LIST OF TABLE . iv
LIST OF FIGURE . v
CHAPTER 1: INTRODUCTION. 1
1.1 RATIONALE OF THE DISSERTATION . 1
1.2 OBJECTIVES . 3
1.2.1 General objectives. 3
1.2.2 Specific objectives . 3
1.3 RESEARCH QUESTIONS. 4
1.4 RESEARCH SUBJECTS . 4
1.5 SCOPE OF THE STUDY. 4
CHAPTER 2: LITERATURE REVIEW . 5
2.1 REVIEW OF NEW TECHNOLOGY ACCEPTANCE
MODEL AND FARMING TRANSFORMATION . 5
2.2 REVIEW ABOUT ECONOMIC EFFICIENCY. 6
2.3 REVIEW ABOUT ENVIRONMENTAL EFFICIENCY . 7
2.4 REVIEWS OF FACTORS AFFECTING EFFICIENCY . 9
CHAPTER 3: THEORY AND RESEARCH METHODOLOGY
. 10
3.1. THEORETICAL CONCEPTS. 10
3.1.1 Intensive farming . 10
3.1.2 Factors affecting changes in farming systems . 10
3.1.3 Economic efficiency and measurement methods. 11
3.1.4. Environmental efficiency and measurement method. 11
37 trang |
Chia sẻ: honganh20 | Ngày: 02/03/2022 | Lượt xem: 368 | Lượt tải: 0
Bạn đang xem trước 20 trang tài liệu Tóm tắt Luận án Economic and environmental efficiency of intensive shrimp farming in the coastal transforming areas of the mekong delta, để xem tài liệu hoàn chỉnh bạn click vào nút DOWNLOAD ở trên
he technology
characteristics-user's context model (Negatu and Parikh, 1999)
argues that technology characteristics are a basic component
affecting an individual people to adopt that technology. The model
also considered the cognitive characteristics of potential adopters as
influences on adoption decisions. The second theoretical model is
the utility maximization (Rahm and Huffman, 1984; Sidibé, 2005)
argues that farmers are more likely to apply new technologies,
innovations or practices if the utility from them (technologies) is
larger than older ones.
6
2.2 REVIEW ABOUT ECONOMIC EFFICIENCY
Economic efficiency was first proposed by Farrell (1957) through
the term total efficiency or overall efficiency. Economic efficiency
is defined as the ability to produce a fixed output with the lowest
input cost or the product of technical efficiency and allocative
efficiency (Farrell, 1957; Schmidt and Lovell, 1979, 1980; Kopp,
1981; Bravo‐ Ureta and Pinheiro, 1997). According to Coelli et al.
(2005); Kumbhakar & Lovell (2003), economic efficiency can be
cost efficiency, revenue efficiency and profit efficiency. Cost-
efficiency or economic efficiency shows the ability to produce a
certain output at the lowest cost with the corresponding input price
(Farrell, 1957; Battese, 1992; Bravo‐ Ureta and Pinheiro, 1997;
Reinhard et al., 1999; Reinhard et al., 2000; Coelli et al., 2002;
Coelli et al., 2005; Khai and Yabe, 2011). Economic efficiency can
be measured by using the SFA method. The approach was first
proposed by Aigner, Lovell and Schmindt (1977) and Meeusen and
Van Den Broeck (1977).
Measuring economic efficiency has been conducted for various
agricultural production activities and is considered as the basis for
assessing whether a certain production model is efficient or not. In
order to estimate the economic efficiency, there are normally two
measurement methods (1) using the profit function or (2) using the
cost function. Some case studies using the profit function include
Pham Le Thong et al (2011); Nguyen Van Tien and Pham Le Thong
(2014); Pham Le Thong and Nguyen Thi Phuong (2015); Nguyen
Minh Hieu (2014). Some other authors have used the cost function
approach to measure economic efficiency towards the direction of
cost minimization such as Ferrier and Lovell (1990); Worthington
(2000); Rosko (2001).
In order to estimate the economic efficiency, recently the one-step
estimation model is recommended by econometric experts instead
7
of the two-step estimation approach as the one-step approach can
control the estimation bias (Caudill & Ford, 1993; Wang &
Schmidt, 2002; Caudill, 2003; Greene, 2005; Belotti et al., 2013;
Kumbhakar et al., 2015 ).
