Efficiency of Ethiopian Garment Factories. Evidence from Addis Ababa

TABLES OF CONTENETS

ACKNOWLEDGMENTS

LIST OF TABLES

LIST OF FIGURES

LIST OF APPENDICES

LIST OF ACRONYMS/ABBREVIATIONS

ABSTRACT

CHAPTER ONE Introduction
1.1. Background
1.2. Problem Statement
1.3. Objective
1.3.1. General Objectives
1.3.2. Specific Objectives
1.3.3. Research question
1.4. Research hypotheses
1.5. Significance of the study
1.6. Scope and limitation of the study
1.7. Organization of the Paper

CHAPTER TWO LITERATURE REVIEW
2.1. Theoretical review
2.2. Empirical review

CHAPTER THREE Methodology, Model Specification and Description of variables
3.1. Sources of Data
3.2. Sampling design
3.3. Description of Variables
3.3. Specification of Econometric Model
3.3.1. Parametric approach
3.3.2. Non Parametric approach
3.4. Identifying sources of technical inefficiency
3.5. Conceptual framework garment factories technical efficiency

CHAPTER FOUR Result and Discussion
4.1. Characteristics and production performance of garment factories in Ethiopia
4.2. The technical efficiency level of garment factories
4.3. Comparison of technical efficiency level by market orientation
4.3.1. Comparison of DEA and SFA result by owner’s citizenship
4.4. Econometric analysis results
4.4.1. Hypothesis testing and Model selection
4.5. Determinates of technical efficiencies

CHAPTER FIVE Conclusion and Recommendation
5.1. Conclusions
5.2. Recommendations

Biblography

APPENDICES

ACKNOWLEDGMENTS

In the first place, I would like to thank my Almighty GOD for his support and blessing in my life.

Next, I would like to express my gratitude to my advisor Dr. Solomon Tsehay for his valuable assistance and constructive advice in preparing this thesis.

I would like express my heart full thanks to my family (Ababa and Eyewa) for their continuous support and encouragement in all aspect my life.

Finally, I would like to acknowledge my deep gratitude to my best friends (metesha and dagi) for their support.

LIST OF TABLES

Table,4.1 Descriptive Statistics of large and medium scale garment factories

Table 4.2 summaries of major variables of large and medium scale garment factories

Table 4.3 Descriptive Statistics of small scale garment factories

Table 4.4 Summaries statistics of main variable of small scale garment factories

Table 4.5 Summarize efficiency score for large and medium scale garment factories

Table 4.6 Summarize DEA and SFA efficiency score for small scale garment factories

Table 4.7 Efficiency of export oriented large and medium scale garment factories

Table 4.8 Efficiency of local market oriented large and medium scale garment factories

Table 4.9 DEA and SFA mean technical efficiency by owner’s citizenship

Table 4.10 DEA SFA efficiency correlation

Table 11 t-test resul

Table 4.12 Estimation result for the Cobb_douglas and Translog for the frontier model

Table 4.13 Technical efficiency effect model estimation

LIST OF FIGURES

Figure 1, Technical and Allocative Efficiencies Input-Oriented measurement

Figure 2, Technical and Allocative Efficiencies Output-Oriented measurement

Figure 3, Mean technical efficiency by market orientation

Figure 5, comparisons of SFA and DEA efficiency score by size

Figure 4, DEA and SFA mean technical efficiency by owner’s citizenship

LIST OF APPENDICES

Appendix 1: DEA efficiency score for large and medium scale garment factories

Appendix 2: DEA and SFA efficiency score of small scale garment factories

Figure 3, SFA technical efficiency of based large and medium scale garment factories

Figure,4, DEA Technical efficiency of small scale garment factories

Appendix 4: DEA and SFA efficiency score of export oriented garment factories

Appendix 5: DEA and SFA efficiency score of local oriented large and medium garment Factories

Appendix 6; DEA and SFA Efficiency score of large and medium scale garment factories owned by foreigners

Appendix 7: DEA Efficiency score of large and medium scale garment factories owned by an Ethiopian

