Estimate the technical efficiency and identify influencing factors of tea farmers in thai nguyen city

THAI NGUYEN UNIVERSITY UNIVERSITY OF AGRICULTURAL AND FORESTRY PHAM HA LUONG ESTIMATE THE TECHNICAL EFFICIENCY AND IDENTIFY INFLUENCING FACTORS OF TEA FARMERS IN THAI NGUYEN CITY BACH

THAI NGUYEN UNIVERSITY

UNIVERSITY OF AGRICULTURAL AND FORESTRY

PHAM HA LUONG

THAI NGUYEN CITY

BACHELOR THESIS

Study Mode: Full-time Major: Agricultural Economics Faculty: Advanced Education Program Office Batch: 2017-2021

Thai Nguyen, 27/11/2021

DOCUMENTATION PAGE WITH ABSTRACT

Thai Nguyen University of Agriculture and Forest

Thesis Title Estimate the technical efficiency and identify influencing

factors of tea farmers in Thai Nguyen city

Supervisor’s Signature

Abstract:

This study aims to estimate the technical efficiency levels and determine the factors correlated with the technical efficiency in tea-planting households in Thai Nguyen City of Vietnam To begin with, the output-oriented Data Envelopment Analysis method is employed to estimate the technical efficiency of the 90 households, residing in Tan Cuong, Phuc Xuan, and Phuc Triu in the year 2020 by using five input and one output variables According to the estimation results, the mean of technical efficiency under constant returns to scale, technical efficiency under variable returns to scale, and scale efficiency are 81.4%, 85.0%, and 95.7% respectively These figures reveal that the average sample tea-growing households may boost production by a certain amount by modifying the allocation of inputs more rationally Tobit regression is applied in the second stage to examine the influences of determinants factors on estimated technical efficiency The empirical findings of the study imply that the ethnic, irrigation, organic standard, gender, extension service, and co-operatives participant have a substantial impact on the technical efficiency of tea-planting households in three communes of Thai

Nguyen city

Planting, Thai Nguyen

Date of Submission

ACKNOWLEDGEMENT

First and foremost, I would like to extend my sincerest thanks to my supervisor Dr

Ho Van Bac, Faculty of Economics and Rural Development, Thai Nguyen University of Agriculture and Forestry Without his enthusiastic support, dedicated advice and continuous incentives, this work would hardly have been accomplished

I would like to convey my appreciation to the staff in the Thai Nguyen City Statistical Office Their guidance and assistance in the data collection stage paved the solid way for conducting this study

Finally yet importantly, I would like to express my heartfelt gratitude to my beloved family and my affectionate friends for believing in me You all are an intense motivation for any journey in my life

1 TABLE OF CONTENTS

LIST OF FIGURES 1

LIST OF TABLES 2

LIST OF ABBREVIATIONS 3

PART I INTRODUCTION 4

1.1 Research rationale 4

1.2 Research’s objectives 5

1.3 Research questions 5

1.3.1 Research subjects 5

1.3.2 Research scope 6

1.3.3 Research limitation 6

PART II LITERATURE REVIEW 7

2.1 Concept of efficiency 8

2.1.1 Technical efficiency 9

2.1.2 Allocative efficiency 10

2.1.3 Scale efficiency 10

2.2 Evaluation of efficiency 10

2.3 Identifying the factor affecting the technical efficiency 12

2.4 Review of studies related to technical efficiency in Vietnam 13

PART III METHODOLOGY 17

3.1 Description of the study area 17

3.2 Sampling and data collection 19

3.2.1 Sample size and sampling techniques 19

3.2.2 Data sources and method of data collection 19

3.3 Data analysis 20

3.4 Definition and measurement of variables 20

PART IV RESULTS AND DISCUSSION 22

4.1 Tea production in Thai Nguyen City 22

4.2 The reality of tea production in the studied area 24

4.2.1 The productive efficiency 24

4.2.2 Description statistic result 26

4.2.3 Distribution of technical efficiency 29

4.2.4 Factors affecting the technical efficiency 32

CONCLUSION 36

REFERENCES 37

APPENDICES 43

LIST OF FIGURES

Figure 1 Location map of study area 17

Figure 2 The total tea planted area and production in Thai Nguyen City 2013-2020 22

