2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd

2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd 2020 tác động của khoa học, công nghệ và đổi mới đến tăng trưởng kinh tế giữa các nền kinh tế oecd và ngoài oecd

Tuyển tập báo cáo hội thảo “Phát triển kinh tế Việt Nam bối cảnh biến đổi toàn cầu” The impact of Science, Technology and Innovation on economic growth among OECD and non-OECD economies Hoang Khac Lich, Nguyen Ngoc Sao Ly University of Economics and Business, Vietnam National University - Hanoi Abstract This paper examines how Science, Technology and Innovation (STI) proxied by Research and Development (R&D) affects the economic growth rate in all nations around the world By employing Fixed-effect model, Hausman test and Imputation data analysis, the results show that investigated determinants, namely capital, government expenditure, labour, export and most importantly R&D have significantly effect on the growth of the economy Keywords: R&D, economic growth, OECD economies Introduction Economic growth has always been a major concern of almost all nations around the world However, it is likely to converge, which means that the growth rate gradually declines, then reaching zero or even negative This idea is also highly supported by the law of diminishing marginal productivity The theory suggests that when a factor of production is kept constant, the increase in one input may first rise output; yet, further additional increases in that input will become less productive and eventually have no impacts, or negative impacts on output Malthus (1798) also gave an explanation of poverty by simple ratio between population growth and the growth rate of nature human food Specifically, population rises faster than food production because population rises by geometric progression, and food production rises by arithmetic progression Consequently, because of population explosion, poverty would threaten the destiny of all humanity Nonetheless, the truth is that human beings are increasingly wealthy, and the richest countries in the world are continuously growing To demonstrate the fact, Lee (1988) examined a model which combined Malthusian population growth theory and population-induced technological progress to accelerate growth In the model, Malthusian population growth theory which is closely attached with diminishing marginal productivity is effectively solved by technical progress because technical changes enable increasing productivity The great importance of science, technology and innovation (STI) in accelerating economic growth is remarkably emphasized in endogenous growth theory The theory suggests that increases in economic growth cannot be simply explained by increases in capital or labour only, but it is come from the level of technology Solow (1957) described an estimate of technological changes as the Total Factor Productivity (TFP) or Solow residual boosting the economy effectively and efficiently Similarly, Romer (1990) concluded that the large population is insufficient to generate growth, yet growth is increased by technological changes that come from intentional investment decisions to utilize outcomes To sum up, 147 Tuyển tập báo cáo hội thảo “Phát triển kinh tế Việt Nam bối cảnh biến đổi toàn cầu” theory of economic growth suggests that growth is driven by five basic factors including capital, labour, human capital, an index of the level of technology, and institutions Many studies with different approaches have shown correlation between technological progress and economic growth In general, economists in the field have indicated that positive outcomes are closely linked with effective science and technology policies This means that effective science and technology policies has created a positive impact on private sectors, especially in research and development, trade, investment and human resources training Adak (2015) studied the impact of technological progress and innovation on Turkey’s economy The study analyzed inter-relation between technological progress and economic growth in two steps The first step is that technological progress and innovation was analyzed by OLS methods In this step, the study suggested that there was a significant relation between technological import and the number of total patent applications The second step is that a long run relation between total patent applications and GDP was tested by Engel Granger and Error Correction Models Consequently, the result found that technological progress and innovation have a significant effect on economic growth Lee and Mathews (2013) also argued that STI play an integral role in accelerating transition to a sustainable mode of development However, the level of technological progress and innovation differs from countries to countries Latecomers experience drawbacks as attempting to catch-up with technological leaders Nevertheless, the significant benefit is that latecomer countries could absorb the experience of technological nations to take a leap in technological progress This requires effective policy measures, both at the domestic and international level, to facilitate technological diffusion in latecomer countries Similarly, this paper aims to bring an empirical study on the