220428/9792947.html
Shaanxi Statistics Bureau. (2015). Transformation of
agricultural development mode and enhancement
of product supply capacity. http://tjj.shaanxi.gov.
cn/sy/ztzl/zgcyksx/201510/t20151019_1615133.
html
Sharma, K. R., Leung, P., & Zaleski, H. M.
(1997). Productive efficiency of the swine
industry in Hawaii: Stochastic frontier vs.
Data envelopment analysis. Journal of
Productivity Analysis, 8(4), 447–459. https://
doi.org/10.1023/a:1007744327504
Shi, Z. Z., Wang, M. L., & Hu, X. D. (2015). A study
on beef market partial equilibrium model in
China. Journal of China Agricultural University,
20, 278–290.
Siafakas, S., Tsiplakou, E., Kotsarinis, M., Tsiboukas,
K., & Zervas, G. (2019). Identification of
ecient dairy farms in Greece based on home
grown feedstus, using the Data Envelopment
Analysis method. Livestock Science, 222, 14–20.
https://doi.org/10.1016/j.livsci.2019.02.008
Soh, N. C., Samsuddin, N. S., Ismail, M. R., &
Habibullah, M. S. (2021). Technical eciency
of commercial stingless bee honey production
in Peninsular Malaysia. Pertanika Journal of
Social Science and Humanities, 29(2). https://
doi.org/10.47836/pjssh.29.2.03
Somwaru, A., Zhang, X., & Tuan, F. C. (2003).
China’s hog production structure and eciency
[Paper Presentation]. American Agricultural
Economics Association Annual Meeting 2003,
Montreal, Canada. https://doi.org/10.22004/
ag.econ.22003
Terry, S. A., Basarab, J. A., Guan, L. L., & McAllister,
T. A. (2021). Strategies to improve the eciency
of beef cattle production. Canadian Journal
of Animal Science, 101(1), 1–19. https://doi.
org/10.1139/cjas-2020-0022

1182 Pertanika J. Soc. Sci. & Hum. 32 (3): 1165 - 1182 (2024)
Yujie Li, Ee Shiang Lim and Ghee-Thean Lim
Tian, X., Sun, F., & Zhou, Y. (2015). Technical
eciency and its determinants in China’s hog
production. Journal of Integrative Agriculture,
14(6), 1057–1068. https://doi.org/10.1016/
s2095-3119(14)60989-8
Wang, G., Zhao, C., Shen, Y., & Yin, N. (2021).
Estimation of cost eciency of fattening pigs,
sows, and piglets using SFA approach analysis:
Evidence from China. PLOS ONE, 16(12),
Article e0261240. https://doi.org/10.1371/
journal.pone.0261240
Wang, Y., Han, Y., Han, Y., & Li, W. (2021b).
Does the grassland ecological compensation
policy improve the herders’ breeding technical
efficiency in China? Based on the parallel
mediation effect model. PLOS ONE, 16(4),
Article e0249990. https://doi.org/10.1371/
journal.pone.0249990
Xiao, H., Wang, J., Oxley, L., & Ma, H. (2012).
The evolution of hog production and potential
sources for future growth in China. Food Policy,
37(4), 366–377. https://doi.org/10.1016/j.
foodpol.2012.02.002
Yan, Z., Wang, M., Li, X., & Jiang, H. (2023).
Impact of African swine fever epidemic on
the cost intensity of pork production in China.
Agriculture, 13(2), Article 497. https://doi.
org/10.3390/agriculture13020497
Yang, C. C., Hsiao, C. K., & Yu, M. M. (2008).
Technical eciency and impact of environmental
regulations in farrow-to-nish swine production
in Taiwan. Agricultural Economics, 39(1),
51–61. https://doi.org/10.1111/j.1574-
0862.2008.00314.x
Zhang, Z., Zhang, Y., Li, F., Yang, H., Yuan, Y.,
& Yuan, X. (2017). Economic efficiency of
small-scale tilapia farms in Guangxi, China.
Aquaculture Economics & Management, 21(2),
283–294. https://doi.org/10.1080/13657305.20
16.1180644
Zhong, S., Li, J., Chen, X., & Wen, H. (2021).
Research on the green total factor productivity
of laying hens in China. Journal of Cleaner
Production, 315, Article 128150. https://doi.
org/10.1016/j.jclepro.2021.128150
Zhou, Y., Zhang, X., Tian, X., Geng, X., Zhang, P.,
& Yan, B. (2015). Technical and environmental
efficiency of hog production in China – A
stochastic frontier production function analysis.
Journal of Integrative Agriculture, 14(6),
1069–1080. https://doi.org/10.1016/s2095-
3119(14)60990-4
Zhu, J. (2009). Quantitative models for performance
evaluation and benchmarking: Data envelopment
analysis with spreadsheets (2nd ed.). Springer.
https://doi.org/10.1007/978-0-387-85982-8