Figure 3 Total tea planted area by communes in Thai Nguyen City in 2020 23

Figure 4 Population and area of three study communes in 2020 24

Figure 5 Distribution of technical efficiency level under variable returns to scale 29

Figure 6 Distribution of technical efficiency level under constant returns to scale 30

Figure 7 Distribution of scale efficiency level 30

LIST OF TABLES

Table 1 Description of variables used 21

Table 2 Number of households and tea-planting households in three communes 23

Table 3 The total revenue, total cost and profit 25

Table 4 Summary statistics of tea-planting households 26

Table 5 Efficiency score summary statistics 28

Table 6 Returns to scale summary statistics 31

Table 7 Factors influencing technical efficiency 32

LIST OF ABBREVIATIONS

VietGAP Vietnamese Good Agricultural Practices

PART I INTRODUCTION

Tea has been regarded as one of the emphasized traditional cultures in Vietnam for centuries (Yanling & Montira, 2019) The elderly have a habit of consuming tea as a regular drink and serving green tea to their beloved guests, whereas a cup of hot tea is an indispensable beverage for the blue-collar labor force every early morning Nowadays, tea

is processed into a variety of goods and products, which makes tea accessible to everyone regardless of their age group, educational level or social status For many centuries, tea has been a product of daily consumption for many consumers, irrespective of location (Wenner, 2011) To fulfill the colossal demand for tea, Vietnam spends 123,000 ha on tea cultivation, with a production of 1019.9 thousand tons in 2019, according to the Annual statistical report of the Vietnam General Statistics Office (GSO, 2019) Vietnam is the fifth largest tea exporter in the world, and has plenty of tea-growing regions The most popular ones are concentrated in the Northern provinces such as Thai Nguyen, Tuyen Quang, Son La and Lai Chau and in central highlands like Lam Dong Based on the Thai Nguyen Statistical Yearbook in 2019, Thai Nguyen has the biggest cultivated land for tea

in Viet Nam with 22,282 ha planted tea, of which 1545 hectares belongs to Thai Nguyen City Located in the west corner of Thai Nguyen city, Tan Cuong commune, Phuc Xuan commune and Phuc Triu commune have a large proportion of tea growing and processing farmers The daily life and income of the local people are highly dependent on tea However, the productivity of tea production amongst tea farmers is heterogeneous, partly due to the techniques using inputs Therefore, the local farmers have to sit down under

precarious living and face tremendous difficulties Starting from that reality, I conducted this research to estimate the technical efficiency of tea-planting households and identify the factors affecting the technical efficiency in the current time to find out advice to help people reduce input costs and improve labor productivity in order to achieve the ultimate goal of profit maximization

1.2 Research’s objectives

In general, estimation the technical efficiency and analyzing the factors influencing the tea process is to give suggestions of improving the efficiency and productivity amongst tea-growing households in the Thai Nguyen City

Specifically, conducting this research is to pursue the following targets:

− To overview the status of tea production in the study area;

− To estimate the level of technical efficiency of tea-planting households and the average technical efficiency;

− To identify the factors that influence technical efficiency in tea production

1.3 Research questions

1.3.1 Research subjects

The coverage area is limited to three communes, namely Tan Cuong, Phuc Xuan and Phuc Triu in Thai Nguyen City Households in the study area produce many types of crops such as paddy rice, maize and cassava However, tea is the dominant crop in the known yield in the study area Therefore, this study focuses solely on tea production in terms of technical efficiency Due to the limitation of capacity, budget and duration, the study conducts a case study of 90 households in three abovementioned communes

1.3.2 Research scope

In terms of spatial scope, the topic focuses on conducting research on tea production situations of 90 random households in three communes, namely, Tan Cuong, Phuc Xuan and Phuc Triu

In terms of time scale, the topic focuses on the tea production in 2020 of certain households in three communes, Thai Nguyen city

