The contribution of information and communication technologies investment on economic growth

Although its contribution is not largely as non-ICT capital, its share in economic growth is not small and positive in all countries including developed and developing groups.. Particula

UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES

VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

THE CONTRIBUTION OF INFORMATION AND COMMUNICATION TECHNOLOGIES INVESTMENT ON ECONOMIC GROWTH

BY

THAN PHUONG LAN

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

HO CHI MINH CITY, DECEMBER 2013

UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES

VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

THE CONTRIBUTION OF INFORMATION

AND COMMUNICATION TECHNOLOGIES

INVESTMENT ON ECONOMIC GROWTH

A thesis submitted in partial fulfilment of the requirements for the degree of

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

Writing process of a thesis is a collaborative experience involving the helps and supports from many people I want to express my gratitude to those who give me the great support to complete this study

I am deeply indebted to Associate Professor Doctor Pham Hoang Van, Lecturer at Department of Economics, Baylor University (USA), Fulbright Scholar, is a super supervisor Thanks for his encouragements and useful comments to pursue this topic from the initial ideas to the final completion as well as enable me to develop a deeply understanding on my thesis

Equally, I also wish to express my grateful thanks to Associate Professor Doctor Nguyen Trong Hoai, Lecturer at Department of Economics Development, University of Economic (HCMC), my second supervisor, for his valuable suggestions during the time I write this thesis His wide knowledge, excellent advice and logical way of thinking have provided me a good basis in present this thesis

Doctor Truong Dang Thuy, Lecturer at Department of Economics Development, University of Economic (HCMC), has provided me many useful comments that help me to improve my academic writing skill as well as critical thinking Thanks for his nice supports and helpful advices

Doctor Pham Khanh Nam, Lecturer at Department of Economics Development, University of Economic (HCMC), has supports me in initial ideas for this thesis and provides me useful data source that help me a lot to continue this thesis

By the way, I also would like to express sincere thanks to Anders Halvorsen, Vice President, Administration of World Information Technology and Services Alliance (WITSA) organization He has provided me the Digital Planet Reports versions

2002, 2004, 2006, 2008 and 2010 that covers all my needed data for ICT spending Without his data, I would not be able to complete my thesis

Equally, I also wish to express my grateful thanks to Dr Vu Minh Khuong, Assistant Professor Lee Kuan Yew School of Public Policy, National University of Singapore (Singapore) Although very busy with his job, he did give me a lot of knowledge, data and useful advices for me to solve my problem when I got stuck and complete my thesis

During 2 years at VNP course, lectures and many friends have encouraged, reposed trust in me and motivated me to overcome all difficulties in learning and my life

Thanks for all

Than Phuong Lan

Dec 2013

ABSTRACT

Growth accounting method to measure the contribution of Information and Communication Technologies (ICT) investment on economic growth has widespread in 1990s in the United States and then in many other developed countries including Europe, G7 This study extends growth accounting exercise in a large sample of 32 countries in recent period from 2005 to 2010 Results from this study prove that ICT capital investment is a very important factor Although its contribution is not largely as non-ICT capital, its share in economic growth is not small and positive in all countries including developed and developing groups Even

in the period after financial crisis in year 2008, ICT investment is recognized a positive contribution to economic growth in all economies in sample

Such country-by-country results, however, “are not sufficient to judge how significant ICT has impacted output growth across economies” as Vu (2005) mentioned Therefore, in my research, I continue to analyze the main channels through which ICT investment influences the economic growth and evaluate its impacts on each channels By applying regression analysis, this study finds a robust and positive statistically significant between ICT capital on output growth and on TFP growth through spillovers effects in the full sample of 32 countries, especially

in developing economies group For developed economies, the empirical evidence

of ICT impact on output growth is smaller and even not statistically significant impact on TFP growth through spillovers

Finally, an empirical regression model is build in order to find out determinants of variation on the contribution of ICT investment on economic growth in this research Regression results show that education, openness, inward FDI, health and English fluency are the most significant and positively impact although institutional quality is not statistically significant in period before 2008 However, those impacts from these variables are smaller in later period of post-2008

