Innovation and productivity in small and medium enterprises a case study of vietnam

To establish and quantify this relationship, this study employs the two-stage process: i the estimation of total factor productivity for each firm; and ii a determination of an innovatio

UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES

VIETNAM – THE NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

INNOVATION AND PRODUCTIVITY IN SMALL AND MEDIUM ENTERPRISES:

A CASE STUDY OF VIETNAM

By PHAM DO TUONG VY

MASTER OF ART IN DEVELOPMENT ECONOMICS

HCMC, NOVEMBER 2016

University of Economics International Institute of Social Study

VIETNAM – THE NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

INNOVATION AND PRODUCTIVITY IN SMALL AND MEDIUM ENTERPRISES:

A CASE STUDY OF VIETNAM

by Pham Do Tuong Vy

A Thesis Submitted in Partial Fulfilment of the Requirements for

the Degree of

Master of Art in Development Economics

Academic Supervisor: Dr Vo Hong Duc

HCMC, NOVEMBER 2016

ACKNOWLEDGEMENT

I would like to take this opportunity to express my deepest gratitude for the help, support and encouragement of the following people, who have contributed to the completion of this thesis in their very own ways

Above all, I would like to express my immeasurable appreciation to my supervisor – Dr Võ Hồng Đức for his precious time, support and advices to make this thesis completed

Furthermore, I would like to send my great thanks to all the lecturers and staffs

at the Vietnam – The Netherlands Programme for their knowledge and supports during my time joining in the program In specific, I am extremely grateful to Dr Phạm Khánh Nam and Dr Trương Đăng Thụy for their valuable guidance and support in the courses and thesis writing process

To all of my friends in Class 21 and my colleagues at TPF, I could never thankful enough for your encouragement and support until the very end of this thesis

Last but not least, my deepest thanks and love to my parents, who have always been beside me Without their unconditional love, none of this would have been possible.

ABBREVIATION

CDM Crépon, Duguet and Mairesse

GMM Generalized Method of Moments

GSO General Statistic Office

OLS Ordinary Least Squares

TFP Total Factor Productivity

ABSTRACT

The majority of enterprises in Vietnam is categorized as small and medium sized (SMEs) firms which play an important role to the sustainable growth of the Vietnamese economy As such, improving the productivity of the SMEs is essentially needed and this request becomes a crucial mission for the governments It is generally accepted that innovation and technology improvement are key drivers of productivity (Bartelsman & Doms, 2000) However, they have not been well-acknowledged by the SMEs in Vietnam even though their huge contribution to firm’s productivity is unarguable

This study aims to examine the relationship between innovation and productivity in the Small and Medium-sized Enterprises (SMEs) in Vietnam To establish and quantify this relationship, this study employs the two-stage process: (i) the estimation of total factor productivity for each firm; and (ii) a determination of an innovation – productivity relationship In the first stage, total factor productivity is estimated based on production function using the input and output approach However, firms might adjust their input level according to expected productivity shock As such, a potential endogeneity caused by possible relationship between input decision and productivity shocks (unobserved productivity shock) might exist To deal with this problem of endogeneity, an approach developed by Levinsohn and Petrin is applied to estimate firm’s total productivity In the second stage, the system-GMM approach is adopted to examine the relationship between innovation and productivity

An unbalanced panel dataset from five Small and Medium-sized Enterprises surveys from 2005 to 2013 is used in this study Findings from this study indicate that, in the context of Vietnam, when innovation is measured as innovation expenditure intensity and high-quality labor share in total firm’s labor force, innovation activities provide positive and significant impact on firm’s productivity

In addition, past value of firm’s productivity also has significant relationship with its current level This finding implies that higher (lower) level of current productivity

could lead to higher (lower) level of productivity in the future The study also provides empirical evidence to confirm that larger firms might perform better than the relatively smaller firms In contrast, capital structure provides negative impact on firm’s productivity However, this study fails to provide any evidence to support the view that longevity of firm does provide significant impact on productivity of firms

Key words: Vietnam SMEs; Total factor productivity; Productivity Shock;

Innovation, GMM

TABLE OF CONTENTS

CHAPTER 1 1

INTRODUCTION 1

1.1 Problem statement 1

1.2 Research objectives 2

1.3 Research questions 2

1.4 Research motivations 2

1.5 Research scope and data 3

1.6 The structure of this study 3

CHAPTER 2 5

LITERATURE REVIEW 5

2.1 Schumpeter Theory of Innovation – How does Innovation play its role in economic development? 5

2.2 Productivity: concept and measurements 7

2.1.1 Concept 7

2.1.2 Measurements 7

2.1.3 General productivity determinants 12

2.3 Innovation: concept and measurements 16

2.1.4 Concept 16

2.1.5 Measurements 17

2.4 How has the relationship between innovation and firms’ performance been analysed in the literature? 18

CHAPTER 3 23

RESEARCH METHODOLOGY 23

3.1 An overview of Vietnamese Small and Medium-sized Enterprises 23

3.1.1 Statistic overview 23

3.1.2 Difficulties 26

3.2 Methodology 27

3.1.3 Conceptual framework 27

3.1.4 Model identification 29

3.3 Research hypotheses and concept measurements 34

3.1.5 Research hypotheses 34

3.1.6 Concept and variable measurements 34

3.4 Data sources 36

CHAPTER 4 39

EMPIRICAL RESULTS 39

4.1 Total Factor Productivity of Vietnamese SMEs 39

4.1.1 Data descriptions 39

4.1.2 Total factor productivity from production function estimation of Vietnamese SMEs 42

4.2 Innovation – Firm’s productivity relationship 45

4.1.3 Data descriptions 45

4.1.4 The relationship between innovation expenditure intensity and firm’s productivity 46

4.1.5 The relationship between high-quality labor share in total firm’s labor force and their productivity 49

CHAPTER 5 52

CONCLUSION AND POLICY IMPLICATION 52

5.1 Conclusion remarks 52

5.2 Policy implications 54

5.3 Limitation and potential future research 54

REFERENCES 56

APPENDIX 1: Empirical studies on general productivity determinants 62

APPENDIX 2: Empirical studies on relationship between innovation and firm’s performance 65

APPENDIX 3: Durbin – Wu Hausman test for endogeneity 69

APPENDIX 4: Durbin – Wu Hausman test for endogeneity 71

LIST OF TABLES AND FIGURES

Table 3.1: Classification of SMEs in Vietnam

Table 3.2: Concepts and measurements of variables used in the study

Table 3.3: Number of observation in selected industries in dataset

Table 3.4: Number of observation after filtering

Table 3.5: Number of observation after filtering in stage 2

Table 4.1: Descriptive statistics of production function variables

Table 4.2: Comparison of OLS, Fixed Effect and LP estimators in Foods, Woods

and Rubber and Plastics Table 4.3: Comparison of OLS, Fixed Effect and LP estimators in Non-metallic

mineral, Fabricated metal and Furniture Table 4.4: Descriptive statistics of TFP and its determinants