Appendix 8 Enumeration Questionnaire

LIST OF ACRONYMS/ABBREVIATIONS

illustration not visible in this excerpt

ABSTRACT

The study measures the level of technical efficiency and its determinants in Ethiopian garment factories. The study employs both Data Envelopment Analysis (DEA) and Stochastic Frontier Analysis (SFA) to compute the technical efficiency of Ethiopian garment factories and Tobit model to examine the determinant of technical inefficiency of the garment factories based on a date set of ninety four garment factories over the period of 2014-2015. The study used Akaike Information Criteria (AIC) to opt for the best function between Cobb-Douglas and Translog functions. The result reveals that Cobb-Douglas production functions better explains the production behavior of garment factories. The DEA estimation shows that the mean technical efficiency of garment factories was around to be 0.43 while such figure in SFA goes as high as 0.89 and the research indicated that export do not promote technical efficiency Ethiopian garment factories. The research indicated that DEA is more appropriate for small garment factors while SFA is more suitable to large and medium garment factories. The result from DEA model indicates that educational level of production manager, local fabric sourcing, collaboration with similar factories and membership status with Ethiopian textile and garment institute have positive contribution to efficiency. On the other hand wage rate of semi/unskilled labour, years of establishments, number of production line, educational level of the manager and size have a negative effect to efficiency. While result from SFA indicates that wage rate of semi/unskilled, year of establishment, size and educational level of the manger have the positive contribution to technical efficiency. On the other hand owners citizenship and collaboration work with similar factories are a negative contribution to SFA technical efficiency. The research recommends that output can be increased by improving the technical efficiency of Ethiopian garment factories. To do so, for large and medium scale firms attention should be given to improve the wage rate of semi/unskilled labour, educational level of the managers and minimize collaboration with similar firms while educational level of the production manager, local fabric sourcing, collaboration with similar factories, and membership status of textile and garment institute should be improved for small firms.

Key words: garment factories, stochastic frontier analysis, data envelopment analysis, technical efficiency

CHAPTER ONE Introduction

1.1. Background

The manufacturing sector in Ethiopia is at its infancy in comparison with the agriculture and service sectors. Following the formulation of the national industrial policy the industry in general and manufacturing sector in particular has been given appropriate national importance. The policy has identified a priority sector that gives prime attention to build the required stage for the industry to play its key role in the economy of the nation. Textile and garment sector is treated as a number one identified sectors by the government industrial development strategy. The government to promote industry created a variety of support program example, duty free incentive and tax exemption (investment proclamation 769/2012 and investment regulation 270/202). The sector contributes 13% (in 2013) and 15.4% (in 2014) to GDP of the economy¹.

Textile and garment industry are mainly targeted to be export oriented and become competitive at international market(GTP, 2010). The reasons for such view is because Ethiopia has a suitable agro climatic condition for the main raw material of the sector, abundant and relatively lower cost of labour, international textile and garment production and consumption shift to least developed countries, and the availability of international market privilege through free trade agreement(AGOA, EBA, COMESA). Some of the major export destinations of textile products are Germany, Turkey, USA, Italy, Belgium, China, and united Arab emirate². Next to food processing and beverage industry, and leather industry textile and garment industry is the third largest manufacturing industry in Ethiopia (Negede. et al 2011)

Historically, the first textile industry was established in Dire Dawa in 1939 and the first garment factory was Agusta garment currently called Addis Garments, was established in the 1960s (Negede. et al 2011). In Ethiopia there are 189 large and medium scale garment factories out of which 52 factories are located in Addis Ababa. Currently there are 41 functional large and medium scale garment factories in Addis Ababa (TIDI, 2015). Practically, no firms use only a single factor of production to a single input. Machine, labour, raw material, energy, and other inputs are required in garment factory to produce a garment. Getnet and Admit (2005) Ethiopian textile and garment industry uncompetitive in domestic and international market. Low levels of productivity and under capacity utilization are the cause for un-competitiveness (Jemal, 2008).This research tried to measuring and identifies the factor that affect the technical efficiency of the sector.

1.2. Problem Statement

Efficient allocation of resources ensures firms to be competitive to the international market. Many developing countries mainly south and East Asian countries such as Bangladesh, China and Pakistan export textile and garment products to the main international markets (Kerry McNamara, 2008). Alderin (2014) showed that the textile industry is considered to be one of the first steps into industrialization and give opportunities for employment and increase the possibility for global trading (Gereffi, 1999).

The Ethiopian government envisions transforming the economy by giving special emphasis to priority industries including textile and garment industry (MOFED, 2010). The justification behind such action of the government emanates from the fact that the sector is labor intensive and could engender the country to secure comparative advantage intensive. Ethiopia has high labour force of young population; lower labour and energy cost compare to other countries therefore this give the country a comparative advantage over countries in becoming a competitive textile industry in global market (Alderin, 2014).