ABBREVIATIONS

ICT Information and Communication Technologies

TFP Total Factor Productivity

TFPG Total Factor Productivity Growth

FDI Foreign Direct Investment

G7 Group of 7 Developed Countries: U.S, U.K, France, Germany,

Italy, Canada and Japan

OECD Organization for Economic Co-operation and Development

WITSA World Information Technology and Services Alliance

BEA U.S Bureau of Economic Analysis

UNDP United Nations Development Programme

CONTENTS

ACKNOWLEDGEMENTS

ABSTRACT

ABBREVIATIONS

Chapter I 1

INTRODUCTION 1

1.1 Problem Statement 1

1.2 Research Objectives 4

1.3 Research Questions 4

1.4 Organization of the Study 4

Chapter II 6

THEORETICAL BACKGROUND 6

2.1 Sources of Economic Growth 6

2.1.1 Two Views of Capital 7

2.1.2 Explaining Sources of Economic Growth 8

2.2 ICT and Its Impacts on Economic Growth 11

2.2.1 Information and Communication Technologies (ICT) 11

2.2.2 Accessing the Channels of Contribution to Growth by ICT 11

2.3 Measuring the Contribution of ICT Investment to Economic Growth 15

2.3.1 Approaches for Evaluating the Impacts of ICT on Growth 15

2.3.2 Measuring the Contribution of ICT Investment as a Source of Economic Growth 16 2.3.3 Empirical Literature 19

2.4 Determinants of the Variations in ICT Magnitude Contribute to Economic Growth 33

Chapter III 38

RESEARCH METHODOLOGY 38

3.1 Estimating ICT and non-ICT Capital services 38

3.1.1 ICT Investment Flows Calculation 40

3.1.2 Real ICT Investment Flows 42

3.1.3 The measurement of ICT Capital Stock using PIM 43

3.1.4 Quantities and User cost of ICT Capital Services 44

3.2 Methodology 47

3.2.1 Growth Accounting For Contribution to Economic Growth of ICT Investment 47

3.2.2 Impacts of ICT on Economic Growth 49

3.2.3 Determinants of ICT Contribution to Growth 54

CHAPTER IV 60

EMPIRICAL RESULTS AND DISCUSSIONS 60

4.1 The Contribution of ICT to Economic Growth 60

4.1.1 Magnitude of ICT Contribution to Growth 60

4.1.2 The Share of ICT Contribution to Output Growth Rate 64

4.1.3 ICT vs Non-ICT Capital 65

4.2 The Impacts of ICT Investment to Economic Growth 66

4.2.1 ICT is Capital Input Contribute to Economy Output 66

4.2.2 Spillover effects of ICT Investment on TFP Growth 71

4.3 Determinants of Variations of ICT Contribution to Economic Growth 74

CHAPTER V 81

IMPLICATIONS AND CONCLUSIONS 81

5.1 Conclusion 81

5.2 Implications 83

5.3 Limitations 86

REFERENCES 88

APPENDIX A 96

INFORMATION AND COMMUNICATION TECHNOLOGIES (ICT) 96

APPENDIX B 99

ICT CONTRIBUTION TO ECONOMIC GROWTH (RESULTS OF GROWTH ACCOUNTING METHOD) 99

LISTS OF FIGURES

Figure 2.2.1: Contribution of ICT to Economic Growth 14

Figure 2.3.1: Share of ICT-Producing Industries in Total Business Sector 20

Figure 2.3.2: Output contribution of ICT Production in the Market Sector in Finland 21

Figure 2.3.3: Share of TFP Growth of ICT Producing Sector in Total TPF and Labor Productivity 1995-2000 23

Figure 2.3.4: Contribution of ICT to Labor Productivity Growth 30

Figure 3.1.1: Summary for capital services calculation 40

Figure 4.1.1: Average GDP Growth of 32 Countries in Sample (in %) 61

Figure 4.1.2: ICT Contribution to GDP Growth (All Samples) 62

Figure 4.1.3: Average GDP Growth- Developed vs Developing 63

Figure 4.1.4: ICT Contribtion To GDP Growth Developed vs Developinng 64

Figure 4.1.5: ICT Contribute to 100% of Overall Output Growth 2005-2010 65

Figure 4.1.6: Shares in Contribution of ICT vs non-ICT Capital Input to Growth 66

LISTS OF TABLES Table 2.3-1: Contribution of ICT Capital on Economic Growth 26

Table 2.3-2: Contributions from non-ICT sector to Aggregate MFP growth 32

Table 3.1-1: ICT’s Investment to spending ratio 41

Table 3.1-2: Data sources for Capital Services Calculation 46

Table 3.2-1: Proxy and expectation sign for each determinants of ICT contribute to Growth 58

Table 3.2-2: Data sources of determinants of ICT contribute to Economic Growth 58

Table 4.2-1: ICT is Capital Input Contribute to Economy Output 67

Table 4.2-2: Specific Tests for ICT is Capital Input Contribute to Economy Output 68

Table 4.2-3: Spillovers Effects of ICT Investment on TFP Growth 72

Table 4.3-1: Determinants of Variations of ICT Contribution to Economic Growth 75

Table B-1: ICT contribution to economic growth – results of growth accounting method 99

CHAPTER I

INTRODUCTION 1.1 Problem Statement

The concept of Information and Communication Technology (ICT) had been used since 1980s by many academic researchers Since then, ICT has become an interest topic and attracts researchers in many different fields The reason is because ICT has many characteristics that make it particularly attractive in the whole development strategy

At first, it is a useful channel of communication that would overcome the geographical distance In any place, you can access information from anywhere else Through ICT, almost every kind of knowledge can be transmitted between countries and regions all around the world So with ICT, different countries have greater opportunity to interact with each other to improve themselves in cultures, politics and other features of society Especially, developing countries have the ability to increase their knowledge, productivity and new introductions of modern technologies Moreover, this capability increases continuously due to advance developments in ICT products including radio, satellite, networking and data processing

Second, ICT has applications in many different areas that significant impact on the ways of education, learning, production, health, economic manipulations… Remote

or distance education and health are among such of those benefits from ICT applications Moreover, in discussion about role of ICT in learning and education in her report of “Investing Effectively in Information and Communications Technology in Schools, 2008-2013”, Minister Mary Hanafin emphasized that a pivotal force that brings the changing in learning is the use of ICT “which provides richer, more immediate, world-relevant educational resources and opportunities”

Hence, “when used well, ICT enriches learning and enhances teaching” (Department of Education and Science 2008) For economic development, ICT applications such as PCs, telephone, mobile, internet or more recently of applications in stock market change the world economy to a new process that have not ever experienced before

Finally, the most important characteristic of ICT is called “network effects” since ICT benefits not only owners but also the society as a whole The more people use ICT products, the increasing benefits for society it can raise In addition, when more and more people use ICT applications, the quality of ICT products could be improved since experiences from users may contribute to the quality of the ICT product, especially software