Table 4.5: Regression results of innovation expenditure intensity and firm’s

productivity Table 4.6: Regression results of high-quality labor share in total labor force and

firm’s productivity Figure 3.1: Number of enterprises at 31/12 (by size of employees)

Figure 3.2: Conceptual framework

CHAPTER 1 INTRODUCTION

This chapter introduces the research topic and presents research objectives, research questions and motivation Research scope and data requirement also are discussed in this chapter

1.1 Problem statement

In line with Decree No 56/2009/ND-CP regarding assistances for the development of small and medium – sized enterprises (SMEs) in Vietnam, the SMEs defined as firms with total employee between 10 and 300, and total equity less than

100 billion dong Following these criterion, up to Mar 2015, total SMEs in Vietnam account for over 90% of all enterprises These firms have created more than half a billion of jobs every year These firms also contribute appropriately 40% overall GDP

SMEs play an important role to the sustainable growth of the economy Improving the productivity of SMEs is essentially and urgently needed and this need becomes a crucial mission of the Vietnamese Government because the growth of the economy depends significantly on the productivity of firms in the economy Key drivers of firm’s productivity are innovation and technology improvement (Bartelsman & Doms, 2000) However SMEs in Vietnam have still struggled in their operations and therefore lead to inefficiency One of the obstacles facing SMEs in Vietnam is the process of acknowledging the important role of creating innovation and applying new technology in production to increase their productivity Innovation has not attracted great attention from the SMEs themselves even though huge contribution to firms’ productivity is widely confirmed

The common measurement for innovation in empirical studies is R&D expenditure of a particular firm Various empirical studies have been conducted to quantify the relationship between R&D expenditure and firm’s performance

Conclusions vary from these studies including strong correlation between the two (Siedschlag, Zhang and Cahill (2010); Belderbos, Carree and Lokshin (2004); Crespi and Pianta (2009)) However, in the Vietnamese context, small and medium-sized firms have not widely reported their spending on research and development activities

In addition, innovation activities of SMEs is less formal and involved in many different exercises as compared with larger firms As such, research on the impact of innovation on SMEs productivity faces many obstacles in the Vietnamese context

A lack of interest in relation to the relationship between innovation and SMEs productivity in Vietnam has opened up the interest of deep investigation It is especially essential in the context of the economy dominated by SMEs and technology level is still low Therefore gaining further knowledge in this field is needed for policy makers to orient the development creation and application of innovative activities toward the growth of firms as well as country

1.2 Research objectives

This study is conducted to provide an additional evidence on the relationship between innovation and firms’ productivity for the Vietnamese SMEs The main

objective of this study can be summarized as Defining and quantifying the

relationship between innovation and productivity in firm level in the context of Vietnamese SMEs

1.3 Research questions

The study aims to provide empirical evidence for the main questions emerged:

Is there any relationship between innovation and productivity in the context of SMEs

in Vietnam? If yes, then how does innovation can affect SMEs productivity?

1.4 Research motivations

This study aims to provide the closer look at the Vietnamese SMEs’ productivity using the approach of Levinsohn and Petrin (2003), how it could be changed due to changes in the level of innovation making Despite the fact that innovation play a crucial role in the development, the outcome of innovation activities are uncertainty We do not know beforehand whether these activities would success

in creating value added to the firms Research results provide policy makers some evidences on how to appropriately allocate the available resources to obtain the target productivity This topic is interesting in the context of developing countries such as Vietnam for two reasons as suggested by Indjikian and Siegel (2005) Firstly the benefit of innovation might not be fully exploited in developing countries Secondly,

in these countries, national resources allocated to creating new innovation still are restricted despite the fact that innovation plays an important role in global growth

1.5 Research scope and data

The study aims to determine the relationship between innovation and productivity in Vietnam SMEs from 2005 to 2013 in six selected industries include: (1) foods; (2) wood and wood-related products; (3) rubber and plastic products; (4) non-metallic mineral products; (5) fabricated metal products and (6) furniture The reason why these six insuctries are selected in the study is that data of these industries

is biggest and have accounted for nearly 70% of total SMEs in the five-round survey, therefore can be representative for the whole dataset At the time data used in this study was collected, the dataset of 2015 survey was not fully gathered and published

As such, data used in this study only ends in 2013

1.6 The structure of this study

This study contains five chapters which can be presented as follow:

Chapter 2 provides theoretical and empirical studies on the relationship between innovation and productivity Chapter 2 begins with Schumpeter Theory of Innovation that explains the role of innovation to economic growth Then this chapter reviews the concept of productivity and the methods of how productivity can be estimated as well as its determinants In addition, the definition of innovation and how it is measured are discussed in the chapter The relationship between these two concepts has been reviewed through literature

Chapter 3 presents the methodology which is utilised in the study An overview

of Vietnam SMEs is discussed On the ground of literature review in Chapter 2, the conceptual framework is constructed The measurement of relevant variables and

regression techniques are described In addition, this section also includes the process

of how to filter data

Chapter 4 presents empirical results Statistical descriptive of data is presented

in this chapter Then, the findings on Vietnam SMEs’ productivity are described and discussed The results of regression in relation to the relationship between innovation and productivity are presented in this chapter

Chapter 5 provides the summary of the main results and proposes some policy implications based on the results described in Chapter 4 This Chapter also includes research limitation and suggests some further research direction in the future

CHAPTER 2 LITERATURE REVIEW

This chapter provides the literature review on the relationship between innovation and productivity At first, Schumpeter Theory of Innovation that explain the role of innovation in the economic growth is presented Then the concept, calculation and determinant of productivity is reviewed After that, this chapter presents the definition and measurement methods of concept innovation In the end

of the chapter, the relationship between innovation and productivity has been reviewed through empirical studies in the past

2.1 Schumpeter Theory of Innovation – How does Innovation play its

role in economic development?