effect of STI on economic growth To be more specific, STI is proxied by Research and Development (R&D) expenditure for two reasons Firstly, R&D is often believed to have innovative influences on the economy and efficiently absorptive capacity for technology transfer (Isaksson, 2007) This means fostering R&D is a chef role of innovation and results in more new production and knowledge, which positively affect the economy Secondly, it is utilized in analyzing the relationship between STI and economic growth by a majority of researchers (Elbagory, 2018; Bozkurt ,2015 and Ulku, 2014) Moreover, Isaksson (2007) stated that R&D activities are costly, they are primarily carried out in OECD countries As a result, this work plans to examine whether a large investment in R&D is positively related to the growth of the economy by examining two groups, namely OECD and the other nations in the world The data is obtained on a yearly basis from the World Development Indicator (WDI) between 1960 to 2019 due to its availability The results show that there exists a positive relationship between R&D expenditure and economic growth in all nations around the world The remaining of this paper is structured as the following Section provides overview of empirical studies on Science, technology, innovation and economic growth Section represents methodology and data Section provides interpretations of the findings Section is conclusion and discussion 148 Tuyển tập báo cáo hội thảo “Phát triển kinh tế Việt Nam bối cảnh biến đổi toàn cầu” Overview of studies on STI and Economic growth In general, Nelson (1959) and Arrow (1962) argued that the value of an idea is not fully expressed, given the fact that total benefits it brings to society are often dramatically larger than expected The reason is that the idea could be easily shared from one person to another at small additional cost On the other hand, individuals are frequently unwilling to pay a higher cost, because they are often risk averse, and not fully appreciate benefits created by an idea In addition, an innovative idea fully embodies characteristics of a public good Romer (1990) suggested that commodities used by many people often exhibit noncompetitive characteristics, and unlimited benefits resulting from its powerful spillover effects This also shows a decreasing marginal cost because of increasing economic of scale, even though the cost for a first user could be relatively high In fact, one has to spend a great deal of money to maintain possession and cease the process of sharing of an idea However, this does not mean that the idea is obliged to provide by governments, or fails to be produced by private sectors The problem with public goods is that the competitive market will be ineffective, which specifically shows that provided goods are less than maximum benefits for the whole society Economists agree that the findings of scientific research may be low-priced when it is directly sold through a perfect competitive market Furthermore, applied research may be more interested than basic research since the risks of basic research are often higher than applied one The fact is that research projects may fall in shortage of capital as it takes a long time for commercialization As the results, small or startup businesses have to hire capital at higher price than large ones, causing many obstacles to innovation and invention This leads to the important role of governments in supporting and regulating supply of public commodities The interaction between technological knowledge, public policy and economic growth is shown in figure below The figure shows that (i) a change in technological knowledge will (i) leads to a change in composition of production and may shift the boundary between natural resources of economic value and of no economic value (ii) a change in the production structure may affect the accumulation of technological knowledge, shifts the boundary between natural resources of economic value and of no economic value, and affect production results (iii) Public policy is closely links with policy structures and operates within policy framework, may shift the boundary between natural resources of economic value and of no economic value, and directly affects the accumulation of technological knowledge, affects the composition of the production structure, and directly has impacts on production results The diagram represents the interaction between technological knowledge, public policy, and economic growth 149 Tuyển tập báo cáo hội thảo “Phát triển kinh tế Việt Nam bối cảnh biến đổi toàn cầu” Source: Lipsey, et al (2005) Diagram annotations: The double arrows represent (A) the flow of resources in facilitating structure in which labor and capital are used to produce (B) output and income The single arrows indicate the affected lines: In order to assess the impact of technological progress on economic growth, quantitative studies are conducted in different approaches Firstly, technological advances come from innovation and creativity The innovation activities are measured by global innovation index The higher innovation index is closely linked with the greater opportunities for scientific and technological progress Zalewski and Skawinska (2009) studied impacts of technological innovation on economic growth by analyzing the relationship between innovation and labor productivity, GDP per capita, and hi-tech exports in European Union and OECD countries Innovative activity is a complex and multidimensional concept measured by Summary Innovative Index (SII) for EU states and Global Summary Innovative Index (GSII) The results suggested that the relation between GDP per capita and SII for EU and other selected countries is curve-linear semi-logarithmic plot, while the correlation is a linear plot for GSII Secondly, technological progress comes from the allocation of economy resources on research and development activities Research and development (R&D) expenditure is considered as the best estimate of resource commitment for innovative activities According to Barron and Sala I Martin (2004) model, R&D is a main factor contributing positively to a Total Factor Productivity (TFP) growth that in turn has a positive impact on economic growth Therefore, R&D is regarded as a proxy of technological progress In practical side, Aghion (1998) and Zachariadis (2003) showed that there is a strong evidence of investment in R&D and growth of TFP in US economy The relationship is also found in studies using international data such as Frantzen & Griffith (2000), Redding and Reenen (2002) Coccia (2012) suggested that R&D intensity positively correlated with the measures of economic growth including GDP, TFP and labor productivity Additionally, resource allocation for R&D between public and private sectors is also of paramount importance R&D spending by private sectors has a greater impact on economic growth in comparison public sectors At the 150 Tuyển tập báo cáo hội thảo “Phát triển kinh tế Việt Nam bối cảnh biến đổi toàn cầu” same time, the spending of private sectors has had a major influence on R&D spending in public sectors Bozkurt (2015) studied the correlation between R&D spending and economic growth in Turkey The correlation was tested and explained by Johansen cointegration test and vector error correction model The results indicated that there is a causal relationship between economic growth and R&D As there is a rise in economic activities and economic growth, R&D must increase to maintain sustainability Additionally, Elbagory (2018) found a positive and non-significant impact of R&D on economic growth in six Arab countries between the year 2000 and 2012 Thirdly, technological advances will bring improvements in products and manufacturing process As the result, in the output side, the number of patents is used as measure of technological progress Josheski and Koteski (2011) investigated relationship between the increasing number of patents and GDP growth in G7 economies The ARDL model shows that there is a positive relevance between quarterly increasing in numbers of patents and GDP in long term In short term, however, there is a negative correlation between these two variables Johansen multivariate cointegration test poses that long-term multiplier between an increasing in the number of patents and GDP growth is positive In addition, Granger causality test shows that a rise in the number of patents enables GDP growth Shina (2008) studies the relevance between the increasing number of patents and economic development in Japan and Korea with Granger causality test Research results indicated that there is a two-way casual relationship in Japan Meanwhile, GDP growth in Korea generates a rise in the number of patents, but the opposite direction does not exist Fourth, the recent decades have witnessed an explosion of Information and Communication Technology (or ICT) In fact, the diffusion of information and communication technology has become a key factor for economic growth Farhadi, et al (2012) evaluated the effects of ICT on economic growth by using panel data of 159 countries from 2000 to 2009 The index of information and communication technology is based on the number of internet users, fixed broadband internet subscribers, and mobile subscribers per 100 citizens The results showed that there is a positive relationship between GDP per capita and the index of using ICT In addition, the impacts of ICT on economic growth is higher in high-income countries than other groups Iscan (2012) studied the influence of ICT on economic growth in Turkey with the support of Johansen cointegration test In this study, R&D expenditure and public investment in telecommunications is a measure of ICT The study posed that information and communication technology are relevant to the proportion of GDP in commercial, industrial, construction and manufacturing sectors Finally, traditional growth accountants not separately measure contribution of technological progress to economic growth, but aggregate it into overall productivity factors of TFP production Corrado & Hulten (2003) suggested an increasingly comprehensive national accounting booklet for national growth accountants which include investment in knowledge-based capital, intellectual property rights (R & D, patents, artistic copyright) and economic power (brand equity, labor training, organizational structure) Muchdie, et al 151 Tuyển tập báo cáo hội thảo “Phát triển kinh tế Việt Nam bối