1.3.3 Research limitation

Due to time and resource constraints, the coverage is limited to three communes in Thai Nguyen City with 90 sample tea-planting households The measure used to collect data is not entirely reasonable; the unit of measure of collected data is mainly attributed to monetary values and binary variables In addition, the results of cross-sectional data provided only a snapshot for technical efficiency estimation at a given point in time and

do not show the changes over time for comparison purposes Further researches should expand the scale and make use of the most proper unit of measurement in order to achieve the superlative accomplishments

PART II LITERATURE REVIEW

Tea is a perennial industrial plant with high economic value It plays a prominent role in daily life, national economics as well as cultural specificity In Vietnam, tea products are not only an indispensable product for domestic consumption but also a potential export item for gathering foreign currencies (Khoi et al., 2015) Currently, Vietnamese tea has been shipped overseas to more than 100 countries and territories with

an average export value of 200 million US dollars (MARD, 2018) For local citizens, tea brings remarkable benefits such as job creation, stable income, and socio-economic life improvement The tea sector ensures steady employments for around 400,000 small households in the rural areas and offers 1.5 million jobs for secondary parts of the tea value chain (GSO, 2011)

Vietnamese tea, meanwhile, has to confront a variety of challenges such as price fluctuation, natural disasters, and productivity, which attract many researchers to work together to address the problem and offer constructive advice to policymakers (Hong & Yabe, 2017) As a result, there has been a great deal of researches and studies done on such kind of crop The tea value chain has been investigated to provide a unique role and outstanding benefits for each actor in each stage of the value chain in order to encourage the position of each actor in the whole value chain (Khoi et al., 2015) The authors identified the factor affecting farmer’s decisions to adopt the feasible type of tea varieties

in Vietnam (Nguyen et al., 2017) Narrow down the scope in Thai Nguyen province, where the tea has been cultivated and processed for hundreds of years (Luyen et al., 2014) There has been a study on applying the Stochastic Production Frontier Model to

determine the level affecting of some certain factors in tea production and using Benefit Analysis to evaluate the process of transition traditional tea into organic tea production in Thai Nguyen (Tran, 2009) In the terms of productivity and efficiency, Ha

Cost-& Tho (2018) researched and analyzed the economic efficiency of household tea plants in Thai Nguyen province through the figure of Gross Output, Intermediation Cost, Mixed-Income, Net Present Value, Internal Rate of Returns and Benefit-Cost Ratio Hong & Yabe (2015) applied the Translog Stochastic frontier model to estimate the resource use efficiency of 243 tea-planting households in Vietnam, particularly in Thai Nguyen province

2.1 Concept of efficiency

To aim for the ultimate goal of profit maximization in agriculture, farmers and producers often first expect to increase efficiency in production There are two common ways to improve productive efficiency: increasing the number of resources inputs and enhancing the productivity of the current inputs For the latter technique, the problem revolves around the proper allocation of resources In the agricultural realm, there is a simple and common way of measuring efficiency by dividing the outputs produced by the area This productivity score uses a single input, whilst farmers and producers have to combine multiple inputs such as land, water, fertilizer, labor, and machines to produce a product, thereby; this score cannot fully generalize the productivity of the product As the result, originating from Farrell’s endeavor (1957) of establishing an index to measure the efficiency of the input factors, economic efficiency arose The theory of this efficiency has been developed for nearly seven decades and received tons of feedback and is popular

use by a wide range of researchers from all walks of life These efficiency scores may orient farmers to produce the obtainable amount of goods and services by utilizing inputs (Hong & Yabe, 2015) Abdulai & Huffman (2000) have been used economic efficiency score to adjust the rice production of farmers in Northern Gana and suggest some solution

to improve the productive efficiency Mediterranean fisheries (Maravelias & Tsitsika, 2008) applied this approach in the fishery industry in the Mediterranean to compare the capacity in the different segments and suggest the potential output level each purse-seiner can achieve at the given status Vietnamese coffee is another example of determining the economic efficiency and allocative efficiency between small farms and large farms in Daklak province (Rios & Shively, 2005) Specifically, economic efficiency has the three most popular components, with names such as technical, allocative, and scale efficiencies (Bojnec & Latruffe, 2007)