Schumpeter was seen as a person who built very first basic foundation to the

theory of innovation and economic development In his famous book The Theory of

Economic Development (published first time in 1912), he has argued that the whole

economy has its own business cycle in which technological innovations play an important role When a new technology has been introduced and the economy is ready

to adapt then the economy would alter itself to fully employ the new technology and resulted in the upward trend of the business cycle If that new technology has been introduce at the time the economy is saturated and became more vulnerable to the any negative shock and easily get into depression then only new technology might not help out the whole economy Schumpeter also argued that firms should willing to take risk and invest in new technologies to take advantage of the profit at the early stage of these new technological innovations when the other firms haven’t applied

Together with The Theory of Economic Development (1934), Capitalism,

Socialism and Democracy (1942) and Business Cycles: A Theoretical, Historical and Statistical Analysis of a Capitalist Process (1939), Schumpeter has contributed to the

economics theories with the role of innovation and entrepreneurship in the economic development He believed that innovation is the core driver of development as well

as emphasized the role of entrepreneur of smoothing the mechanism in which revolutionarily technical changes occurs via innovation and push the economy out of its steady state

Schumpeter explained the development of the economy is mainly driven by innovation which he categorized into five types:

(i) launching new products, whether these are about improving a part of products or totally new to the market,

(ii) introducing new method of production,

(iii) opening new markets which have not entered in the past yet,

(iv) searching/discovering new sources/suppliers for raw material and other inputs in production process,

(v) acquiring new market structures in any industry (i.e changing the monopoly position)

How innovation could become driven to the growth of economy? According to Schumpeter, innovation can be expressed in a process of four steps: invention, innovation, diffusion and imitation (Schumpeter, 1942) in which the first two steps have less impact while the last two have much more influence on the economic growth His arguments relied on the vague of economic achievements in the early stage of innovation, after that economies would realize the potential of increasing sales or cost deduction when they come to the period of diffusion and imitation and

at that time they invest more in these innovation However purely ideas could not alone play the whole game, they need a power to implement these ideas At that necessarily, entrepreneur play important role of allocating the resources to the process

of replacing old technologies with new ones which Schumpeter labelled as creative

destruction In other words, Schumpeter explained the economic development

through the process of creative destruction driven by innovations

2.2 Productivity: concept and measurements

2.2.1 Concept

Productivity is the efficiency of the process in which firm, industry and country convert input factors in to output Therefore productivity is generally defined

as the ratio between output and inputs in the manufacturing process Productivity is

a good indicator to economic performance of firm, industry or country as a whole

There are two things could affect productivity: through the availability of input resources and through value adding to the products in producing process In a further details, firm’s productivity could be decreased in the circumstances of lacking inputs

or inputs were not used efficiently However through creating value added with available inputs and certain activities in manufacturing process helps to improve productivity

2.2.2 Measurements

There are many ways to measure productivity, but they could be classified into two groups: single factor productivity measures (in which productivity is the ratio of output over single input) and multifactor productivity/total factor productivity measures (a measure of output to several inputs)

In the group of single factor productivity, there are two ways to measure productivity: labor productivity and capital productivity In both ways, productivity has been expressed as quantity index of labor input/capital input over an index of gross output or value added These measures are easy to calculate but they only reflect the partial productivity of workers’ capacity or capital intensity, how efficiency they are in combine with other input factors in production process To have a better index

of productivity in which take into account contribution of more than an input, multifactor productivity/total factor productivity turns out to be more efficient measure Therefore in this research, total factor productivity has been used to estimate firms’ productivity

Estimating total factor productivity through Production function estimators have been regularly used to address many relevant issues in the literature: the

relationship between foreign direct investment and domestic firms’ productivity (Javorcik, 2004), impact of R&D (Hall et al., 2009), impact of information technology (Chun et al , (2015) These relationships are mostly estimated based on simple Cobb-Douglas production function regression

𝑌𝑗 = F(𝐴𝑗, 𝐾𝑗, 𝐿𝑗) = 𝐴𝑗𝐾𝑗𝛽𝑘𝐿𝑗𝛽𝑙 (1) Where 𝑌𝑗 represents firm j’s output, 𝐾𝑗 is physical capital stock, 𝐿𝑗 is labor input and 𝐴𝑗 denotes for firm’s level of efficiency, 𝛽𝑘 and 𝛽𝑙 are output elasticities with respect to capital and labor

Based on the definition of productivity above, 𝐴𝑗 is referred to Total Factor Productivity and could be derived by taking natural logs of (1):

𝑦𝑗𝑡 = 𝛽0+ 𝛽𝑘𝑘𝑗𝑡 + 𝛽𝑙𝑙𝑗𝑡 + 𝜀𝑗𝑡 (2)

Where t subscript denotes time series and lower case letters are represented

for log value In equation (2), Total Factor Productivity has two components: 𝛽0and 𝜀𝑗𝑡, in which 𝛽0 is average productivity for all firms across time and 𝜀𝑗𝑡 captures firm’s deviation productivity from that average caused by unobserved factors affect firms’ output outside of inputs 𝜀𝑗𝑡 then can be separated in two components: firm-level productivity 𝑤𝑗𝑡 and i.i.d component 𝑣𝑗𝑡:

𝑦𝑗𝑡 = 𝛽0+ 𝛽𝑘𝑘𝑗𝑡+ 𝛽𝑙𝑙𝑗𝑡 + 𝑤𝑗𝑡 + 𝑣𝑗𝑡 (3) Therefore researchers can get firm’s productivity from estimating (3) and solving for 𝑤𝑗𝑡:

𝑤̂ = 𝑦𝑗𝑡 𝑗𝑡 − 𝛽̂𝑘𝑘 𝑗𝑡− 𝛽̂ 𝑙𝑙 𝑗𝑡 (4)

Then, the exponential of 𝑤̂ is the result of firm-level productivity 𝑗𝑡

Mainly there are two trends of approach of research in how to calculate total factor productivity, non-parametric and parametric With non-parametric technique, growth accounting is the most used based on a paper of Robert Solow in 1957 about technical changes and production function Under the assumptions of constant return

to scales and competitive factor markets, growth accounting method expresses how

much changes in output growth can be explained by changes in different types of input and changes in total factor productivity Although growth accounting technique

is well–established and consistent, it cannot address the problem of causality, which

is investment in technological changes can be driven and resulted of productivity growth at the same time With parametric technique, econometric method has been applied to estimate total factor productivity in the relationship between production inputs and output (production function estimators) There are several benefits by using econometric techniques: the parameters can be check for the statistical significance; solving problem of endogeneity

2.2.2.1 Growth accounting method

Non - parametric growth accounting method was developed by Robert Solow

in his paper about the technical change through analyzing aggregate production function (Solow, 1957) Growth accounting approach aims to determine how much economic growth was due to contribution of inputs (growing by the movement along the production function) and how much growth was due to the improvement in technology (shift the production function) (Nelson, 1973) This approach has the assumption of constant return to scales, which means total elasticities of all input factors in production function equal one (from the equation (1), 𝛽𝑘 + 𝛽𝑙 = 1) Typically these input factors are weighted by their income shares (in case of calculating productivity at country level) (Cardona et al., 2013), or by their cost shares (when calculating firm’s productivity)

Productivity is calculated by solving equation (4) without econometric sense

𝑤̂ is called “Solow residual”, has positive value whenever the growth rate of output 𝑗𝑡rises faster than the growth rate of input factors “Solow residual” not only reflects growth by changes in technological progress but also other factors that affect the efficiency in general outside of input factors (Schreyer, 2001)

2.2.2.2 Production function estimator methods

As mentioned above, there could exist problem of endogeneity caused by possible relationship between input decision and productivity shocks (unobserved

productivity - 𝑤𝑗𝑡), which means firms might adjust their input level according to productivity shocks For example, firms tend to increase their investment if they observe a lucrative productivity shock, in another way, if an unfavorable shock occur, firms might reduce their level of workforce Therefore the result of input coefficients

in the OLS regression might be biased and inconsistent (Eberhardt and Helmers, 2010)