Ethiopia has become very active in producing textile and garment products for its local and foreign customers. These industries are expected to have effective and efficient manufacturing processes in order to success in global market competition. At the end of the plan period earns USD 21.8 million from exporting textile and garment products but the targeted is to generate USD 500 million. There is a wide gap between the actual and targeted revenue generated from export of textile and garment product. One of the main reasons for the gap is under performance of textiles and garment sector which is associated to the prevalence of poor level of efficiency among garment factories (GDP, 2010). Melaku (2013) showed that low productivity; inefficiency and lack of finance are the major factors which limited the contribution of the sector to employment, production and export earnings. According to Alderin (2014), inexperienced labour and production structure is the main challenge of inefficient production for Ethiopian textile industry. In domestic and international market the competitiveness of Ethiopian textile and garment industry has not been satisfactory. According to the study conducted by Jemal,(2008) that the textile and garment industry of Ethiopia are uncompetitive both in domestic and international market as measured by unit cost ratio.

Farrell (1957) efficiency means producing maximum level of output from the given set of factor of production. Measuring the technical efficiency of an industry is important to know the gap between the actual and maximum possible output level and the amount expected to increase its output without increasing input. To estimate and calculate technical efficiency there are different techniques. These techniques can be broadly categorized in to parametric and non-parametric approach. Both techniques have their own advantage and disadvantage.

Efficiency studies in garment industries have been conducted for different countries for example Joshi & Singh(2009) and Sudeshna(2015) for India, Vixathep and Matsunaga(2012) for Cambodia and Walujadi(2004) for Indonesia but for Ethiopia they are still very limited. Up to my knowledge, to date, there are no published studies particularly analyzing the technical efficiency of Ethiopian garment factories.

Among the studies available on industrial sector efficiency in Ethiopia, Jemal( 2008) conducted a study to analyze the productivity and competitiveness of Ethiopian textile and garment subsector. Melaku(2013) conducts a study assesses the total productivity and technical efficiency in the Ethiopian manufacturing sector over the period 1996-2008.

Therefore, this thesis aims to analyze the technical efficiency of Ethiopian garment factories by using both parametric and non parametric approaches.

1.3. Objective

1.3.1. General Objectives

The general objective of this study is to assess the technical efficiency of Ethiopian garment factories evidence from Addis Ababa.

1.3.2. Specific Objectives

The specific objectives of this study are;

- To characterize the production performance of the garment factories in Ethiopia.
- To measure the technical efficiency of garment factories.
- To compare and contrast the technical efficiency of export oriented and local oriented garment factories.
- To identify factors affecting the technical efficiency of garment factories in Ethiopia.

1.3.3. Research question

To achieve these objectives, this research investigates the following questions:

- Are our garment factories efficient?
- Do exports promote technical efficiency of Ethiopian garment factories?
- What are the determinant factors for inefficiency?

1.4. Research hypotheses

On the basis of the literature reviewed, the following null hypotheses were formulated for the above objectives so as to guide the study:-

- Hypothesized; Ethiopian garment factories are technically efficient. Thus the study tries to test by formulating the null hypothesis that the existence of no technical inefficiency.
- The study also tries to test by formulating the null hypothesis the existence of no difference between local oriented and export oriented garment factories

1.5. Significance of the study

The study would help the investors that invested in garment industry, new entrants in the business and policymakers to understand the technical efficiency trends of garment industries and factors affecting their inefficiencies. The study helps the garment factories to identify their drawbacks and which variables affecting the efficiency. It would enable to find remedies for industrial inefficiencies as well as policy measures with the proper utilization of the scarce resource of the country. It could also used to other interested researchers for reference.

1.6. Scope and limitation of the study

The scope of the study is assessing the efficiency of garment factories in Ethiopia. This study focused on large and medium and small scale garment industries engaged in export and domestic market. However, the study does not focus on allocative and economic efficiency but the analyses technical efficiencies of garment factories. In addition, due to lack of data that shows the exact number of small scale garment factories the sample selection of small scale garment factories do not supported by sample size selection formula.

1.7. Organization of the Paper

The rest of the paper is outlined as follows. The second chapter deals with review of related literature (both theoretical and empirical). In the third chapter data source, methodology and definition of variables are described. The fourth chapter discusses the descriptive and econometric results. And the last chapters provide conclusions and recommendation.

CHAPTER TWO LITERATURE REVIEW

In this section both theoretical as well as empirical literatures are summarized. The first section portrays the theoretical part while the second part features the empirical reviews of the research.