In summary, the above studies normally estimated economic
efficiency via the profit or cost frontier by a two-step approach. In
addition, these previous studies using the Cobb-Doughlas function
and the DEA method make it impossibly to separate the noise
effects apart from inefficiency effects and to investigate the causes
of inefficiency. Therefore, the dissertation focuses on estimating
economic efficiency towards the direction of cost minimization by
using one-step SFA to overcome the limitations of two-step
estimation.
2.3 REVIEW ABOUT ENVIRONMENTAL EFFICIENCY
Pittman (1983) is probably considered to be the first to concern
about environmental issues when estimating the efficiency for
production activities. In this study, the author considered the
environmental aspect as an unexpected output from production
process. The study is crucial for policy makers to control pollution
under the context of undesirable–desirable outputs trade-off.
However, measuring undesirable output is a difficult task, especially
in agricultural production.
Färe et al. (1989) proposed the term enhanced hyperbolic
productive efficiency measure. This term considers simultaneously
the difference among the maximum equiproportional increase in
desirable outputs, the maximum equiproportional decrease in
undesirable outputs and the maximum equiproportional decrease in
inputs. However, the study measured productive efficiency using
nonparametric approach, which cannot separate noise effect apart
from deterministic frontier. In addition, again measuring undesirable
output is a difficult task, especially in agricultural production.
8
In order to overcome the drawbacks of the previous studies and
respect the material balance principle, Reinhard et al. (1999) treated
environmental pollution as input surpluses (e.g., fertilizers,
pesticides, energy) to estimate environmental efficiency. As the
environmentally detrimental inputs such as chemical fertilizers,
pesticides, fuels, ... have a close relationship with the unexpected
output (pollution), minimizing the unexpected output can be done
through minimizing the environmentally detrimental inputs.
Some case studies that estimated environmental efficiency include
Vo Hong Tu (2015); Tu et al. (2015). These studies presented the
method of measuring environmental efficiency for agricultural
production by using SFA and the results on the environmental
efficiency of ecological engineered rice production, also known as
"rice fields surrounded with flowers" in An Giang province. The
study defined the environmental efficiency as the ratio of possible
minimum environmentally detrimental inputs (fertilizers, pesticides,
fuels) to its observed amount or in other words it reflects the ability
to reduce environmentally detrimental inputs.
Hong et al. (2016) employed the approach of Reinhard et al. (2000)
to measure environmental efficiency for 243 tea producing
households in Thai Nguyen province. This study considered two
inputs that have a negative impact on the environment: chemical
fertilizers and pesticides and other normal inputs include labor,
capital, irrigation costs and other costs. The research results show
that the average environmental efficiency of tea growers is 76.03%
and there is a large variation in environmental efficiency among
farmers. Tu (2015) also uses the approach of Reinhard et al. (2000);
Reinhard & Thijssen (2000) to measure the efficiency of using input
resources for rice farmers in An Giang province. The research
results showed that the input and output-oriented technical
efficiency is 91.92% and 85.39%, respectively. The returns to scale
of rice farmers is decreasing. The study also found that the least
9
efficiently used inputs were pesticide and fuel with the average
efficiency indexes of 51.39% and 45.53%, respectively. This
inefficient use has resulted in a significant economic loss of about
VND 8.2 million/ha.
Therefore, this dissertation uses the one-step method to specify
production frontier function, from which to measure the
environmental efficiency for the intensive shrimp farming in the
coastal transforming areas.
2.4 REVIEWS OF FACTORS AFFECTING EFFICIENCY
The Tobit regression model was first developed by Tobin (1958) to
consider the correlation between dependent and independent
variables, in which the dependent variable is censored and non-
negative. In the field of agricultural economics, Tobit model is used
to investigate the effects of independent variables namely socio-
economic conditions (gender of household head, educational level,
participation in training ...) on technical efficiency, cost efficiency
and economic efficiency of agricultural production models (Thai
Thanh Ha, 2009; Tu & Yabe, 2015)
The Tobit regression is considered to be the second step in the
efficiency studies because the results from the Tobit regression will
be an important basis for finding the gaps in efficiency levels among
households, from which one can propose appropriate solutions to
improve efficiency (Färe & Lovell, 1978; Bravo-Ureta & Rieger,
1991; Bravo-Ureta & Pinheiro, 1993; Bravo ‐ Ureta & Pinheiro,
1997; Khai & Yabe, 2011). Because the efficiency level is bounded
in a range or in other words, it is censored within a certain limit.