It is also recognized that basing on those characteristics; ICT makes the global development strategy changed such as globalization, trade extension, and economic development through the ways of learning, working or communication Particularly

in the economic development field, contribution of ICT on output and labor productivity growth is not a new subject and also proved to be significant on many economies especially in US and other developed as well as developing countries (Oliner and Sichel 2000, Daveri 2002, Schreyer 2000, Vu 2005) Global economists even predict that investment in ICT drives the economic growth (Kooshki and Ismail 2011, Colecchia and Schreyer 2002 for several OECD countries) Moreover, ICT is considered as the key factor that explains growth differentials across countries

US economy is the most special and well-known case which attracts most economists in 1990s since it is the leader in producing and using ICT of the world Both Oliner and Sichel (2000), Jorgenson and Stiroh (2000) and Stiroh (2002) agree that ICT plays the most significant sources of surprising fast pace in the US in 1990s and “constitutes the longest-ever recorded period of sustained growth” as mentioned by Schreyer (2000)

The contributions of ICT to output and labor productivity in other developed countries are not as high as that in the U.S Daveri (2002, 2003) studied fourteen European economies and found significant contributions in only six countries Schreyer (2000) found the important role of ICT investment in growth in all G7 countries, but it is most accentuated in the U.S

However, there are mixed results on the contribution of ICT investment on growth

in developing countries Some studies do not support for statistically significant of growth impacts from ICT investment (Dewan and Kraemer 2000) Vu (2005) found that ICT investment is significant impacts but not much as that in developed countries On the other hand, positive and significant evidences are recognized for Asia in 1990s (Lee and Khatri 2003), Newly Industrialized Countries including Hong Kong, Singapore, South Korea and Malaysia in 1990-2007 (Kooshki and Ismail 2011) or in transition economies (Piatkowski - 2003)

In addition, there are debates about the importance of ICT investment on growth in developing countries and different views on effective investment in ICT in those countries In the first view, ICT is considered the best way to allow poor countries

to have leap-fogging in economic development (Alleman et al., 1994, Roller and Waverman 2001) People who support this view believe in the potential benefits of ICT (ITU 2004) Moreover, they also emphasize that bridging “digital divide”, which is described the gap separating between those who have computers and internet accessing and those who do not (Young 2001), may “fostering economic equality, educational potential, and earning potential” (Fong, M W 2009) On the other view, ICT investment in poor countries would drain needed resources for other requirements in the economy Moreover, developing countries may not exploit all the benefits from ICT investment because of lacking the necessary infrastructure (Avgerou 2008) Hence, from this view, investing in ICT is not necessary and not efficiency (Negash 2006, Dewan and Kraemer 2000)

This study measures the contribution of ICT investment on economic growth in each country in 32 countries in period from 2005 to 2010 Such country-by-country results, however, “are not sufficient to judge how significant ICT has impacted output growth across economies” as Vu (2005) mentioned Therefore, evaluating the impacts of ICT investment on economic growth across countries at global basis

is necessary In addition, confront with debates about the efficiency of ICT investment in developing countries, this study make a comparison between the impacts to economic growth of ICT in developed and developing and find out determinants of the magnitude in contribution from ICT investment

1.2 Research Objectives

This study aims to measure for the contribution from ICT investment to each economic as well as examines the impacts of ICT on economic growth through some important channels at global basis In addition, this study compares those impacts to growth from ICT investment between two groups of developed and developing After that, it investigates the determinants of the variations of ICT investment contribution to economic growth across countries and discusses some implications to increase the efficiency of ICT investment on growth

1.4 Organization of the Study

This study consists of five major chapters Continuing with this introduction chapter, the rest of the study is organized as follow

Chapter II is the theoretical framework for this study This chapter starts with the theory of sources of economic growth After that, I present about ICT and some main channels through which ICT influences the economic growth as well as its mathematical framework of each impact and empirical results discussions Finally,

an analysis for determinants of ICT magnitude contribution to economic growth is employed

Chapter III is the research methodology and data In this chapter, I will present steps

to calculate ICT capital services as well as other types of capital services and implement growth accounting method to measure the contribution of ICT investment on growth I also present models to evaluate the significantly impacts of ICT investment on economic growth through some important channels and compare those impacts between developed and developing countries Empirical model of determinants that affect to the variations of ICT investment contribution on economic growth across countries is presented at the end of this chapter

Next, the results analysis is discussed in chapter IV

Chapter V draws this study’s conclusion and implication suggestions that would help policy makers have better policies to improve efficiency in investment of ICT Limitations of this study are also addressed at the end of this chapter

CHAPTER II

THEORETICAL BACKGROUND

2.1 Sources of Economic Growth

During the years of 1950s and 1960s, “neoclassical growth theory” developed by Solow (1956) becomes a main branch for theories of economic growth This theory tries to explain long-run economic growth by exploring factors such as capital accumulation, growth rate of population and productivity While marginal productivity of capital and labor are assumed to exhibit a diminishing return, technology progress is considered as a main source of changes in income per capita

of countries as well as long-run growth of economies

However, the technological progress in neoclassical growth model by Solow (1956)

is treated as exogenous factor Therefore, it leaves the long-run growth in an unexplained manner Explaining sources of technological progress as well as how it affects economic growth attracts many latter economists

There are two approaches of technical change that differ mainly in how capital is treated In neoclassical growth model by Solow (1956), technological progress is interpreted as “disembodied” technical change in which capital in different vintages