After the problem of endogeneity arises in production function estimation, there are several solutions have been developed and applied in the literature: Instrumental Variables (IV) regression; ‘dynamic panel estimators – developed by Arellano and Bond (1991) and Blundell and Bond (1998), commonly known as Generalized Method of Moment (GMM) approach and the works of Olley and Pakes (1996) which is categorized as ‘structural estimators’, then been further developed by Levinsohn and Petrin (2003)

In the standard IV regression, to generate the consistent and unbiased coefficients, independent variables that causing endogeneity (in this case is input quantities - K and L) need to be instrumented by variables that satisfy two conditions: these variables have relationship with input quantities, but are exogenous with unobserved productivity With the assumption of perfectly competitive input and

output markets, input prices (r, w) have been introduced as instruments for input

quantities However input prices seem not to be good instruments as summary by Eberhardt and Helmers (2010) and Van Beveren (2012) for the following reasons (i) Lack of information about input prices in most of dataset Even those information exist, they do not vary across firms enough to estimate the production function (ii) The assumption of perfectly competitive inputs market seems hard to be hold because

of the argument that productivity shocks might create market power for firms, then

in turns affect to input prices, causing the relationship between instrument variables and error term (iii) Even the perfectly competitive inputs market assumption is strictly hold, input prices might correlate with unobserved productivity in other ways That is the changes in ‘input price’ wages might be because of the unobserved labor quality, and this unobserved labor quality become a part of unobserved productivity, then wages could not act as valid instrument for labor input in production function

estimation Similar mechanism with rental rate and capital stock and unobserved productivity Because of the above reasons, standard IV regression using input prices

as instruments for input quantity could not yield consistent results

It seems hard to find a strong instrument for input quantity in production function regression to yield satisfactory results, Arellano and Bond (1991) and Blundell and Bond (1998) have contributed to the literature by proposing Generalized Method of Moment (GMM) estimator In this approach, past values of dependent and independent variables have been used as instruments to correct endogeneity problems These instruments also are valid with the argument that input choices

before time t are uncorrelated with productivity shock at time t, 𝑤𝑗𝑡

Although GMM approach was an efficient tool in addressing endogeneity problem and yield satisfy results, this approach are not constructed from structural model based on firm’s behaviors (Eberhardt and Helmers, 2010) Olley and Pakes (1996) (OP) has developed a new approach that explains firm’s production function using observed firm’s behaviors Put in the simpler explanation, OP solved the

problem of endogeneity in the production function by using observed firm’s

investment decision to proxy for unobserved productivity 𝑤𝑗𝑡 They have made two assumptions when building this approach, one is known as “monotonicity assumption”, in which stating that firm investment has strong positive relationship with capital stock and (unobserved) productivity, 𝑖𝑗𝑡(𝑘𝑗𝑡, 𝑤𝑗𝑡) and this relationship is continuous in 𝑘𝑗𝑡 and 𝑤𝑗𝑡 Therefore productivity 𝑤𝑗𝑡 can be determined by inverting investment function: 𝑤𝑗𝑡= 𝑓𝑡(𝑖𝑗𝑡, 𝑘𝑗𝑡) Labor is not included in this function because

it is assumed to be fully flexible that can proper alter immediately at the time of observing 𝑤𝑗𝑡 Another assumption is that only unobserved productivity 𝑤𝑗𝑡 is the factor that can affect firm’s investment decision This assumptions is called “scalar unobservable” condition OP “structural estimator” determined through two steps Details are below:

Firstly, from (3) output 𝑦𝑗𝑡 has been regressed on labor input 𝑙𝑗𝑡 and a proxy

of firm-specific productivity:

𝑦𝑗𝑡 = 𝛽𝑙𝑙𝑗𝑡 + 𝜑𝑡(𝑖𝑗𝑡, 𝑘𝑗𝑡) + 𝑣𝑗𝑡 (5)

where:

𝜑𝑡(𝑖𝑗𝑡, 𝑘𝑗𝑡) = 𝛽0+ 𝛽𝑘𝑘𝑗𝑡+ 𝑓𝑡(𝑖𝑗𝑡, 𝑘𝑗𝑡) (6) Equation (5) is in partially – linear form and 𝛽𝑙 is assumed to be exogenous with error term 𝑣𝑗𝑡 OP suggested a method based on third-order polynomial expansion to estimate equation (5) then get the unbiased and consistent result on labor coefficient 𝛽𝑙 and non-parametric part 𝜑𝑡(𝑖𝑗𝑡, 𝑘𝑗𝑡)

In the second step, OP has regressed 𝑦𝑗𝑡 - 𝛽𝑙𝑙𝑗𝑡 on 𝑘𝑗𝑡 and 𝜑𝑡(𝑖𝑗𝑡, 𝑘𝑗𝑡) then yield unbiased and consistent result for capital input coefficient 𝛽𝑘

“Structural estimator” has been further developed from OP method by Levinsohn and Petrin (2003) (LP) Instead of using investment as proxy for productivity, 𝑤𝑗𝑡, Levinsohn and Pertrin proposed to used intermediate inputs They have pointed out that intermediate inputs as proxy might be better satisfy with the

“monotonicity assumption” made in OP approach since the argument that firms with higher level of capital and productivity would consume more intermediate inputs is more reasonable than investment decision Furthermore data on intermediate inputs are generally available in most of firm level datasets while data on firm investment might either been missing or reported at nil value thus eliminate the situation of drop many observations in OP approach LP method still rely on two assumptions were made from OP (“monotonicity assumption” and “scalar unobservable”) and employed the similar procedure with OP to determine firm productivity

2.2.3 General productivity determinants

It is important to know about the sources as well as determinants of productivity From the definition of productivity, inputs of production (such as capital, labor, material, energy, etc.) are general well-known as direct factors affecting firm’s productivity In addition, there are other factors also have significant effect on productivity such as: firm age, firm size, ownership status, credit accessibility, export intensity These factors could be allocated into two groups: (i) exogenous factors including firm age, firm size, ownership status and (ii) endogenous factors: credit accessibility, export intensity