cảnh biến đổi toàn cầu” (2016) also emphasized a positive and strong relationship between total factor productivity (TFP) growth and economic growth at both national and regional level in Indonesia The summary of quantitative studies on technological progress and economic growth No Author Time Scope Proxies for STI Findings Zalewski & 2006 European Global Innovation Index (+)7 Turkey R&D expenditure (+) R&D expenditure (+) Skawinska countries & (2009) OECD countries Bozkurt (2015) 19982013 Ulku 1981- OECD countries (2004) 1997 Non-OECD (o)8 countries Sinha 1963- Japan and South Number of patents Causal (2008) 2005 Korea Josheski & 1963- G7 countries Number of patents (+) Koteski 1993 159 countries ICT (+) Turkey R&D expenditure and public (+) relationship (2011) Farhadi, Ismail 2000 - and Fooladi 2009 (2012) Iscan 1980- (2012) 2011 Muchdie 1984- (2016) 2010 Elbagory 2000- (2018) 2012 telecommunications investment Indonesia TFP (+) Arab countries R&D expenditire (o) (Source: Author’s Compilation) In conclusion, most empirical studies on the impact of STI on economic growth are conducted in developed countries The time used in the research is mostly conducted in longrun Those studies all find the positive or non-significant impact of STI on economic development Methodology and data The data is computed from WDI database including OECD economies and the other economies in the world The period investigation spans from 1960 to 2019 due to the Positive relationship Non-significant relationship 152 Tuyển tập báo cáo hội thảo “Phát triển kinh tế Việt Nam bối cảnh biến đổi toàn cầu” availability of data The dependent variable is GDP growth At the same time, there are five independent variables: R&D expenditure: The fact is that there are several factors that could be chosen as representative for science, technology, and innovation However, Research and Development is regarded as the most common ones The Romer endogenous growth model tried to explain why and how advanced countries of the world hold sustained growth In the model, technological progress is driven by R&D sector R&D expenditure is considered as an incentive factor for developments of science and technology The reason is that the expenditure is invested in human resources development, capital accumulation and innovation (Bozkurt, 2015) As the result, R&D activities enable technological progress, which allows individuals to produce more in the same amount of resources This leads to a rise in productivity, and eventually increases economic growth Therefore, in the model, R&D expenditure which is represented as lnRD is expected to have a positive sign Labor: Labor is a major source of production and indispensable part in economic activities Enhancing human capital could lead to the effective application of technology, which in turn increase production efficiency Therefore, labourg variable which measures by annual growth rate of labour force (%) is predicted to have positive relationship with economic growth This is highly supported by the endogenous growth theory The theory outlines that economic growth could accomplish by three necessary driving forces: labor, capital and technological progress Investment in capital: In general, investment is considered as all economic activity which uses resources to produce goods and services Furthermore, increasing investment in fixed capital could generate employment opportunities as it opens up construction works, expanding production size Anderson (1990) shows that investment is of great importance in a country’s growth if it is used effectively to boost the output The Solow Economic Growth model suggests that a sustained increase in capital investment leads to a rise in economic growth in short term Hence, invest variable is predicted to have positive sign Government consumption expenditure: The relationship between government consumption spending on economic growth could be negative or positive depending on chosen countries, the period of estimation, and variables Marta (2017) shows that this relationship is negative in European countries as a whole over the period 1994-2012 However, this correlation is positive in Portugal and United Kingdom In this research, an explanatory variable govconsum which measures by the percentage of government consumption expenditure in GDP is predicted to be negative Export: 153 Tuyển tập báo cáo hội thảo “Phát triển kinh tế Việt Nam bối cảnh biến đổi toàn cầu” The independent variable export which is measured by the growth rate of exports of goods and services is predicted to have a positive sign The role of exports is of paramount importance in promoting economic development This is particularly valid in the context of increasing openness in terms of trade between countries Chien-Hui & Bwo-Nung (2002) explained several reasons that have been put forward to relate these two variables First, because of higher rates of capital formation the growth of exports has a stimulating effect on total factor productivity growth Second, competition created by opening markets ensures an efficient price mechanism that facilitate optimum resource allocation and increases the pressure on industries to improve technological change, thereby fostering economic growth