2.1. Theoretical review

The term productivity and efficiency often used interchangeably but this is unfortunate. Because they are not the same thing, productivity is the ratio of output that produce to input on the other hand efficiency is the ratio of actual productivity to highest possible productivity of the firm (Coelli et al, 1998).

Farrell was the first to analyze the efficiency of the firm; he proposed that the efficiency of a firm consists of two components to measure total economic efficiency. These components are technical efficiency and allocative efficiency. Technical efficiency is defined as from a given set if input the firm’s ability to produce maximum output level and allocative efficiency is the firm’s ability to use optimal proportion of factor of production with given their respective price and the production technology (Farrell, 1957).

The firms must be technically efficient in order to be economically efficient because technical efficiency is one part of economic efficiency. Technical efficiency has an output orientation and input orientation techniques. To maximize profit of firms the one requirement is within a given level of input produce the maximum level of output. (Harold et al, 1993)

There are two most widely used method of measuring inefficiencies which are input orientated and output orientated. Input orientated measures focused on minimizing the factor of production without reducing the output produced. (Farrell, 1957), illustrated the concept of efficiency graphically using a single example of the two inputs (x 1, x 2 ) to produce a single output q under the assumption constant return to scale. In Figure 1, below, the horizontal axis represent the ratio input one (x 1) to output and the vertical axis represent ratio of input two (x 2) to output. P, is represent the input of two factors to produce a unit of output, the distance QP represents the technical inefficiency of the firm, which is without the reduction of output the quantity of all input proportionally reduced. SS ', Represents perfectly efficient of the firms which shows the various combinations of the two inputs to use to produce unit output. The technical efficiency

(TE)of a firm is most commonly measured by the ratio TE OQ / OP. The value of technical efficiency (TE) is between zero and one and hence, provides an indicator of the degree of technical efficiency of the firm. The value of technically efficiency is one implies that the firm is fully efficient. For example, the point Q is technically efficient because it lies on the efficient isoquant.

illustration not visible in this excerpt

Figure, 1 Technical and Allocative Efficiencies Input-Oriented measurement Source: Timothy Coelli… et al (1998)

If the input price ratio represented by the slope of isocost line A ' A in Figure, 1 also the allocative and technical efficiency measures can be calculated using the isocost line. The allocative efficiency of firms at P is defined to the ratio; AE OR / OQ. The distance RQ represents the reductions in production costs that would occur if production were to occur at the allocative and technically efficient point Q ' instead of at technically efficient, but allocative inefficient, point Q. The total economic efficiency (EE) is given by OR / OP.

Output orientated measures evaluate the ability to produce the highest possible output within the given set of input. Farrell also illustrates by output oriented by using simple example involving in one input (x) and two outputs (q 1and q 2 ). In figure, 2 below the distance AB the technical inefficiency, this is the quantity output increase without requiring additional input.

illustration not visible in this excerpt

Figure, 2 Technical and Allocative Efficiencies Output-Oriented measurement Source: Timothy Coelli…et al (1998)

Figure, 2 shows that the output oriented measure of Technical efficiency is TE OA / OB.If we have price information then we can draw the isorevenue line D ' D and define the allocation efficiency measures as AE OB / OC. Revenue efficiency can be defined for any output price vector P represented by the line D ' D defined as RE OA / OC.

Francesco (2009) non parametric and parametric approach is the two main methodologies for measuring technical efficiency of the firms; parametric is econometric approach and non- parametric is mathematical approach. DEA is a non-parametric approach, based on linear programming, to estimating frontiers of technology sets. In DEA approach any deviations from the frontier is interpreted as inefficiency and in stochastic frontier analysis deviation from the frontier are not necessarily interpreted as inefficiency. SFA a parametric approach the deviation from the frontier are interpreted under two components, inefficiency and random component. Both methodologies have their advantage and disadvantage. The stochastic approach attempt to separate the noise effect and inefficiency effects, while the DEA approach both the noise and inefficiency effect is interpreted under inefficiency effect. The SFA approach is parametric and as a result suffers from functional form misspecification, while the DEA approach is non- parametric and so it is not sensitive to any form of functional misspecification

A panel data method gives the ability to contain more observation than cross sectional data. Because of this, in panel data technique the advantage to obtain better efficient estimators of unknown parameter and efficient predictor of technical efficiencies (Coelli et al, 1998). Panel data give the opportunity to clarify the different effect of inefficiency and noise get reliable prediction of technical efficiency, and investigate technical efficiency changes over time.