“differ only by some fixed factor associated with wear, tear and retirement” (Hulten 1992) Hence, change in technological progress is represented for the shifts of aggregate production function Solow (1960) then interpreted technical change basing on new view of embodied-capital where differences in technical design are taken into account According to that, a significant source of productivity growth may be recognized from technical change of new capital-embodied

In the late of 1950s, there are debates between two previous different views in capital treatment and hence in interpreting source of technological progress as well

as how it affects economic growth Although some previous studies support for

unimportant of embodiment hypothesis (Denison 1964, Mathews 1964, Gordon 1988), some others believe on its important roles on productivity (Gordon 1990) Moreover, Jorgenson (1966) argued that “it has frequently been suggested that embodied and disembodied technical change are two different aspects of reality” Explaining the actual roles of disembodied and embodied technological change is significant since most capital increase their quality accompanied to their technical design overtime like fast quality improvement of ICT assets (Jorgenson and Stiroh 2000)

Hulten (1992) extends (Denison 1964) and (Jorgenson 1966) in order to show how important embodiment actual is Start with different in two views of capital which are main drivers of differences in two views of technical change, this study try to explain the gap between them and show how significant embodiment is

2.1.1 Two Views of Capital

In the first view of capital in which disembodied technical change based on, capital stock K(t) available at year t is calculated as the weighted sum of its all past investment I(t), accompanied with its efficiency rate remaining at year t

The efficiency units of investment H(t) is:

H(t) = q(t)I(t) (2.1 − 2)

In addition, under assumption of equivalent of “better” and “more”, efficiency units

of investment is used to calculate aggregate capital stock instead of investment I(t)

Therefore, new aggregate capital stock J(t) is:

J(t) = ∑(1 − δ)τH(τ) = (1 − δ)τq(τ)I(τ)

t

τ=0

) (2.1 − 3)

If ϑ(t) is “average embodied technical efficiency”, then, technical efficiency index

of investment q(t) and the ϑ(t) can be calculated from:

q(t) = H(t)⁄I(t) and ϑ(t) = J(t)⁄K(t) (2.1 − 3)

q(t) and the ϑ(t) are also calculated basing on the price ratio instead of its units (Jorgenson 1966) which is implied that:

𝑃𝐻(𝑡)𝐻(𝑡) = 𝑃𝐼(𝑡)𝐼(𝑡) 𝑎𝑛𝑑 𝑃𝐾(𝑡)𝐾(𝑡) = 𝑃𝐽(𝑡)𝐽(𝑡) (2.1 − 4) Where: PI(t), PH(t) are prices of quality-unadjusted and quality-adjusted investment goods, PK(t) and PJ(t) are cost of one unit capital and one unit quality-adjusted investment goods in turn

2.1.2 Explaining Sources of Economic Growth

There are two significant assumptions in study of Solow (1960) The first one is a constant exponential rate of embodied technical change And the second one, which

is implicitly, is the perfect substitutions in production of consumption and investment goods conventionally measured As the result, the investment goods in output side are not quality-adjusted term Jorgenson (1966) makes improvement on Solow (1960) with the first objective is to provide an embodied technical change

model that can free such restrictive assumptions Moreover, the aggregate production function by Jorgenson (1966), which captures both improvements in technical design of investment output as well as capital input, has the form of:

O(t) = C(t) + q(t)I(t) = f(ϑ(t)K(t), L(t), t) (2.1 − 6)

In previous equation, C(t) and I(t) represented for consumption and investment goods in output side K(t) and L(t) are capital and labor in the input side Recall that q(t) is technical efficiency index of investment I(t) and ϑ(t) is “average embodied technical efficiency”

Noted that in previous equation, if q(t) = 1 is the case of model by Solow (1960)

Moreover, under the assumption of all factors are assumed to be paid at their marginal products and constant return to scale of production function, the rate of economic growth can be expressed as following:

O(t)̂ = (1 − β(t))C(t)̂ + β(t)I(t)̂ + β(t)q(t)̂

= (1 − π(t))L(t)̂ + π(t)K(t)̂ + π(t)ϑ(t)̂+ λ(t) (2.1 − 7)

The hat over each variable represent for its growth rate 𝜆(𝑡) is disembodied technical change rate and is interpreted as the shift of aggregate production function β(t) and π(t) are investment and income share in turn that can be calculated from:

𝛽(𝑡) =𝑃𝐻(𝑡)𝐻(𝑡)

𝑃𝐼(𝑡)𝐼(𝑡)𝑉(𝑡) 𝑎𝑛𝑑

𝜋(𝑡) = 𝑃𝐽(𝑡)𝐽(𝑡)

𝑃𝐾(𝑡)𝐾(𝑡)𝑉(𝑡) (2.1 − 8)

V(t): Total value of input as well as output which are assumed to be identical for both sides

V(t) = PH(t)H(t) + PC(t)C(t) = PL(t)L(t) + PJ(t)J(t) (2.1 − 9)

Where:

PC(t): Prices of consumption goods C(t)

PL(t): Cost of one unit of labor L(t)

PI(t), PH(t) are prices of quality-unadjusted and quality-adjusted investment goods, PK(t) and PJ(t) are cost of one unit capital and one unit quality-adjusted investment goods in turn

Economic growth from equation (2.1 − 7) shows that quality-adjusted capital influences economic growth in two ways

 Direct contribution from technical design improvement of investment goods

on output side (in term of 𝛽(𝑡)𝑞(𝑡)̂)

 Indirect contribution to productivity growth of new capital 𝑞(𝑡)̂ and leads to increase of future 𝜗(𝑡)̂

If the output side is quality-unadjusted variable 𝑄(𝑡)̂ instead of 𝑂(𝑡)̂, then equation (2.1 − 7) can be rearranged to become:

2.2 ICT and Its Impacts on Economic Growth

2.2.1 Information and Communication Technologies (ICT)

OCED (2002) noted about the definition of ICT producing sector as the agreement

by OECD members which is based on ISIC Rev 3 (International Standard Classification) According to that classification, ICT sector is “the combination of manufacturing and services industries that capture, transmit and display data and information electronically”

Products from ICT producing sectors can be categories into 4 types including computer hardware, computer software, communications and computer services as defined by Digital Planet Report 2010 which published by WITSA (The World Technology and Services Alliance) This study follows the widely adopted definitions for ICT of WITSA organization Evaluating the contribution on and impacts to economic growth from ICT investment only based on three kinds of ICT products including hardware, software and communication equipments, except for computer services type which include outsourced services due to lacking of data More detailed for each kind of ICT types are presented in Appendix A

2.2.2 Accessing the Channels of Contribution to Growth by ICT

ICT goods and services are outputs from ICT producing industries as well as inputs for ICT using industries Therefore, in order to evaluate the impacts of ICT on economic growth, we must first distinguish carefully between ICT-producing industries in which ICT goods and services are produced and ICT-using industries where ICT products are used to produce other goods and services Evaluating the impacts of ICT must be analyzed in all both ICT-producing and ICT-using industries

Scheyer (2000) suggest that ICT influences economic growth in three ways The first is ICT production ICT producing industries are part of the economy and thus, despite their small share, their rapid growth in productivity may contribute

remarkably to the productivity growth of the overall economy The second influence of ICT is through its role as a capital input With the rapid fall in ICT products’ prices, ICT capital is used more and more intensively in substitution for other inputs The role of ICT capital together with its substitution for other inputs is

an important aspect in evaluating the contribution of ICT investment to growth The third is the network or spillover effect It is argued that producing and using ICT products benefit not only for investors or owners but also for the rest of the economy Such network or spillovers effect may lead to an improvement in economy-wide TFP

Jorgenson and Stiroh (2000) emphasized on the massive increase in the quality and power of new ICT products which would contribute to growth through capital deepening and TFP growth Technological progress contributes to productivity gains in ICT producing industries and thus to the economy-wide productivity Accumulation such rapidly quality and power improvement reflects the substitution

of ICT products for other inputs in ICT-using industries Since labor is using more and better ICT equipments, it leads to an increase in labor productivity The authors also noted that if the spillovers do not exist, the increase of labor productivity is from capital deepening thus TFP growth rate is not affected TFP is only affected in case of spillover effect from ICT-producing to using of ICT or in problems of measurement errors

Qiang, Pitt,, and Ayers (2004) summary three channels through which ICT impacts

to economic growth The first is TFP growth in ICT producing sectors through technological progress The second is increasing in ICT capital per worker or capital deepening due to falling relative ICT prices comparing to the others The third is TFP growth in ICT using sectors through reorganization of production or spillover effects of using ICT products

Piatkowski (2003) and Lee and Khatri (2003) summarized previous studies for major ways of impacts to economic growth from ICT which also separated into ICT

producing and ICT using industries Therefore, direct contribution to aggregate output of the economy and rapid growth of ICT productivity are the first two channels represented for contributions from ICT producing industries The last two channels focus on the influences on ICT-using industries through using ICT as capital inputs and the economy-wide TFP growth due to spillover effects from producing and using ICT products

Generally, most of economic researchers agree with four main channels through which ICT influences the economic growth

ICT capital input

Firms invest in ICT products as a type of capital, and combine them with other types of capital and labor in order to produce output Labor productivity is also raised through capital deepening due to on the substitution of ICT capital for the others

TFP spillovers effect

ICT is a special capital which leads to network or spillover effects and hence contributes to economy-wide TFP growth

The summary of all four previous channels is represented in Figure 2.1 to show the roles of ICT on the economic growth.

This paper, with the objective of examining the contribution of ICT investment to economic growth, focuses only on ICT-using industries but not ICT-producing industries Therefore, this paper evaluates the impacts of ICT on economic growth

on two channels where ICT is capital input and spillover effects on TFP growth

TFP (Technological Change)

ICT G&S

ICT as Capital Input

TFP (Produce & use ICT)

Capital Deepening

Economy wide TFP

Labor Productivit

y Growth

Aggregate

Output

ICT Using Sectors

Figure 2.2.1: Contribution of ICT to Economic Growth

2.3 Measuring the Contribution of ICT Investment to Economic Growth

2.3.1 Approaches for Evaluating the Impacts of ICT on Growth

Measuring the impact of ICT on economic growth can be accessed commonly through two economic techniques including non-parametric or growth accounting and parametric method or regression analysis

Growth Accounting Approach

Non-parametric or growth accounting approach was initiated by Solow (1957) Under this exercise, total output (or output per-capita) growth rate is decomposed into the contribution from growth rate of capital, labor and a term of technical change or Total Factor Productivity

Generally, growth accounting method provides an easy determination for sources of output (or productivity) growth This economic framework has characteristics of well-designed and consistent Moreover, this well-established tool provided by growth accounting framework allows means to examine “how much of a country’s

or industry’s growth in output can be explained by growth in inputs, especially by growth in different types of capital input” (Daveri, 2002) However, growth accounting exercises are justified by the neoclassical model assumptions of competitive input markets and input exhaustion Those assumptions lead to the equality between each input’s factor share and its output elasticity as mentioned in Stiroh (2002)