This section provides the empirical studies on how exogenous factors (firm age, firm size) and endogenous factors (credit accessibility) could affect firm productivity Because the research scope is limited in small and medium-sized firms which mostly operating as private enterprises therefore the ownership might have less impact on firm performance in Vietnamese context despite the fact that ownership does matter for productivity as confirmed by Cucculelli et al (2014), Margaritis and Psillaki (2010) and Kim (2006) Likewise, not many small and medium-sized firms

in Vietnam have international transactions so export intensity might be not the sources of productivity differentials in Vietnamese SMEs

by applying Levinsohn and Petrin (2003) approach In the next stage, other variables (such as firm age, firm size, family-managed status, ownership concentration, capital intensity) are included in the regression of total factor productivity obtained from first stage to examine the impact of these factors on firm productivity They have concluded that family-managed firms are less productive than non-family-managed firms, and this relationship is significantly and robust In addition, they have found the evidence for the increasing relationship between firm age and family-managed firm productivity, but no relationship between age and non-family firm productivity Huergo and Jaumandreu (2004) studied on over 2,300 Spanish manufacturing firms from 1990 to 1998 and found that firms at early stage of operation enjoy high productivity growth (at 5%), and this rate decreases continuously for 8 years until equal the average productivity (at 2%)

In term of firm size, Dhawan (2001) when doing the study for US firms for the period of 1970 to 1989 found that small firms are more productive than large firms The positive relationship between firm size and their productivity is then confirmed by Cucculelli et al (2014) and Margaritis and Psillaki (2010) Tovar et al (2011) have analysed data of 17 Brazilian electricity distribution firms from 1998 to

2005 to examine the effect of firm size on the productivity of Brazilian Electricity Industry They determined productivity by using Stochastic Frontier Analysis and

then TPF is decomposed into three components: (i) pure technical efficiency change, (ii) scale efficiency change and (iii) technical change They argued that through scale

efficiency change and technical change, different in firm size can explain productivity

differential Because firm size is proved to have positive impact on TFP, they suggested that mergers of small electricity distribution firms could lead to gain in productivity

2.2.3.2 Endogenous factors

Factors that classified as endogenous because they are related to firm decision that affect their productivity and lead to the different in those among firms Endogenous factors reviewed in this section are firm capital structure and international trade decision

The relationship between firm capital structure and productivity has been analysed carefully in both theories and empirical studies The theory of agency suggests the negative effect of debt on firm performance Agency theory assumes that there are conflicts between owners and managers, both who are motivated by self-interest, these conflicts lead to agency cost of equity (Jensen and Meckling, 1976) The agency cost could come from default risk in which firms are under the pressure

of paying high interest rate As a result, firms with higher debt ratio appear to perform

less efficient than firms with lower debt ratio (Myer, 1977) In another hand, free cash

flow theory suggests that free cash flow in companies should be paid out to investors

as it would prevent managers from using it improperly (Jenson, 1986) Therefore higher leverage induces firms to avoid misusing free cash flow and put firms under the pressure of generate more cash to service their debt In this case, debt has positive

influence on firm’ performance Margaritis and Psillaki (2010) using data from three industries in France: chemicals, computers and textiles from 2002 to 2005 have found that firm’s capital structure is positive correlated with their efficiency and this effect

is stronger with firms in chemicals and textiles industries They proposed two-stage approach to estimate this relationship First, Data Envelopment Analysis and distance function approach are applied to estimate firm efficiency In the next stage, firm’s efficiency obtained from first stage is regressed against leverage and other control variables (ownership status, profitability, asset structure, growth opportunities and size) using dynamic OLS estimation to examine the impact of leverage on firm’s efficiency In contrast, studying on Korean manufacturing firms, Kim (2006) found the significant negative effect of debt ratio on firm productivity However they also found the positive impact of debt ratio in Chaebols firms (which defined as family-managed, debt-dependent, diverse business activities firm)

In term of international trade decision, many studies have confirms that involving in international activities does impact on firm’s productivity and could be

a source of productivity Bernard and Jensen (1999) and Bernard and Wagner (1997) suggested two hypotheses about the reasons why productivity of exporters could higher than non-exporters The first hypothesis about self-selection mechanism in which higher productivity firms are more likely to export because of additional costs (transportation, distribution, marketing, human, etc.) that creating barriers to the export market for less productive firms In addition, firms desiring to export tend to improve their productivity themselves today in order to have the ability to export in the future The second hypothesis emphasized the important of learning-by-exporting Through exports, firms could learn about new process or technology from their importers or competitors, then they could improve their performance In addition, exporters have to compete against a lot more competitors in severe environment then they have to push themselves harder and enjoy higher productivity than firms which only trade domestically Aw, Roberts and Winston (2007) have confirmed the relationship between export behaviour and firm’s productivity: export experience have significant positive relationship with firm productivity

As mentioned before, for the reason of not many Vietnam SMEs involving in international trade, this study does not take international trade as an endogenous determinant of Vietnam SMEs productivity

Appendix 1 provides the summaries on the empirical studies about general determinants of productivity

2.3 Innovation: concept and measurements

2.3.1 Concept

Oslo Manual (OECD 2005, p.46) suggested the definition for innovation “is the implementation of a new or significantly improved product (good or service), or process, a new marketing method in business practices, workplace organization or external relations” This definition has been widely used in studies and referred by many institutions when conducting surveys on innovation aspects

There are four type of innovations which proposed in Oslo Manual (OECD 2005):

(i) Product innovation in which new goods or services are introduced to the

market and their characteristics are much improved in compared with current goods or services in term of appearance, technical identification,

new function, or user friendliness, etc

(ii) Process innovation implies improvement in the way of how goods and

services are produced and distributed which help to reduce costs or raise product’s quality Process innovation directly relates to technique applied, supporting equipment and software in the process of producing goods and services

(iii) Marketing innovation refers to improvements in product design, pricing

determination, promotion campaign, etc through new marketing method objecting to best suit to the client need, occupy the market and increase sales

(iv) Organizational innovation refers to changes in organizational structure,

business environment, etc that helping in reducing administrative cost,

transaction cost, improving work efficiency In addition, organizational innovation not only involves internal activities but also external relations improvements (with suppliers, clients, state agencies, etc.)

It is necessary to clearly distinguish among these types of innovation because there are many innovations which their characteristics can be belong to more than one type of innovation For instant, introducing new product which it needs a new

production process to be produced, therefore it can be categorized both product

innovation and process innovation Oslo Manual (2005) provided detail guidelines

on how to distinguish these types of innovation

2.3.2 Measurements

Different types of innovation can be easily represented by dummy variables which applied by many studies such as Hall, Lotti, & Mairesse (2008), Griffith, Huergo, Harrison, & Mairesse (2006), Mairesse, Mohnen, & Kremp (2005), Mairesse

& Robin (2009), Polder, Van Leeuwen et al (2009) However this measurement could not adequately reflect differential in innovation intensity among firms (Mohnen and Hall, 2013) It could generate misleading results when examine innovation activities across firms in different sizes that large firms could more innovative than small firms In fact, it is argued that large firms might involve in many activities thus these activities could fall into one of four innovation types Therefore innovation dummy variables might not suitable to come up with the conclusion that large firms are more innovative than small firms

Innovation can be measured in term of both input and output approaches Input approaches refer to firm’s effort to introduce new products, improve their production process, open new markets or raise firm’s efficiency In the output sides, innovation are reflected through new products introduced, successfully improved production process, costs deduction or gain in efficiency (Mohnen and Hall, 2013)