2.1.1 Technical efficiency

Some people decide to adopt new technologies to achieve higher productivity, whilst others endeavor to improve efficiency by using existing technology more efficiently at a given level of inputs Therefore, the technical efficiency is measured as the capacity of expanding the current output level under the assumption of fixed inputs (namely, output-oriented framework), otherwise, as the figure of potential input reduction that could not affect the output level (namely, input-oriented framework) (Farell, 1957; Coelli et al., 2005) It is argued by some researchers that the farmers should take full advantage of available technology for cultivation, which brings more cost-effectiveness, instead of technological innovation (Belbase et al., 1985; Ali & Flinn, 1989) In this study, the

output-oriented is more appropriate The author attempts to estimate the technical efficiency level in order to evaluate how tea-planting households perform without changing their use of input resources

2.1.2 Allocative efficiency

In contrast to physical quantity and technical, whenever the issues involved to prices, costs minimization, revenue maximization, profits maximization, or behavioral assumption, hence, it is possible to consider allocative efficiency While the technical efficiency tries to estimate the ability of farmers to get the maximum output goal by using the given and reachable inputs, the allocative efficiency in favor of measuring the ability

of farmers to allocate the amount of input with the optimal proportion and corresponding prices (Farell, 1957; Coelli et al., 2005) Through the index of allocative efficiency, farmers may make adjustments in their level of input to achieve output maximization

2.1.3 Scale efficiency

Scale efficiency is to evaluate the management practice According to Farrell (1957) and Kounetas & Tsekouras (2007) that a unit of scale efficiency is defined as the point of time when the study subject achieves simultaneously the maximum level of efficiency and

at the optimum scale size Scale efficiency is measured as a ratio of technical efficiency in constant returns to scale over the technical efficiency in variable returns to scale

2.2 Evaluation of efficiency

There exist two common approaches for estimating the level of technical efficiency The first is the parametric methods, which include the ordinary least square method

(OLS), Stochastic Frontier Analysis and Translog frontier production function (Hussain, 2014)

Data Envelopment Analysis is a non-parametric model based on the mathematical programming that is used to estimate the technical efficiency scores (Fare et al., 1994) This method can be easily applied to the objectives with multiple inputs These inputs and output can have extremely distinct measurement units, such as currency, time, person, kilogram, and hectares It means the DEA allows the use of variables without requiring any pre-determined weight, information on prices and a clear relationship between inputs and outputs (Kavoosi & Shahnazari, 2018; Dessale et al., 2017; Luik, 2020) However, this approach has some disadvantages in measurement error, the effect of noise, and misspecification

In DEA, two models, including CCR (Charnes et al., 1978) and BCC (Banker et al., 1984) were applied for evaluating the performance of different decision-making units The CCR model tends to favor an input-oriented situation in which the possibility of minimizing inputs while maintaining the same level of output is prioritized (Chauhan et al., 2006) Moreover, the assumption of the constant returns to scale illustrates that every multiple of input produces the same multiple of output The BCC model, meanwhile, represents a multidimensional production function, which is not hindered by the stability

of returns to scale It demonstrates three options for output, which include increasing, staying unchanged and decreasing returns to scale in the multiple inputs and single outputs situation In this study, the BBC approach was pursued, and the technical

efficiency at constant returns to scale, technical efficiency at variable returns to scale and scale efficiency will be interpreted in detail further

2.3 Identifying the factor affecting the technical efficiency

To determine the affects the technical efficiency, there are several certain approaches such as OLS, Tobit regression, Fractional Regression of Papke Wooldridge, Classification and Regression Free, of which, Tobit regression was widely applied by many prestigious authors The examples of this are the article of measuring the level of resource use efficiency of tea production in Vietnam (Hong & Yabe, 2015); the research

on the performance of rice farming households in Vietnam (Linh, 2012); and Li Dongpo

in his doctoral thesis about factors of agricultural development and production efficiency

in China (Li Dongpo, 2012)

Yoshimoto (2008) describes the standard Tobit model as follows:

The dependent variable y∗, which has a linear form, is determined by:

= 0 ∶ ∗ ∶ ∗∗≤ 0> 0 The bounded nature of efficiencies is between zero and unity Coincidentally, a two-limit Tobit regression takes account of the fact that its upper limit and lower limit on the scores are set at one and zero, respectively These two compatible properties make the Tobit function widely used to reflect the factors affecting technical efficiency The following function is the Tobit model for this study:

Where, yi= a latent variable representing the level of technical efficiency, which was

calculated using DEA method;

δ0= constant;

δi= parameters to be estimated each input;

Zij= explanatory variables;

ω= an error term

2.4 Review of studies related to technical efficiency in Vietnam

A variety of previous studies have been conducted on the technical efficiency of different crops in Vietnam

Son et al (1993) have performed an assessment of estimate the technical efficiency

of rubber in the Southeast region of Vietnam They derived a time-varying frontier production function by using linear programming technique to the date of 33 farms in five consecutive years from 1986 to 1990 and reported that 59% was the average technical efficiency in rubber production of those sample farms

The publication by Khai & Yabe (2011) relates to applying the Cobb-Douglas production function to the data of 3733 households in Vietnam and reported that the mean technical efficiency score of the sampled household was 81.6%, and individual scores were ranged from 16.5% to 98.5% He also applied Tobit regression to show the role of intensive labor, irrigation, and education in technical efficiency improvement

Kiet et al (2020) analyzed technical efficiency in three seasons of mango production

in Vietnam He estimated the stochastic frontier production function of Cobb-Douglas specification on the date of 1889 observations and reported that the average technical efficiency scores in three main seasons were 54.8%, 49.38%, and 55.28% The positive

determinants of technical efficiency and constraints of mango production were obtained

by maximizing likelihood estimation on the STATA15.0 program

The paper realized by Ly et al (2016), evaluated the performance of pig farm in Hung Yen Province at the level of 80.40% under the constants returns to scale for input-oriented Data Envelopment Analysis In the next step, the Tobit regression was used to identify the determinants of pig production, which were weight per fattening pig, breed time, experience, education, family member joining in pig production, income from pig, access to credit, and veterinary services

Using the DEA approach, Nguyen et al (2018) measured the technical efficiency of striped catfish The mean technical efficiency and scale efficiency of 84% and 80% correspondingly were used in the bootstrap truncated regression of the distance model to identify the factors that could improve the technical efficiency These were education, flooding and salinity intrusion

The article published by Duyen (2020), presents a non-parametric approach for examining the performance of black tiger shrimp farming in Gia Rai town She used data

of 44 intensive farming households in the year 2017 She applied the Data Envelopment Analysis technique to determine the technical efficiency levels of the household and Tobit regression was used to study the impact of households on the technical efficiency of shrimp farming She reported that the average technical efficiency and average scale efficiency of sampled farms were 59% and 79.95%, respectively, and three factors having positive impacts were education, seed density and testing

Nguyen et al (2020) studied the technical efficiency of oranges in 107 farm samples

in Tuyen Quang Province of Vietnam The input-oriented Data Envelopment Analysis approach under variable returns to scale was applied to find out the mean technical efficiency of 82% and scale efficiency of 94%

Le et al (2021) determined the efficiency of coconut production in Vietnam She used the data of 120 microenterprises and based on the output-oriented Data Envelopment Analysis technique She reported that the mean technical efficiency under constant returns

to scale, mean technical efficiency under variable returns to scale and scale efficiency of the sample observations was 58%, 68.4%, and 87%, respectively Tobit regression was applied in the stage of examining the influences of determinant factors on variable returns

to scale technical efficiency

The publication by Linh (2012), gave a comparison between the technical efficiency level by two approaches, that is the parametric approach of Stochastic Frontier Analysis and the non-parametric approach of Data Envelopment Analysis amongst 595 rice farm households throughout Vietnam The technical efficiency was revealed at the level of 63.4% for SFA and 78.5% for DEA The estimation results were applied to the Weight Tobit model accompanied with the information obtaining from bootstrap production to prove the positive influence of education, land holding, farm size and region factor Hong & Yabe (2015) have presented the first assessment of measuring the performance of farms in tea production in the Northern mountainous region of Vietnam They applied a Translog Stochastic production frontier model to determine the efficiency

in the resource use of 243 selected samples under two assumption of output-oriented and

input-oriented at 82.21% and 92.29% respectively In the next step, the Tobit model was used to identify the factors affecting these efficiencies amongst the tea farmers According to the result of the study, gender, soil and water conservation, farm value and extension were the factors that most influenced the technical efficiency