Parametric Approach

Parametric method or regression analysis is where econometric models are applied

to estimate each factor contribution of production function directly

Parametric approach can avoid the restrictive assumptions of growth accounting method of perfect competition Therefore, it can avoid postulating the equilibrium

equation between production elasticity and each factor income share that is unsuitable in reality

However, this approach for contribution of ICT to growth has limitations due to the issues of endogeneity The main argument is that “ICT investment is not exogenous but the decision is linked to output and productivity” (Kretschmer 2012) Therefore, most of previous studies based on econometric techniques to solve this problem

2.3.2 Measuring the Contribution of ICT Investment as a Source of Economic

Growth

In order to assess the impacts of ICT on economic growth, previous studies build a mathematical framework to measure the impacts on each channel

At a period of time t, the aggregate output Q is assumed to consist of ICT products

QICT and other products Qnon−ICT The output Q is produced from labor L, aggregate capital inputs of ICT capital KICT and non-ICT capital Knon−ICT The multi-factor productivity (MFP) or technological level is presented as disembodied Hicks-neutral technical change 𝐴𝑡

Qt = ( QICT, Qnon−ICT) = Atf(KICT, Knon−ICT , Lt) (2.3 − 7)

Output contribution and TFP growth effects in ICT-producing sector

To capture the output contribution of ICT producing industries to aggregate valued added in the economy, each input is split up into shares that are contributed to the production of outputs in ICT industries The portion of 𝑄𝐼𝐶𝑇, which contributes to aggregate GDP from production of ICT goods and services, is measured as the

output contribution of ICT producers

The TFP growth rate that is generated in these industries is measured as

contribution to total MFP of the economy

ICT capital input

Again, firms invest in ICT products as a type of capital, and combine them with other types of capital and labor in order to produce output Contribution from using ICT capital to economic growth can be measured by using previous growth accounting approach (Schreyer 2000, Oliner and Sichel 2000, Jalava and Pohjola

2001, Daveri 2001, 2002 and Vu 2005) or parametric approach (Dewan and Kraemer 2000, Vu 2005)

If the growth accounting approach is applied, previous studies follow research of Jorgenson (1966) which first employed by Jorgenson and Griliches (1967) Under the assumptions of constant return to scale and competitive products and factors markets, the output growth rate can be decomposed into growth rate of each factor accompanied with its income share and growth rate of TFP

Q̂ = SICTK̂ + SICT non−ICTKnon−ICT̂ + SLL̂ + Â (2.3 − 8)

In previous equation, hat over each variable indicates the rate of change for each one SICT, Snon−ICT, SL are nominal income share of ICT non-ICT and labor capital services in turn

The output growth from equation (2.3 − 8) can be decomposed into:

 The contribution of ICT on economic growth is calculated by term SICTK̂ ICT

 The contribution of ICT on economic growth is calculated by term

Snon−ICTKnon−ICT̂

 The contribution of labor inputs SLL̂

 Contribution of TFP growth rate Â

If parametric approach is applied, previous studies start with the Cobb Douglas production function form of equation (2.3 − 7), and then take the natural logarithms

Log(Q) = α + βICTLog(KICT) + βnon−ICTLog(Knon−ICT) + βLLog(L)

The value of βICT, βnon−ICT and βL are estimated contribution of ICT on growth

In addition, if one focuses on impact of ICT investment on labor productivity

growth, equation (2.3 − 8) is rearranged to capture this impact which is denoted

 Share from growth rate of total factor productivity Â

Schreyer (2000) noted that while as capital input, ICT is seen as “ordinary” as other types of capital inputs In this case, the impacts of ICT on labor productivity through ICT capital deepening but not by increase in TFP

TFP spillovers effect

In order to evaluate the impact of ICT on the last channel in which ICT is special input that may increase overall productivity via spillover effect, a mathematically model is employed

From equation (2.3 − 7), the spillovers effect of ICT capital denoted by θ > 0 is added to term of growth contribution of ICT capital

Q

̂ = SICT(1 + θ)K̂ + SICT non−ICTKnon−ICT̂ + SLL̂ + 𝑇𝐹𝑃̂ (2.3 − 11)

The spillovers effects θ can be capture by estimating the term of SICT(1 + θ)

The econometric techniques can be applied to estimate for coefficient SICT(1 + θ) However, as Schreyer (2000) mentioned that the non-parametric approach described

by Barro (1999) is preferred since it is difficult to obtain unbiased econometric estimates of this term

Figure 2.3.1: Share of ICT-Producing Industries in Total Business Sector Source: Summarized by Pilat, D and F C Lee (2001) from OECD (2000)

Evidence from Finland which is one of the largest ICT sector in EU (Figure 2.3) by Jalava and Pohjola (2002) in period 1974 to 1994 has showed a four-fold increase in ICT-producing sector contribute to total output growth in this economy This contribution accounted from 0.3 percentage points in period 1975-1990 to 2.0 percentage points in late 1990s ICT production is also recognized to play significant role in recovering the recession in 1990-1995 in Finland

Share of ICT-Producing industries in

total business sector (%)

Figure 2.3.2: Output contribution of ICT Production in the Market Sector in

Finland Source: Jalava, J., & Pohjola, M (2002) collected from Statistics Finland’s National Accounts Database

Although these output contributions are small share, however, Schreyer (2000) also mentioned that output growth increases rapidly in these industries and this growth rate is much faster than other industries in the economy Therefore, the contribution from ICT industries to overall output generated in the economy is significant and can be much larger since “a small sector can make a large contribution to growth and productivity performance if it experiences much more rapid volume growth than the remainder of the economy” (Pilat and Lee 2001)

