Tóm tắt Luận án Economic and environmental efficiency of intensive shrimp farming in the coastal transforming areas of the mekong delta

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

pdf37 trang | Chia sẻ: honganh20 | Ngày: 02/03/2022 | Lượt xem: 368 | Lượt tải: 0download
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:

  • pdftom_tat_luan_an_economic_and_environmental_efficiency_of_int.pdf
Tài liệu liên quan