The previous studies, however, were conducted quite a long time ago, thereby becoming obsolete at present due to the variability of data Studies in the field of technical efficiency have not mentioned the participation in co-operatives and the application of agricultural standards, which are the paramount factors in the widely popular trend in Vietnam agriculture, namely, “One Commune, One Product” Moreover, the studies on efficiency and productivity in Thai Nguyen area are still limited For this reason, this study have been carried out by applying the previous method of using Data Envelopment Analysis and Tobit regression to bridge the abovementioned gap Have the tea households

in Thai Nguyen successfully cultivated using the available resources and technologies? What factors may influence their tea production? These answers can partly be found in this study

PART III METHODOLOGY

The study was carried out in Thai Nguyen province, which is situated in the Northern Midland and Mountainous Region of Viet Nam Thai Nguyen is border by Bac Kan province in the North, Tuyen Quang province and Vinh Phuc province in the West, Lang Son province and Bac Giang province in the East, and Ha Noi capital in the South The province consists of nine administrative units, including two cities, one town, and six districts with 32 wards and 139 communes

Figure 1 Location map of study area

Lying in the latitude of 21°35'39.19"N and the longitude of 105°50'53.41"E, with an altitude of about 600m above sea level, Thai Nguyen is a tropical province, whose weather is warm and humid The mean air temperatures at Thai Nguyen station remain

24.7-Celsius degrees with the total sunshine duration being 1389 hours in 2019 The average humidity stands at a high level of 80.8% while the total rainfall is 1889.3 millimeters in 2019 The distribution of rainfall is concentrated in summer commencing from April and ending in October According to CCE (2007), Thai Nguyen is covered by alluvial soil and ferralitic soil on limestone, which makes it more favorable for planting midland tea and new generation tea In fact, Thai Nguyen Statistical Yearbook (2019) had published that the total land area of Thai Nguyen province is 352,664 hectares of which 303,555 hectares belonged to agricultural land Thai Nguyen city has a total area of 22,

293 hectares and around 341.707 inhabitants living (TNSB, 2019) Thai Nguyen city encompasses three renowned tea regions, Tan Cuong, Phuc Triu, and Phuc Xuan in particular The total tea planted area in Thai Nguyen city is 1,545 hectares with 220,006 tons of fresh tea leaves production (TNSB, 2019)

The Thai Nguyen gross domestic product is 107.8 trillion VND, of which, the agricultural sector accounted for 10.3% of the total value Despite being an industry province, the agriculture sector still plays an important role in the provincial economy and receives enthusiastic attention and development encouragement from both government organizations and provincial institutions In terms of demographics, the average population in 2019 of Thai Nguyen province reached 1,290.9 thousand people, of which, 303.2 thousand people working in agriculture (occupied 39.6%)

3.2 Sampling and data collection

3.2.1 Sample size and sampling techniques

From personal experiences and suggestions from reliable people, Phuc Xuan, Phuc Triu, and Tan Cuong communes were selected from 139 communes in Thai Nguyen city

by purposive sampling technique Gathering information and data revealed the number of households and tea-planting households in three selected communes These figures were used to calculate the sample sizes based on the probability proportional of size sampling technique The sample size for this study was determined based on Yamane (1967) formula with a confidence level of 90% and a 10% level of precision According to the result of the following formula, the total sample size of the study was 97 households The

97 observations were picked from 3737 tea planting households via simple random techniques Seven observations were dropped after calculating the efficiency on the first test to reduce the duplication Therefore, this study was conducted with 90 observations

of tea-planting households

n = 1 + N(e )NWhere, n = sample size;

N = total number of tea planting households;

e = the acceptable sampling error

3.2.2 Data sources and method of data collection

The primary data were gathered by the questionnaire through face-to-face interview The questionnaire was designed and prepared under the guidance of the supervisor and refer to the agricultural questionnaire form of the Thai Nguyen Statistical Office In the