In East Asia, ICT producing sector is significantly important to growth Besides Korea which experienced a large share in ICT-producing industries and especially

in ICT manufacturing comparing to other developed countries, Japan is also a special case ICT production in Japan has accounted about 28% of the manufacturing exports in this region (Kenny 2003)

-2 0 2 4 6 8

Oliner and Sichel (2000) focus on three sectors in non-farm business including semiconductors producing, other computers manufactures, and other industries in the US economy in 1990s and conclude that ICT is largely a story Their results show a large share of TFP growth rate especially in ICT-producing industries contribute to overall output growth More detail, in 1.0 percentage points of labor productivity growth between 2 halves of 1990s, technical improvement in ICT-producing sector contributed around 0.3 percentage points which contribute about one-haft attributed to total TFP growth

Jorgenson and Stiroh (2000) also focus on the US economy in the same period with Oliner and Sichel (2000) Although the analyzing based on a larger number of industries (37 US industries) comparing with only 3 industries as in Oliner and Sichel (2000), these studies find similar results

Qiang, Pitt and Ayers (2004) summary findings in period 1995-2000 from Van Ark

et al (2003) for developed and from Lee and Khatri (2003) for developing countries Results of this show the significant role of TFP growth in ICT producing industries contribute to total TFP growth and then largely contributed to labor productivity growth in the economy This contribution is remarkable in many developed

countries (Ireland, US) and especially in most developing countries (Singapore, Taiwan, Thailand, South Korea, Malaysia, Hong Kong)

Source: Qiang, Pitt and Ayers (2004) summary findings in period 1995-2000 from Van Ark et al (2003) for developed and Lee and Khatri (2003) for developing

countries

Figure 2.3.3: Share of TFP Growth of ICT Producing Sector in Total TPF and

Labor Productivity 1995-2000 ICT-using industries:

 ICT Capital Input

Most of previous studies showed positive and significant contribution of ICT to output and labor productivity (output per labor or output per hours worked) growth Evaluating the impact of ICT to growth through this aspect attracts attention of many researchers They focus not only for specific firm, industry, one or many

Share of TFP Growth in ICT Producing Sector

in Total TPF and Labor Productivity

1995-2000

Other Labor Productivity Growth

sectors in the economy but also cross-countries analysis for specific region, or the large numbers of countries in the world

Subjects of contribution from ICT capital to economic growth are most attractive in the US since US it is the leader in producing and using ICT of the world

Brynjolfsson and Hitt (1993) analyze 380 of 500 largest US companies, which represent one-third of GDP in the US in the period of 1987-91 They found that the return on computers investment account for 54.2%, while this value is just 4.1% for investment in other types of capital

Also focus on US, Oliner and Sichel (2000) applied growth accounting framework

to measure the contribution of ICT on growth in three periods from 1974-1990, 1991-1995 and 1996-1999 The authors found a relative increase in the magnitude

of ICT contribution to growth in all three periods These values increase from 0.5 (% point) in 1974-90 to 0.6 (% point) in 1991-95 and reach 1.1 percentage points in total 4.8 percentage points of output growth in the third period Moreover, the contributions from ICT exceed those contributions from other types of capital stock

in two periods of 1991-1995 and 1996-1999

Jorgenson and Stiroh (2000) reveal an accelerated decline of computer price nearly 28% per year in 1995-1998 in the US economy Although this study uses the dataset

of 37 industries in the US, they show similar results with 3-sectors dataset in Oliner and Sichel (2000) Jorgenson and Stiroh (2000) showed that the growth contribution per year of computer hardware was 0.46 percentage point (increase more than fivefold), communication and software were an additional 0.3 percentage point in late 1990s

Study by Daveri (2001) finds a triple increase in the contribution from ICT in US economy from 0.53 (percentage points) in 1990-95 to 1.45 (percentage points) in 1995-2000 Results by Colecchia and Schreyer (2002), Vu (2005) in two same periods also finds a rapid increase in ICT magnitude but with a double increase

O’Mahony and Vecchi (2005) applies dynamic panel estimation method on the pooling of industries data for both United States and United Kingdom to measure the impact of ICT capital on growth rate of real output Their results show a positive and significant impact of ICT investment on growth The estimation for individual country suggests stronger impact on the US but less conclusive in the UK

Together with the US, the contributions of ICT capital on growth also significant in many other countries although these contributions seem much less benefit than in

US In fact, most of national institutions of each country (even in OECD) do not provide enough data for ICT investment, ICT capital stock as well as the price indices of ICT assets These cause important problems when analyzing the impacts

of ICT on growth and it is also difficult to make comparisons between countries

Schreyer (2000) tapped the private data source for G7 countries from International Data Corporation (IDC) The results by applying growth accounting framework from this paper show significant contribution of ICT on economic growth of all seven countries although these contributions are not as high as that in the US In the period from 1990 to 1996, the average contribution per year of ICT to GDP growth was about 0.4 percent in the US, 0.3 percent in UK and Canada, and 0.2 percent in Italy, France, Germany and Japan Moreover, these contributions from ICT are relatively stable in all countries over the period from 1980 to 1996

Daveri (2001a) covers all US and EU countries in their research According to that research, not only ICT spending but also the investment in ICT in 1990s are much smaller than those in the US The contributions of ICT to growth in EU countries are around 0.3 to 0.6 (in percentage point) compared to 0.9 (in percentage point) in