Thus,the estimation results from the Tobit regression will be less
biased than OLS regression (Tobin, 1958; Grigorian & Manole,
2006; Tu & Trang, 2015).
10
CHAPTER 3: THEORY AND RESEARCH METHODOLOGY
3.1. THEORETICAL CONCEPTS
3.1.1 Intensive farming
According to Nguyen Thanh Phuong et al. (2014), intensive shrimp
farming is a farming method with productivity <200 tons/ha/year,
good control of farming conditions; high farming techniques and
high production efficiency; tend to actively control all farming
conditions (feed and water quality); and highly artificial farming
system”.
3.1.2 Factors affecting changes in farming systems
According to Negatu & Parikh (1999), the characteristics of
technology are a fundamental component in identifying an
individual adopting new technology. Besides, Rahm & Huffman
(1984); Sidibé (2005) argues that farmers are more likely to adopt
new agricultural technologies, innovations or practices if the utility
from new technology is greater than that from the old ones.
Combining these two theoretical models, groups of variables are
often used in analyzing factors influencing new technology adoption
decisions including (1) socio-demographic characteristics (age,
education, experience, labor and female workers), (2) perceived
risk, (3) perceived usefulness (output, price, benefit), (4)
environmental awareness (pollution and biodiversity), (5) perceived
ease of use (technical aspect); (6) farm characteristics (size of land
and number of land plots), (7) social networks (membership in
organizations) and (8) financial characteristics (awareness of
external support and access to credit) (Adesina and Zinnah, 1993;
Barreiro-Hurlé et al., 2010; Davis, 1989; Negatu and Parikh, 1999;
Sidibé, 2005; Wang et al., 2016). The studies on this aspect
normally use the logit regression model or generalized ordered logit
model or structural equation modeling (SEM) in determining the
factors that influence the decision to convert or to adopt new
techniques.
11
3.1.3 Economic efficiency and measurement methods
Economic efficiency is defined as the ability to produce a given
output level at the optimal cost or regarded as the product of
technical efficiency and allocative efficiency (Farrell, 1957; Kopp,
1981; Bravo ‐ Ureta & Pinheiro, 1997).
In order to estimate economic efficiency by using SFA, the study
uses translog variable cost frontier by one step method to estimate
parameters and economic inefficiency as a farm is assumed to
achieve a static equilibrium with respect to a subset of normal inputs
conditionally on observed levels of quasi-fixed inputs (Brown &
Christensen, 1980; Caves et al. ., 1981). In addition, we cannot
estimate the total cost function because the price of some inputs is
not available in the market (Grisley & Gitu, 1985).
3.1.4. Environmental efficiency and measurement method
To measure environmental efficiency, there have been two main
approaches: DEA and SFA. As the DEA approach is non-
parametric, which calculates efficiency indexes based mathematic
programming. Therefore, it is impossible to separate noise effects
apart from deterministic frontier. Thus, the dissertation measures
environmental efficiency by using the SFA approach. SFA
approach is based on econometric model so it can overcome the
drawbacks of DEA (Tu & Yabe, 2015).
Suppose a farmer uses two types of inputs, denoted by X and Z, to
produce an output, denoted by Y (Y ), where X ( ) is a
vector of normal inputs such as labor, capital, ... and Z ( ) are
environmentally detrimental inputs such as feed, medicines and
fuel. Environmental efficiency is the ability to reduce
environmentally detrimental inputs while other inputs and outputs
are fixed. Similar to economic efficiency, the dissertation also uses
a one-step approach to measuring environmental efficiency.
12
3.2. RESEARCH METHODOLOGY
3.2.1. Theoretical framework
Figure 3.1 describes in detail the theoretical framework of the study.