2.2. Empirical review

This section review empirical studies of the technical efficiency for the garment factories. The reviewed studies bring out in methodologies used to estimate technical efficiency and their findings.

Joshi & Singh(2009), studied measuring production efficiency of readymade garment firms located in Bangalore, India. The study used DEA approaches based on the primary data collected from eight readymade garment firms. To measure the efficiency they consider the number of stitching machines and number of operators as input variable and the number of pieces of garment produced as an output variable. They found, under the CRS assumptions, on average production (technical) efficiency in garment firms is 0.75, this indicate that with the existing level of input the firm could increase their output by 25%. Under the assumption of VRS they found the average pure production efficiency is 0.83 implying that an individual firm inefficient in managerial performance by 17%. The average scale efficiency is 0.91, which suggests that an average firm may have to correct its scale size by 9% to be scale efficient.

Mahmood(2007) investigates efficiency of large scale manufacturing in Pakistan using production frontier approach. He used the stochastic frontier analysis to estimate the technical efficiencies of various industries. The study used two cross section data (1995-1996) and (2000- 2001) and covers 101 industries. He found the average efficiency of the large scale manufacturing sector increased from 0.58 in 1995-1996 to 0.65 in 2000-2001. The average efficiency of wearing apparel also increased from 0.47 in 1995-1996 to o.56 in 2000-2001.

Vixathep and Matsunaga(2012) investigate firms’ efficiency and its determinants for Cambodia’s garment industry. The study used DEA model to calculate the technical efficiency score for firm and a regression model to evaluate determinates of firms efficiency. The study used two stage econometric approaches to determine the influence of exogenous variable on efficiency. They found the average PTE, OTE and SE is 0.558, 0.434 and 0.757, respectively.

Margono and Sharma(2004) assess the technical efficiency and total factor productivity growth in manufacturing industries during 1993 to 2000 in Indonesia by using stochastic frontier approaches. The study in addition analyzed the determinant of inefficiency. The result for the average technical efficiency of textile sector is 0.47889.

Walujadi(2004) focused on estimate industrial level of technical efficiency of medium and large scale garment firms in DKI Jakarta, Indonesia and analyzes the change in technical efficiency of this industry over the last 1990-1995. He used the stochastic frontier analysis to assess technical efficiency of garment firms based on panel data.

Among the studies available the study conducted by Melaku, (2013) the total productivity and technical efficiency in the Ethiopian manufacturing sector over the period 1996-2008. He uses stochastic frontier analysis to estimate the production of large and medium industry based on panel data. This study covered industrial group such as manufactures of food products, beverages, non-metallic mineral products, chemical and chemical products, rubber and plastic products, fabricated metal products(except machinery and equipment), textiles, publishing and printing, tanning and dressing, and wearing apparel.

The result show that the technical efficiency range from 88% in tanning and dressing industry which is relatively efficient sector to 10% in the food industry group with lowest efficiency score. The average technical efficiencies wearing apparel is 0.87.

The studies conducted by Jemal(2008), analyze the level of total factor productivity and competiveness of Ethiopian textile and garment industry. The study show that the average technical efficiency of Ethiopian textile and garment factories showed a declining trend. Within the study period 2001-2005 the average technical efficiencies decline from 0.92 to 0.72

CHAPTER THREE Methodology, Model Specification and Description of variables

3.1. Sources of Data

In order to measure technical efficiencies as well as analyze factors that affect technical efficiency of garment factories in Ethiopia, primary source of data was used. Based on the levels of garment factories produced by ministry of industry if the factories had 1.5milion capital, more than 100 workers and used power driven machineries this thesis consider as large scale garment factories. The factories had between 1-1.5million capital, 50-100 workers and used power driven machineries considered as medium scale garment factories and the factories had below 1millon capital and below 50 workers considered as small scale garment factories. Almost all export oriented garment factories are under medium or large scale level. Therefore this thesis considered large and medium scale garment factories as a one group and small scale garment factories on the other group. All data on technical efficiency were collected by using questionnaires. The data collected include all factors of production i.e. machine, labour, electricity and fuel used for production and output produced by garment factories. The determinants that suppose to explain technical inefficiency of garment factories were part of data gathered though questioners.

In order to conduct the study, various literatures have been surveyed from various sources like books, journals and thesis. Collected primary data through questioner from the factories have been organized, analyzed and presented.