With the input approached, innovation is commonly measured by Research and Development (R&D) expenditures which involved in developing process of introducing new products or production methods However, there are many non-R&D activities that firms are involved in and considered as innovation Oslo Manual (2005)

has pointed down several non-R&D innovation activities that firms might have They can buy patents, pay royalties or scientific information then modify in compliance with their need They can improve their labor knowledge and skill via internal training They can buy new equipment, software or improve their facilities that affect innovative process They can improve their current management structure or creating new method to introduce products to the market Together with purely R&D activities, these above non-R&D activities also have the common objective of innovation to improve firm’s efficiency Therefore expenditures of these non-R&D activities should be taken into account in measuring innovation together with R&D expenses Another suggestion from Oslo Manual (2005) about measuring innovation with the input approach is by the skilled employees of firm As skilled employees is considered as key assets for innovation activities in which they help to facilitate the process of adopt new technologies, manage manufacturing operations and resolve the technological problems might occur

In the output side, innovation is generally measured based on its quality which can be represented through revenue The percentage revenue share of new products

or current products with added value through innovation process is a proper indicator for showing the results of innovation in overall firm’s performance This indicator has been used to proxy for innovation in many studies such as Miguel Benavente (2006), Jefferson et al (2006), Siedschlag et al (2010) Another measurement of innovation effect on firm’s performance is cost deduction due to process innovation (Peters, 2008)

2.4 How has the relationship between innovation and firms’

performance been analysed in the literature?

The relationship between innovation and firm’s performance has been well analysed in the literature However the conclusion about this relationship has not come to the consensus In one hand, innovation can positive influence firm’s performance, which can be measured by several indicators In another hand, many studies pointed out negative impact caused from innovation

How innovation affect firm’s performance? From innovation to firm’s performance is argued as process in which innovation inputs generate innovation output and then these output contributes to the overall firm’s performance (Crépon, Duguet and Mairessec, 1998) Crépon, Duguet and Mairessec (1998) proposed a method to estimate the relationship between innovation and performance called CDM model which is applied and developed by many researchers over the world such as Miguel Benavente (2006), Janz, Lööf, and Peters (2003), Mairesse, et al (2005) In the research of Siedschlag, Zhang and Cahill (2010), CDM model also is employed with panel data of 723 firms from Community Innovation Survey of Ireland in period

of 2004-2008 and control for foreign ownership and international trade activities CDM model with three stages of estimation: (i) firm's decision to invest in innovation; (ii) determine innovation output using innovation inputs and (iii) innovation output and other production inputs in the relationship with final output production Through these three step, they concludes that: (i) foreign owned firms and domestic firms involved in export activities are more likely to invest in innovation than firms with domestic activities only; (ii) foreign owned firms and domestic firms involved in export activities are more likely to have innovation output Innovation expenditure have no significant effect on innovation output; and (iii) innovation outputs have positive relationship with labor productivity

The innovation – firm’s performance relationship is argued as causal relationship (Belderbos, Carree and Lokshin (2004); Lokshin, Belderbos and Carree (2008); Parisi, Schiantarelli and Sembenelli (2006); Santos, Basso, Kimura and Kayo (2014)) In one hand, innovation creates new products, improves production process

or business practices then leads to an improvement in firm’s efficiency and their performance as well In another hand, firms with better performance tend to put more effort on innovation creation Because of the causal relationship between innovation and firm’s performance, IV or GMM estimation have been applied in many studies

to correct the endogeneity problem arises

Belderbos, Carree and Lokshin (2004) using data from Community Innovation survey in 1996 and 1998 for 2056 manufacturing firms to analyse the impact of innovation and firm performance in Netherlands Not only internal innovation

activities but also external innovative collaboration have taken into account in determining this relationship They measure firm’s innovation by two set of variables: (i) internal innovation activities represented by internal innovation expenditure per sales and (ii) external innovation collaboration through R&D cooperation dummies with competitors, suppliers, customers and universities or other research institutions Firm’s performance is expressed in labor productivity growth and sales of the products that new to the market growth IV regression has been applied to identify the causal relationship between innovation and productivity, control for firm size, 2-digit industry dummies, ownership status, demand-pull and cost-push Productivity

in the previous period also is included in the model because of the argument that past level of productivity can impact on current productivity growth They found that different types of R&D collaboration and innovation intensity significantly and positively affect productivity growth, but no significant effect of innovation intensity

on innovation sales growth

Also using internal and external innovative activities to represent for innovation, Lokshin, Belderbos and Carree (2008) found the significant positive relationship between internal innovation and labor productivity Internal innovation

is defined in-firm’s R&D expenditure while external innovation is expenditure on contracted R&D with other firms They applied GMM estimation for the dynamic panel equation from augmented Cobb-Douglas production function for 304 Netherlands manufacturing firms from 1996 to 2001 Innovation variables included

in the model beside internal and external R&D expenditure are their quadratic forms and interaction form The authors concluded that internal and external R&D are complement in the relationship with productivity with decreasing returns to scales effect, internal R&D plays important role in firm’s productivity and external R&D only have significant impact in the circumstance in which firm has invested enough

in internal R&D

Parisi, Schiantarelli and Sembenelli (2006) did the research on 941 manufacturing firms in Italy from two surveys in 1995 and 1998 to examine the innovation – firm’s performance relationship They used product and process dummies and R&D expenditure as percentage of output to represent for firm’s

innovation effort The study use two approaches to identify impact of innovation on firm performance At first, Cobb-Douglas production function is applied against output growth and innovation variables, instrumented by lag of ln(output/labor), ln(material/labor), ln(capital/labor), R&D intensity, size) In the second approach, TFP (calculated using Levinsohn and Petrin (2003)) is regressed against innovation variables, instrumented by the same set of variables in the first approach The results showed positive impact of process and product innovation on productivity In details, the impact of process innovation is bigger than product innovation These results are robust in both approaches: TFP growth and Cobb-Douglas production function estimation

In contrast, many papers found no significant effect of innovation on productivity Santos, Basso, Kimura and Kayo (2014) concluded that innovation efforts from innovative investment do not significant explain firm's performance In addition, Li and Atuahene-Gima (2001) explained the insignificant impact of innovation on firm’s performance through the uncertainty characteristic of innovation They argued that innovation activities are risky while consuming considerable resources but we do not know beforehand whether these activities creates value added to the firms Furthermore, it is required special resources and capabilities in terms of organizational structure for innovation activities to generate positive outcome to the firms as suggested by Branzei and Vertinsky (2006)