Figure 3.1: Theoretical framework of the study
Source: Author
In the context of climate change and market variability, the
transformation in agricultural production has been taking place as an
inevitable phenomenon. In order to propose solutions to manage
transformation of agricultural production models, the dissertation
focuses on analyzing the factors affecting the transformation and
conducting comparison of financial indicators between the two
farming activities. For intensive shrimp farmers, the study used a
SFA approach to measure economic and environmental efficiency,
thereby contributing to proposing solutions to improve economic
13
efficiency and to mitigate environmental pollution for production
activities in the coastal transforming regions of the MD.
3.2.2 Selection of study sites
The study selected two coastal provinces in the MD having the
highest rate of conversion to intensive shrimp, namely Soc Trang
and Kien Giang, of which Kien Giang is influenced by the West sea
and Soc Trang is affected by the East sea. In the period from 2011-
2015, the shrimp production area of Soc Trang province increased
by an average of 13.3%/year, which is the highest growth rate
among other coastal provinces such as Tra Vinh 7.5%/year, Bac
Lieu 7.2 %/year, Ben Tre 4.3%/year and Ca Mau 4.1%/year (GSO,
2015). Kien Giang province was selected as the study site because
this is the only coastal province in the MD affected by the West sea.
3.2.3. Analytical methods
- The study uses descriptive statistical tools to describe the
current production situation and changes in farming activities
from sugarcane to shrimp in Soc Trang province and from rice
- shrimp to shrimp in Kien Giang, and CRA method to analyze
the financial performance of the converted shrimp model.
- To identify the factors influencing the decision of changes in
farming activities from sugarcane to shrimp and from rice-
shrimp to shrimp farming, the dissertation uses logit regression.
- To estimate the economic and environmental efficiency, the
dissertation use the SFA method proposed by Aigner. Lovell &
Schmindt (1977) and Meeusen & Van Den Broeck (1977). The
translog function will be used to specify agricultural production
technology (Coelli et al., 2005) and to estimate efficiency
indexes (Reinhard et al., 1999).
14
+ For estimating economic efficiency, the dissertation
estimates economic efficiency towards the direction of
minimizing costs by using the one-step approach.
+ For environmental efficiency, in the case of shrimp farming,
the environmentally detrimental inputs are feed, medicine
and fuel.
- Because the study uses a one-step approach, the factors
affecting economic efficiency were carried out simultaneously
with the process of estimating the cost frontier. However, for
environmental efficiency, it is calculated through the output-
oriented technical efficiency, which is also estimated through a
one-step approach. Therefore, to propose solutions to improve
environmental efficiency, the dissertation employs Tobit
regression.
15
CHAPTER 4: RESULT AND DISCUSSION
4.1 STATUS OF CHANGES IN FARMING SYSTEMS TO
SHRIMP
4.1.1 Status of changes in farming systems
Transformation of agricultural production activities is an inevitable
phenomenon with the expectation of greater profits. The
comprehensive picture of the transformation of agricultural
production in coastal areas in the study sites is described in detail in
Figure 4.1.
Figure 4.1: Trends in changes of farming systems in Soc Trang
and Kien Giang
Figure 4.1 shows that in the fresh water area, farmers in the study
sites mainly produce rice and sugarcane. In the brackish water area,
also known as the transforming zone affected by salinity intrusion,
many farmers have shifted their production activities from
sugarcane to shrimp in Soc Trang and from rice - shrimp to shrimp
in Kien Giang. This means that salinity intrusion is an important
factor affecting the transformation of agricultural production
Increasing of salinity levels & salinity affected period
SEA Rice and
sugarcane
Mono-
shrimp
Rice-shrimp and
sugarcane
Integrated
shrimp-mangrove
L
ev
el
o
f
a
d
o
p
ti
o
n
/c
h
a
n
g
es
Level of salinity (from fresh – brackish – saline
water)
Rice
and
sugar-
cane
Rice-shrimp and
sugarcane
Mono-shrimp
16
activities. To test this hypothesis, the dissertation uses the variable
namely distance from fields/ponds to river as the independent
variable to replace for the variable salinity magnitude due to the
lack of values of salinity at individual farmers.