US

Estimation for ICT contribution to growth in EU for 2 periods of 1992-1994 and 1995-1999 by Roeger (2001) showed the contributions of ICT from 0.2 to 0.3 (in

percentage point) for the former period and 0.3 to 0.6 (in percentage point) for the latter period

The study by Colechia and Schreyer (2002) for 9 OECD countries found that ICT contribution to growth in period 1980-1995 accounted for 0.2-0.5 (in percentage point) while this contribution increase to 0.3-0.9 in the later period from 1995 -

1999

Besides those studies, the study by Oulton (2001) for United Kingdom, Van der Weil (2000) for Netherlands, RWI and Gordon (2002) for Germany, Cette et at (2000) for France and Javala and Pohjola (2002) for Finland found strong contributions of ICT on growth in both two periods from 1990-95 and 1995-2000

by using growth accounting framework Their results together with large samples in

EU countries by Daveri (2001a) and Colechia and Schreyer (2002) are all summarized by Daveri (2002)

Table 2.3.1 which represents the ICT contribution to growth in the US and EU countries are collected from Daveri (2002) and the results by Vu (2005) which covers 50 economies that aimed to measure the ICT contribution to growth in US and EU countries All the contributions from ICT are positive and has significant role on output growth Moreover, there are also increases in magnitude of ICT contributions most countries between two periods of 1990-95 and 1995-2000

Table 2.3-1: Contribution of ICT Capital on Economic Growth

Although the contribution of ICT has proved to be strong and significant impact on growth, there are still little empirical studies in this field for developing countries

Piatkowski (2003b) follows up the previous study 2003a for Poland in period

1995-2000 and extends his estimation to eight transition economies including “Bulgaria, Czech Republic, Hungary, Poland, Russia, Slovakia and Slovenia” The author concludes that contribution to labor productivity and output growth of ICT investment “was much higher than what might be expected on the basis of the level

of their GDP per capita” Explaining for that large contribution from ICT, the acceleration of 20% per year on average rate in real ICT investment for almost all the transition countries is considered as the most significant factor The author also mentions that the main reason for extraordinary acceleration in real investment of ICT is the substitution process of ICT for non-ICT capital due to the falling ICT prices compared to other goods and services

Contribution from ICT capital deepening to growth increase from one-haft of a percentage point in first halve of 1990s to three-quarters of a percentage point in the second halve in Japan as result by Japan Economic Planning Agency (2000)

Haacker (2010) adapts growth accounting approaches to the dataset of 80 countries with low-income in 1996-2000 to quantify the growth impacts of ICT equipment Findings of this study show that capital deepening contributed to the growth of low –income countries is about 0.2 (in percentage points), and about 0.3 (in percentage point) in low-middle-income countries

Capturing the contributions of ICT on growth in some regions such as Africa, South Asia is scarce Explains for the lacking empirical evidence from those regions by Piatkowski (2002) is due to the quite small quantities of ICT investment or the spending for ICT in such countries much less than higher-income countries Qiang, Pitt and Ayers (2004) also mentioned about lacking available data for growth

accounting analysis in those regions such as “South Asia, the Middle East, and Africa”

Based on results by Lee and Khatri (2003) and Van Ark et al (2003), Qiang, Pitt and Ayers (2004) make a comparison for the contribution of ICT-related capital deepening to labor productivity growth between developed and developing countries Besides NIE countries such as Hong Kong, Singapore, South Korea, Taiwan and Malaysia and developed countries such as US, UK and Ireland which experience high percentage point of ICT contribution, in other Asia countries, the contribution is really small

Figure 2.3.4: Contribution of ICT to Labor Productivity Growth

Source: Qiang, C Z W., Pitt, A., & Ayers, S (2004) Based on results by Lee and Khatri (2003) and Van Ark et al (2003)

With a large of sample size which covers 36 countries over the period from 1985 to

1993, Dewan and Kraemer (2000) estimate the inter-production function of ICT and non-ICT capital inputs to the output (real GDP) Results from this study confirm statistically significant and positive returns to GDP from ICT investment in

developed countries Empirical estimation shows that the output elasticities by ICT capital, non-ICT capital and labor respectively equal to 0.057, 0.160 and 0.823 However, the contribution from ICT in developing countries is not statistically significant

Vu (2005) provides cross-country evidence on the contribution of ICT investment

on economic growth for 50 countries which are major ICT spending countries His result shows that ICT contributions to growth in 1990-1995 and 1995-2000 were positive for all countries although non-ICT capital remains the most significant source of economic growth in all groups, except for Eastern Europe The average magnitude of this contribution was surging from 0.17 to 0.37 percentage points through 2 periods Moreover, the ICT magnitudes contribute to economic growth increase at least doubled between two periods for nearly two-third of economies In addition, in order to evaluate the impacts of ICT investment on growth, Vu (2005) examines if ICT accumulation is significant influence to growth performance The result on both OLS and IV showed significant impacts of ICT on economic growth for all developed and developing countries However, it is also indicates that the roles of ICT on developed economies is much larger than developing groups

So far, the contribution of ICT as “normal capital input” has gone through mostly in

US, developed countries including EU, G7, OECD and some countries in Asia and other developed countries As noted before, there is another aspect of ICT as special capital input as in Schreyer (2000), Jorgenson and Stiroh (2000) and Qiang, Pitt and Ayers (2004) which causes spillovers effect on economy-wide TFP growth Although not much of studies research in this area, there are also some evidence for the relationship between ICT investment and the growth of TFP

 TFP Spillovers Effect

Focus on the US, Oliner and Sichel (2000) estimate efficiency in ICT producing sector (including semiconductor producing) and the growing use of ICT goods and