In Vietnam context, there is an emerging interest in studying about innovation, productivity and related concepts Nguyen et al (2007) has studied about the relationship between innovation and export of Vietnam SMEs in 2005, they have used dummy of whether firm introduces new product, new process or improves existing product to represent for innovation Vu and Doan (2015) also used these dummy as well as marketing changes to proxy for innovation when investigate the relationship between innovation and performance of Vietnam SMEs Firm’s performance is measured as gross profit within a year In their research, they also detected the problem of endogeneity in innovation – firm’s performance relationship and have used 2SLS model to deal with that problem This research found that innovation

efforts in product, production process or marketing do have positive impact on firm’s performance

Productivity for Vietnamese firms have been analysed widely in the literature

Ha and Kiyota (2014) used data from Annual Survey on Enterprises from 2000 to

2009 and non-parametric method to estimate firm’s TFP when analysing the relationship between firm productivity and turnover Yang and Huang (2012) has applied Levinsohn and Petrin approach to estimate firm’s TFP for Vietnamese SMEs However they focused on the effect of trade liberalization on productivity which is different with this study

This study expects to have a closer look at the important of innovation to firm productivity of SMEs in Vietnam – a transition economy Innovation effort of firm is represented in broader measurements including: innovative expenditure intensity, dummy for innovation and share of high-quality employee in total labor force Productivity in the study is measured using Levinsohn and Petrin approach which can solve the problem of endogeneity arisen due to the potential relationship between input decision and productivity shock This method of estimating TFP is rarely employed in the literature of Vietnam context

Appendix 2 provides the summaries on the related empirical studies on the relationship between innovation and firm’s performance

CHAPTER 3 RESEARCH METHODOLOGY

This chapter begins with the overview of Vietnam small and medium-sized enterprises After that, based on the literature reviews in Chapter 2, the study’s conceptual framework is constructed and described The regression techniques then are discussed together with construction of related variables employed in the model The hypotheses of the relationship between innovation and productivity are determined In the end of the chapter, source and process of filter data are presented

3.1 An overview of Vietnamese Small and Medium-sized Enterprises 3.1.1 Statistic overview

The common classifications about Small and Medium-sized enterprises of most international institutions and country over the wall are usually based on number

of employees, total revenues and/or total assets or equity In line with The Decree

No 56/2009/ND-CP regarding assistances for the development of small and medium – sized enterprises in Vietnam, small and medium – sized enterprises have been defined as firms with total employee between 10 and 300, and total equity less than

100 billion dong The details of the classification following Decree No

56/2009/ND-CP is presented in Table 3.1 below:

Table 3.1: Classification of SMEs in Vietnam

Micro

Industry Average no

construction < 10 10-200 < VND 20 bil 200-300 VND 20–100 bil

Source: Government’s decree No 56/2009/ND-CP

The number of enterprises in Vietnam has been increased significantly after implementing Enterprises Law in 2005 (Figure 3.1) Together with the increase of total enterprises, total SMEs also gone up dramatically, from 2006 to 2013 total Vietnamese SMEs has been doubled up, from 120,074 (in 2006) to 367,300 (in 2013)

In Vietnam context, the number of total micro, small and medium – sized firms is the largest and accounted for approximately 96% to 98% total firms every year from 2006

to 2013 The number of total enterprises has been increased year to year, but the size

of enterprises has been gradually tighten down In further details, in 2006, 61% of total firms are micro firms (which have 1-10 employees) compared with 66% in 2006,

in the second rank is small firms with 35% in 2006 and 32% in 2013, and large firms decreased from 4% in 2006 to only 1.6% in 2013 The downsizing trend of Vietnam enterprises leads to the insufficient medium and large – sized enterprises to guide the economy in the international integration

Figure 3.1: Number of enterprises at 31/12 (by size of employees)

Source: General Statistic Office (2006-2011, 2012, 2013)

SMEs has been growing significantly, accounted for the biggest part of total enterprises and contributed significantly for the economy According to the Mid-term Evaluation Report of the Implementation of Developing Small and Medium – sized enterprises Plan in the period of 2011-2015, SMEs has four main contributions to the economy: GDP, government budget, total investment and employment

In the structure of GDP, non-state owned enterprises (which SMEs account up

to 98.6%) has been contributed the most to the total GDP of the whole economy in the period of 2009-2012, at around 48-49% In the second rank, the contribution of state owned enterprises, in which 59.3% are SMEs, has been decreased from 37.32% (2009) to 32.57% (2012) due to the government privatization plan applied for this kind of firm Foreign direct invested enterprises (in which 78,8% are SMEs) is in the third rank with the stable contribution to total GDP at around 17-18% during 2009 to

2012 Because of this GDP structure and the privatization process, it is expected that SMEs in the non-state owned enterprises sector would become more important in the growth of GDP in the future

The total amount of taxes and fees that SMEs has contributed accounting for

a large part in government budget since they are the key sector in the economy In

2010, SMEs contributed over VND 181,060 billion dong to the state budget, accounting for 41% total collected taxes and fees from all enterprises In 2011 and

2012, these figures are VND 181,210 billion dong (34%) and VND 205,260 billion dong (34%)

In term of the contribution to the total national investment, SMEs accounted for a large part of total investment of the whole economy For further details, in 2010 total investment of SMEs was VND 236,119 billion dong, accounting for 32% total investment of enterprise sector In 2011, this amount of investment increased significantly to VND 699,690 billion dong (57%) (almost from micro and small –sized enterprises) then decreased to VND 235,463 billion dong in 2012 (29%) In the structure of total investment from enterprise sector, investment from small-sized firms accounts for the largest proportion, 61%-68% from 2010 to 2012

Another contribution of SMEs to Vietnam economy is maintaining and creating jobs for employees In 2010 Vietnamese SMEs had provided more than 4.35 million jobs, accounting for nearly 45% total jobs in enterprise sector After 2 years,

in 2012 total jobs created by SMEs has increased 16.83% to 5.09 million, accounting for 47% total jobs The worker’s income in SMEs sector also has been improved overtime In 2010, the average income of an employees working in SMEs was VND

42 million dong/employee/year and this figure reaches to VND 46 million in 2011 and continuous increases to VND 61 million in 2012, approximately 90% average income of employee in the whole enterprise sector

It is obvious that SMEs play important roles in the economic growth of Vietnam They have had huge contributions in national income, government budget, and total investment and creating more jobs, stabilize worker’s income and improving their living standards However Vietnamese SMEs are still facing a lot of difficulties

3.1.2 Difficulties

Insufficient internal capital, hard to access external capital: because these

firms are in small and medium – sized, so their internal capital stock are limited and

usually in shortage In addition they found it is hard to access to external capital sources like banks and other financial institutions due to lack of collateral, untrustworthy financial operation or complex procedures or lack of information Therefore they usually fail to raise capital when they want to expand the market, change technology or invest in new projects

Outdate applied technology: technology used in most of these firms is not

updated as compared with other countries over the world A survey of UNDP has pointed out that technology import percentage of Vietnam only at 10% total export and import while this average figure for the others country is 40% The low level of technology applied in Vietnamese SMEs could be explained by three reasons: (i) relatively low profit as compared with large firms so the accumulate capital for renew technology is still limited; (ii) the ability of accessing to government preferential programmes or supporting policies; (iii) Vietnamese SMEs are still standing out of supply chain, in-developed sub-industries Low level of applied technology in Vietnamese SMEs could affect labor productivity, product quality and their competitiveness