4.1.2. Factors affecting changes in farming systems to shrimp
The estimated results of the effects of the socio-economic variables
on the transformation decision are presented in Table 4.1 below:
Table 4.1: Logit results of factors affecting changes in farming
systems
Variables
Soc Trang province
(Sugarcane mono-
shrimp)
Kien Giang province
(Rice-shrimp mono-
shrimp)
Coef. s.e dy/dx Coef. s.e dy/dx
SEX
0.132 0.863 0.0329 0.441 0.772 0.1093
AGE
-0.027 0.029 -0.0066 0.013 0.019 0.0033
LABOR
0.016 0.416 0.0039 0.260 0.348 0.0650
FEMALE
-0.994* 0.592 -0.2483 0.429 0.499 0.1074
EDUCATION
0.153* 0.081 0.0382 0.124* 0.073 0.0309
ORGANIZATION
-1.650* 0.887 -0.3682 -1.209 0.853 -0.2807
CREDIT
-2.682*** 0.608 -0.5853 0.158 0.526 0.0395
AGRILAND
-0.168*** 0.053 -0.0422 -0.015 0.012 -0.0038
DISTANCE
-0.004*** 0.001 -0.0012 -0.004*** 0.001 -0.0009
EXPERIENCE
0.253*** 0.056 0.0633
Intercept 4.731** 2.026 -3.476** 1.503
Note: * indicates the significant level; *p < 0.1; **p < 0.05; ***p < 0.01
s.e stands for standard error; dy/dx indicates marginal effects.
Source: Own estimates; data appendix available from authors.
For the case of Soc Trang province (conversion from sugarcane to
shrimp), Table 4.1 shows the variables namely female labor, credit
access, membership in organizations, agricultural land area and
distance had negative effects on the adoption decision while
17
educational level is the only variable that has a positive relationship
with the dependent variable.
In the case of Kien Giang province (conversion from rice - shrimp
to shrimp), Table 4.1 shows that the distance from the rice field to
the river also had a negative effect while the educational level and
experience of shrimp farming had positive effects on the adoption
decision of new farming activities.
4.2. ECONOMIC AND ENVIRONMENTAL EFFICIENCY OF
TRANSFORMED INTENSIVE SHRIMP FARMING
4.2.1. Economic efficiency of intensive shrimp farming
4.2.1.1 Estimation of economic efficiency
Prior to specifying the cost frontier, we conduct tests to determine
whether the data is best fit with Cobb-Douglas or translog function
by using LR - log-likelihood ratio test (Coelli et al., 2005 ; Greene,
2012; Kumbhakar et al., 2015). The LR test result shows that the
value of , which is much greater than the critical
value and significant at 1%. This result shows that the collected data
is best fit with translog function. The results also show that the
translog cost function by a one-step method (taking into account the
correlation between economic inefficiency and socio-economic
characteristics) is accepted compared to the two-step cost function
(excluding independent variables affecting economic inefficiency)
through the value of = 34.49. This value is much larger than the
critical value at 1%. The correlation matrix results also show that
there is no multi-collinearity between independent variables,
specifically the correlation coefficients are less than 0.6.
Because the data were collected from two different provinces and
shrimp farmers had different levels of intensification of white-
legged shrimp, it is necessary to test whether we can estimate the
pooled cost function or not. The results show that there is no
18
significant difference according to the t-test between the two data
sets, except for the variable fuel. Thus, we can estimate cost
function by pooling the data of two groups of shrimp farmers in
Kien Giang and Soc Trang. Regression results are presented in
detail in Table 4.2:
Table 4.2: Estimated results of translog variable cost frontier
Estimated parameters of translog cost frontier
Variables Coef. s.e Variables Coef. s.e
lnW1 8,578 54,603 lnW2lnZ1 -0,053** 0,023
lnW2 2,838 3,716 lnW2lnY 0,024 0,034
lnW3 -0,986 4,252 (lnW3lnW3)/2 -0,002 0,028
lnW4 -7,975 42,042 lnW3lnW4 0,047 0,310
lnW5 -18,267 45,160 lnW3lnW5 -0,152 0,254