3.2. Sampling design

In order to represent both small scale and large and medium scale garment factories used stratified random sampling techniques. To get the final sample size of large and medium scale garment factories for the survey the formula developed by Yamane (1967) used accordingly,

illustration not visible in this excerpt

Where n is the sample size of large and medium scale garment factories, N is the population size which is 41 large and medium scale garment factories, and e is the level of precision or sampling error which is equal to 10%. Therefore, the total sample size expected to be covered in this study was 29 large and medium scale garment firms. I disseminated the questionnaire to 30 large and medium scale garment factories, out of which 3 factories did not replay. Small scale industries are highly promoted in Ethiopia and there are many number of small scale industries. But there is challenge to get the exact number of small scale garment sub sectors. Hence, the sample selections for small scale garment factories are randomly select and disseminated questionnaire to 72 small scale garment factories out of which 5 factories did not replay. Therefore, data for the analysis of the study was collected through questioner from 27 large and medium scale and 67 small scale garment factories covering the period 2014-2015.

3.3. Description of Variables

The following variables were considered to estimate the efficiency scores and the inefficiency effects:

1. Output (Yit) ; Output is amounts of t-shirt and t-shirt equivalent garment produced and manufactured. Production is the result of the interplay of raw materials, fixed assets and other industrial inputs and it is relatively less affected by measurement errors when calculated at the firm level. Thus, total pieces of garment produced considering as an output variable (dependent variable).
2. Machine (M) : It represents the total number of power derive machine such as stitching machine, automatic fabric cutter machine and others.
3. Labour (L) : It represents the total number of skilled and semi/unskilled labours.
4. Electricity (El) : it measures the total consumption of electricity in the process of production in kilo watt.
5. Fuel (Fu) : represents consumption of fuel in the process of production in litter

3.3. Specification of Econometric Model

The data analysis part constitutes both descriptive and econometric techniques. To measure efficiency with frontier approach there are several techniques used. Data envelopment analysis (DEA) and stochastic frontier analysis (SFA) are most common used technique. DEA approach is non-parametric while SFA are parametric (Coelli eta al, 1998).

This paper will apply two different econometric methods in order to measure efficiency. Firstly, I will use DEA followed by regression of output on input and secondly, SFA methods, include the estimation of multi-dimensional cost minimizing model for inefficient effects. DEA measure the production of efficiency of garment firm located in Addis Ababa, Ethiopia. This technique using only observed output and input data of the firm and evaluate how efficiently the input are convert to output.

3.3.1. Parametric approach

To evaluate the performance of factories there are two most popular stochastic frontiers production function form such as Cobb-Douglas and general Translog production function. These two functional forms are widely used in stochastic frontier function analysis (Coelli, 1996). Cobb-Douglas function has been widely and extensively used in stochastic production analysis. It is simple and less flexible. Translog production function is less restricted (unrestricted) form allowing more flexible than Cobb-Douglas production function.

The socialistic frontier production frontier for the panel data model incorporate the usual stochastic error term which is exogenous to the system and the firm leave effect to distributer a truncated normal variable. Aigner, Lovell and Schmidt (1977), proposed the stochastic frontier production function can be specified as a Cobb-Douglas or Trans log function form.

The Cobb-Douglas production function of large and medium scale garment factories represent as:-

illustration not visible in this excerpt

Trans log production function of large and medium scale garment factories defined as:

illustration not visible in this excerpt

The Cobb-Douglas production function of small scale garment factories is defined as:

illustration not visible in this excerpt

Trans log production function of small scale garment factories defined as:

illustration not visible in this excerpt

Where: i =1, 2, 3k representing the selected factories, in my case k=94 garment factories. t =1, 2…T representing the time periods of 2014-2015

Yi; represents the outputs of the ith firm at time period; M is the number of power driven machine.

L is the number of all skilled and semi/unskilled labour.

El is the amount of electricity consumption in kilo watt.

Fu is the amount of fuel consumption in litter.

[Abbildung in dieser Leseprobe nicht enthalten] is a vector of unknown parameter.

ui is a non negative (one side) random variable associated with technical inefficiency.

Abbildung in dieser Leseprobe nicht enthalten

vi Are the systematic random errors measuring the positive and negative (two side) effect of

exogenous shock (statistical noise).

Abbildung in dieser Leseprobe nicht enthalten

[...]

---

¹ https://www.cia.gov/library/publications /the-world-factbook/geos/et.html

² https://www.cia.gov/library/publications /the-world-factbook/geos/et.html