Therefore, the variable export is expected to be positive The model, therefore, can be constructed as below: 𝑔𝑑𝑝𝑔𝑖𝑡 = 𝛽̂0 + 𝛽̂1 𝑙𝑛𝑅𝐷𝑖𝑡 + 𝛽̂2 𝑙𝑎𝑏𝑜𝑢𝑟𝑔𝑖𝑡 + 𝛽̂3 𝑐𝑎𝑝𝑖𝑡𝑎𝑙𝑖𝑡 + 𝛽̂4 𝑔𝑜𝑣𝑐𝑜𝑛𝑠𝑢𝑚𝑖𝑡 + 𝛽̂5 𝑒𝑥𝑝𝑜𝑟𝑡𝑖𝑡 + 𝛼𝑖 + 𝜀𝑖𝑡 The detailed variables are presented in table 1: Table 1: The list of variables Variables Description Expected sign of coefficient gdpg Growth rate of GDP (annual % growth) lnRD The natural log of R&D expenditure + (million current US dollars) labourg Growth rate of total labor force + (annual % growth) capital Growth rate of gross fixed capital formation + (annual % growth) govconsum The percentage of government consumption expenditure - (% of GDP) export Growth rate of exports of goods and services + (annual % growth) Follow Sokolov-Mladenović (2016), the study employs a Fixed Effect model (FEM) for multiple regression It is because the selected model is able to analyse the impacts of chosen variables that are changing over time FEM investigates the relationship between in not only dependent but also independent variables and control variables within each observed country individually Every country processes its own properties, which determines the influences of explanatory and control variables on the economic growth One important assumption of the FEM is that those time-invariant characteristics are unique and not correlated with other entities’ characteristics One entity is different from the others, so the entity’s error term and constant should not be correlated with the other entities’ error terms (Wooldridge, 2002) In the case the error terms of two entity are not uncorrelated, the assumptions of the FEM are violated, then Random Effect model (REM) is applied instead The choice whether FEM or REM should be employed is decided by running a Hausman test 154 Tuyển tập báo cáo hội thảo “Phát triển kinh tế Việt Nam bối cảnh biến đổi toàn cầu” In the case of OECD countries, the Hausman test result was negative It is because the data suffered from heteroskedasticity To solve this problem, the data was clustered, since Wooldridge (2012) states fixed and random methods can be employed to cluster samples, then test over-identifying restrictions by applying Sargan-Hansen test Results 4.1 Descriptive statistics The scatter plots of economic growth and the figure for R&D expenditure in OECD countries and non-OECD countries between 1960 and 2019 is presented in figure and figure The noticeable point is that the fitted value which shows linear relationship between economic growth and R&D expenditure goes downward in OECD countries This means GDP growth and R&D expenditure is subjected to have negative relationship in OECD countries At the same time, the linear relationship in the other non-OECD countries keeps relatively stable at about 0, and the data of GDP growth is located around fitted value Furthermore, the value representing GDP growth seem to be more converge towards the fitted line in figure than that in figure That means the regression model of non-OECD countries better illustrate the nexus between GDP growth and R&D compared to figure More specifically, the annual average GDP growth in OECD countries is 3.356%, while that number for the other countries is higher at 3.941% However, the average natural log of R&D expenditure in OECD countries is higher at 26.698 million US dollars Meanwhile, 23.052 million US dollars are the average natural log of R&D expenditure in non-OECD countries (see Appendix 1, table and 2) Figure 1: lnRD and GDP growth in OECD countries Figure 2: lnRD and GDP growth in the other countries The specific figure (mean, standard deviation, min, max) for all variables presented in Appendix 1,2 In general, the average GDP growth rate in OECD countries is 3.356% per year, while the figure for non-OECD countries is slightly higher, at 3.941% Additionally, the figure for a mean value of natural log of R&D expenditure in OECD countries is higher, at roughly 26.69 million US dollars Meanwhile, the figure for non-OECD countries stands 155 Tuyển tập báo cáo hội thảo “Phát triển kinh tế Việt Nam bối cảnh biến đổi toàn cầu” at 23.052 million US dollars Moreover, the spending for R&D is fairly equal between OECD countries with the value of standard deviation is 2.055 4.2 Regression analysis results The initial estimated results hold several significant results First, Hausman test shows that FEM is appropriate in both groups Secondly, R&D is tested not to affect the economic growth Nonetheless, the result inverses, which means R&D has a significantly positive impacts on the economic growth after missing data analysis is applied The detailed regression result is presented in table and Table 1: Initial estimated empirical results from 1960 to 2019 Variables lnRD labourg capital govconsum export _cons N OECD countries Non-OECD countries REM1 FEM1* REM2 FEM2* -0.105* -0.305 -0.067 -0.062 (0.042) (0.189) (0.0668) (0.204) 33.937* 31.240* 17.027* 16.01 (7.47) (8.305) (7.225) (9.861) 0.212* 0.203* 0.0767* 0.069* (0.019) (0.02) (0.008) (0.008) -0.056 -0.0237* -0.118* -0.155 (0.0288) (0.045) (0.0295) (0.074) 0.161* 0.1515* 0.1429* 0.138* (0.017) (0.018) (0.01) (0.011) 4.603* 13.512* 7.471* 6.666 (1.25) (5.65) (6.78) (4.945) 37 37 85 85 t, z-statistics in parentheses, + p