lnZ1 -0,636 6,673 lnW3lnZ1 0,017 0,022
lnY -4,923 8,646 lnW3lnY -0,050 0,048
(lnW1lnW1)/2 0,414 1,451 (lnW4lnW4)/2 1,227 1,627
lnW1lnW2 -0,172 0,262 lnW4lnW5 -0,141 2,358
lnW1lnW3 0,266 0,328 lnW4lnZ1 0,337 0,510
lnW1lnW4 0,317 2,522 lnW4lnY -0,373 0,309
lnW1lnW5 -0,555 4,184 (lnW5lnW5)/2 1,338 2,053
lnW1lnZ1 -0,399 0,518 lnW5lnZ1 0,239 0,300
lnW1lnY -0,182 0,608 lnW5lnY 0,629 0,481
(lnW2lnW2)/2 0,015 0,018 (lnZ1lnZ1)/2 0,048 0,042
lnW2lnW3 0,008 0,015 lnZ1lnY -0,004 0,058
lnW2lnW4 -0,225 0,227 (lnYlnY)/2 0,207** 0,103
lnW2lnW5 0,039 0,173 Constant 117,308 551,5
7
Estimated parameters of factors affecting inefficiency (Mu)
Variables Coef. s.e Variables Coef. s.e
Education 0,029 0,129 No. of ponds 1,039** 0,436
Experience 0,041 0,118 Distance -0,004 0,005
Organization 0,356 1,894 Labor -0,003 0,632
Pond size -1,137** 0,457 Constant -0,124 2,188
Density -0,027* 0,015
Usigma -0,607 0,437
Vsigma -2,919*** 0,179 L-Likelihood -9,27
Lamda 3,176*** 0,165 Wald χ2 value 228,33
Source: Household survey in 2017, n = 125
19
From the estimated results of Table 4.2, one can estimate the
economic efficiency for individual shrimp farmers in the study sites.
The economic efficiency is summarized in Table 4.3:
Table 4.3: Economic efficiency of intensive shrimp farming
Economic
efficiency
Soc Trang Kien Giang
Frequency % Frequency %
≥90 42 46,67 23 65,71
80-90 38 42,22 9 25,71
70-80 6 6,67 2 5,72
60-70 2 2,22 0 0
50-60 0 0 1 2,86
40-50 1 1,11 0 0
30-40 0 0 0 0
<30 1 1,11 0 0
Mean 86,95 89,98
Min 22,73 55,35
Max 97,58 97,96
t-value 1,55
Pooled efficiency 87,80
Standard deviation 9,85
Source: Household survey in 2017, n = 125
The results of Table 4.3 shows that the average economic efficiency
of the shrimp farming in Kien Giang province is 89.98%, which is
not significantly different from that of Soc Trang province 86.95%.
This result partly reflects inefficient use and allocation of inputs.
There is a great variation in economic efficiency among farmers in
Kien Giang, specifically the highest efficiency score of 97.96%
while the lowest efficiency score of only 55.35%. Similarly, the
economic efficiency of shrimp farming households in Soc Trang
province is also greatly variated, the highest efficiency score of
97.58% while the lowest one of only 22.73%. In terms of
minimizing costs at the current output level, this great variation may
be explained that there are big differences in knowledge and
technologies among shrimp farmers who recently converted their
20
farming activities. These differences results in different output
levels.
The majority of shrimp farmers in Soc Trang province has the
efficiency scores over 70%, accounting for 95.56%. With this
average economic efficiency, at the current output level, shrimp
farmers in Soc Trang Province can reduce about 13.05% of total
variable cost. The total costs that farmers can reduce or in other
words the difference between actual and potential minimum costs
are shown in Figure 5.5:
Figure 4.2: Observed and possible minimum cost in Soc Trang
Source: Household survey in 2017, n = 125
From Figure 5.5, we can calculate the losses due to economic
inefficiency. In other words, the cost that shrimp farmers in Soc
Trang province can reduce on average (observed cost minus
possible minimum cost) 78.03 million VND/ha/season.
In the case of Kien Giang province, the economic efficiency scores
distributed mainly over 90%, accounting for more than 65.71%. In
Kien Giang, the proportion of shrimp farmers had economic
efficiency distributed above 70% was about 97.14%. For farmers in
Kien Giang province, the study shows that on average, farmers can
Possible reduction of cost
78.03 million
VND/ha/season
21
contract about 10.02% of total current variable
Các file đính kèm theo tài liệu này:
- tom_tat_luan_an_economic_and_environmental_efficiency_of_int.pdf