Low level of management and labor force: in 2012, 75% workers are not going

through technical training process in official facilities In term of managers, only 40% SMEs owners have university degree Moreover not many managers have been trained in economic knowledge, management business and law; management decisions are mostly based on experience Therefore Vietnamese SMEs’ operational efficiency is still lower than other country in the region and over the world

3.2 Methodology

3.2.1 Conceptual framework

From the theories and empirical studies, the conceptual framework for this study is built and illustrated as Figure 3.2 below:

Figure 3.2: Conceptual framework

Source: Author’s construction

The impact of innovation on productivity can be estimated through two stages: (i) Total factor productivity estimation: at this stage, total factor productivity for each firm will be determined using Levinsohn and Petrin (2003) approach applied in production function Input variables is labor, capital and intermediate inputs Output variables is value added

(ii) Innovation and productivity relationship examination: the result of total factor productivity for each firm generated from stage (1) then will be regressed against innovation proxies: dummy variable of invest in innovation activities; innovation expenditure intensity; share of high-quality employees in total labor force This relationship are built on the ground of Schumpeter Theory of Innovation and empirical studies such as Siedschlag, Zhang and Cahill (2010); Belderbos, Carree and Lokshin (2004); Crespi and Pianta (2009); Santos, Basso, Kimura and Kayo (2014) and Lokshin, Belderbos and Carree (2008)

Beside the innovation variables, other control variables also will be added

to the regression at this stage, which are firm age, firm size, capital structure and past value of firm productivity These variables are suggested to have relationship with firm’s productivity by many empirical studies: Cucculelli et al (2014); De Kok et al (2006); Huergo and

Jaumandreu (2004); Dhawan (2001); Margaritis and Psillaki (2010) and

Aw, Roberts and Winston (2007)

The detail methodology for above two stages are presented in the next sections

3.2.2 Model identification

3.2.2.1 First stage: Total factor productivity estimation using Levinsohn

and Petrin (2003) approach

Levinsohn and Petrin (2003) built up a new approach to calculate TFP in the framework of Olley and Pakes (1996) As mentioned in the Literature review, instead

of using investment, LP used intermediate inputs to control for unobserved productivity (part of error term that correlated with input) and proved that using suggested proxy would generate more consistent results in coefficients estimated

LP developed the estimation in two cases: one with output variable is gross revenue and one with output variables is value added This study will employ LP approach in value added case to estimate firm’s productivity The details of this method are described following LP method in value-added case starts with Cobb-Douglas production function in logs with two inputs: the freely variable input labor

(l) and the state variable capital (k) and output: value added (y) of firm i in year t:

𝑦𝑗𝑡 = 𝛽0+ 𝛽𝑘𝑘𝑗𝑡 + 𝛽𝑙𝑙𝑗𝑡+ 𝜀𝑗𝑡 (7) Error term 𝜀𝑗𝑡 can be divided by two components: unobservable productivity (productivity shocks that correlated to inputs) 𝑤𝑗𝑡 and an i.i.d component which does not affect inputs choice decision 𝑣𝑗𝑡 Rewrite (7) we have:

𝑦𝑗𝑡 = 𝛽0+ 𝛽𝑘𝑘𝑗𝑡 + 𝛽𝑙𝑙𝑗𝑡 + 𝑤𝑗𝑡 + 𝑣𝑗𝑡 (8)

LP made an assumption of perfect market competition in which inputs and outputs price are identical for all firms in the market Therefore intermediate inputs

(i) demand can be expressed through capital and unobservable productivity without

presentation of inputs and output prices: 𝑖𝑗𝑡 = 𝑖𝑡(𝑘𝑗𝑡, 𝑤𝑗𝑡), which means that intermediate inputs have correlation with capital stock and unobservable productivity

shocks The function is index by time t to allow for any changes in inputs and output

prices across time

Another assumption was made to the intermediate inputs called monotonicity

condition: conditional on capital, an increase in productivity (or a positive

productivity shock) would lead to an increase in intermediate inputs used This assumption is straightforward, firms would produce more output in the situation that productivity shock make marginal revenue of intermediate input(s) increased, by

producing more, firms in turn use more intermediate inputs Monotonicity condition

is clearer under perfect market competition condition If the market is not perfectly

competitive, let says oligopolistic, firms would not produce more output when they experience a positive productivity shock because producing more could lead to decrease in prices and drive down their sales

Under monotonicity assumption, intermediate inputs demand function can be inverted then productivity shocks 𝑤𝑡 are determined as a function of intermediate

inputs and capital: 𝑤𝑗𝑡 = 𝑤𝑡(𝑖𝑗𝑡, 𝑘𝑗𝑡) At this step, productivity shocks are proxy using intermediate inputs

Substituting productivity shock function into (8) get:

𝑦𝑗𝑡 = 𝛽𝑙𝑙𝑗𝑡 + 𝜑𝑡(𝑖𝑗𝑡, 𝑘𝑗𝑡) + 𝑣𝑗𝑡 (9) where: 𝜑𝑡(𝑖𝑗𝑡, 𝑘𝑗𝑡) = 𝛽0+ 𝛽𝑘𝑘𝑗𝑡+ 𝑤𝑡(𝑖𝑗𝑡, 𝑘𝑗𝑡)

To estimate the coefficients of input variables in equation (7), LP suggested a process of two stages:

The first stage: Estimating coefficient of freely variable labor:

Equation (9) is partially linear: linear in variable labor and non-linear in variable intermediate inputs and capital As developed by Robinson (1988), partially linear equation can be estimated using semiparametric estimation approach In this approach, 𝑦𝑗𝑡 and 𝑙𝑗𝑖 is predicted on given 𝑖𝑗𝑡 and 𝑘𝑗𝑡 (𝐸(𝑦𝑗𝑡|𝑖𝑗𝑡, 𝑘𝑗𝑡) and 𝐸(𝑙𝑗𝑖|𝑖𝑗𝑡, 𝑘𝑗𝑡)) using weighted least squares in second order approximation in (𝑖𝑗𝑡, 𝑘𝑗𝑡)

Equation (9) then can be rewritten as:

𝐸(𝑦𝑗𝑡|𝑖𝑗𝑡, 𝑘𝑗𝑡) = 𝛽𝑙 𝐸(𝑙𝑗𝑖|𝑖𝑗𝑡, 𝑘𝑗𝑡) + 𝜑𝑡(𝑖𝑗𝑡, 𝑘𝑗𝑡) (10) (it is noted that: 𝐸(𝑣𝑗𝑡|𝑖𝑗𝑡, 𝑘𝑗𝑡) = 0