Chapter 4 Methodology Framework and Data Source
4.2 Measurement of Variable and Data Sources
國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
Diagnosis of Causality
We study another important effect that the direction of causality between telecommunications development and economic growth can be tested using the Granger Test technique.25 Ding (2007) researched that it is to avoid causality issues by identifying the determinative effects of initial per capita income and a series of other conditional variables on the conditional
convergence approach. Hence, we will use the Pairwise Granger Causality Test with the lagged values in the regression to detect the causality direction according to the approach implemented by Ding (2005) and Liu (2008).
4.2 Measurement of Variable and Data Sources
This section discusses the measurement of variables, the sources of data and the research limitation for the analysis in this research.
Spatial Units and Time Period
The dataset in this study is constructed from 2003 to 2008 compared with Ding’s (2005) and Liu’s (2008) research, respectively. Their study covered 29 regions in the 31 provinces, autonomous regions and municipalities from 1986 to 2002, except for Province of Tibet and the Chongqing municipality, and the initial year of 1986 is selected because the size of telecommunications infrastructure in China was small and insignificant. In this research, the 31 regions can be estimated separately.
25 Granger causality is a statistical concept of causality that is based on prediction. According to Granger causality, if a signal X1 "Granger-causes" a signal X2, then past values of X1 should contain information that helps predict X2 above and beyond the information contained in past values of X2 alone. Its mathematical formulation is based on linear regression modeling of stochastic processes (Granger 1969).
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
Measurement of Variables
The variables considered in growth studies range from traditional economic variables and there are many variables being used in previous literature for measurement of this variable. The empirical findings for a panel of around a hundred countries strongly support the general notion of conditional convergence. For a given starting level of real per capita GDP, the growth rate is enhanced by higher initial schooling and life expectancy, lower fertility, lower government consumption, better maintenance of the rule of law, lower inflation, and improvement in the terms of trade (Barro, 1996). For given values of these and other variables, growth is negatively related to the initial level of real per capital (Barro, 1991). In this research, we follow the consideration that China is a developing economy. Traditional economic inputs should play a dominating rule in its economic growth but those data such as cultural and political
variables will not be considered because of the limitation of data (Ding, 2005).
Regional economic output consists of the dependent variables showed in Equation 4.2 including: per capita income indicators from demand perspective and per capita output indicators from perspective of supply. Except for the lagged growth variable, the choice of independent variables follows the line of analysis on growth determinants in the growth theory literature. But the lagged growth variable is suggested by Ding’s (2005), Liu’s (2008) and Zahra’s (2009) investigation of regional economic growth. For given values of the other
explanatory variables, the neo-classical model predicts a negative coefficient of initial GDP, which enter in the system logarithmic form. The coefficient of the log of initial GDP has the interpretation of a conditional rate of convergence (Barro, 1991).
Chinese Statistical Yearbook defines the data on GDP are computed by
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
Department of National Accounts of the National Bureau of Statistics (NBS) based on different approaches. Data on GDP and related indicators of the most recent year published in the Yearbook are not final and are subject to changes when more information from financial data and administrative records become available. Where a census has been conducted, historical data of GDP of the previous years may also undergo change. Gross Domestic Product is a measurement of value which changes depending on changes of price and production. GDP at constant prices converts the gross domestic product based on the current price into a value based on the price of the base period. Since China started GDP calculation, seven constant-price base periods have been used, i.e., 1952, 1957, 1970, 1980, 1990, 2000 and 2005. As the calculation of constant-price GDP is based on different base periods, the constant-price GDP data in Chinese Statistical Yearbook shall also be announced in accordance with various periods.
Considering the general structure applies to most growth model, three factors: physical capital K(t), labor L(t) and knowledge T(t) are the major compositions of production function: Y(t) = F﹝K(t), L(t), T(t)﹞. Capital K(t) represents the durable physical. Labor L(t) represents the inputs associated with the human body. This input includes the number of workers and the amount of time. In Chinese Statistical Yearbook 2008, it defines “number of employed persons at year-end” as the statistics is the population aged 16 and over of the whole country. Data on economically active population, employed persons and employed persons by urban and rural areas since 1990 are estimated on the basis of the 2000 National Population Census. Regarding the total investment in fixed assets, it refers to the volume of activities in
construction and purchases of fixed assets and related fees in China,
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
expressed in monetary terms during the reference period. Foreign investment refers to foreign funds received during the reference period for the construction and purchase of investment in fixed assets (covering equipment, materials and technology), including foreign borrowings (loans from foreign governments and international financial institutions, export credit, commercial loans from foreign banks, issue of bonds and stocks overseas), foreign direct investment and other foreign investments.
Initial human capital addresses the quality of labor force. The role of human capital has been highlighted by endogenous growth theory (Romer, 1990). Countries and regions with higher levels of human capital are supposed to expect higher growth rates than territories with lower levels. Different
explanations have been provided by the literature which is usually proxied by the average number schooling of years or the percentage of population with secondary or tertiary studies (Ramos, 2009). This study follows Ding (2005) and Liu (2008) in approximation of human capital for regions in China with average years of schooling for population aging greater than 6 following the World Bank approach.
Urbanization is the physical growth of urban areas from rural areas as a result of population immigration to an existing urban area. Effects include change in density and administration services. It is also defined by the United Nations as movement of people from rural to urban areas with population growth equating to urban migration.26 At the end of 2007, China's total population was 1.32 billion, with 727 million (55%) and 593 million (45%) residing in the rural and urban areas respectively. The rural population fraction
26 http://en.wikipedia.org/wiki/Urbanization
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
was 60.9% in 2001 and 73.6% in 1990.27 The annual population growth rate was estimated at 0.52% (2007 estimate). This study uses the urbanization rate as an independent variable for explanation of regional economic growth
performance.
Telecommunications infrastructure development is measured as the number of fixed-line telephone user density, mobile phone user density and Internet user density, respectively. This kind of approaches has been applied in previous literature but only involved fixed-line telephone user density (e.g., Ding, 2005 and Liu, 2008). Inclusion of mobile phone user density in the proxy of telecommunications is important because of the fact that mobile phone subscribers in China have overtaken fixed line subscribers since 2003. Ding (2005) found that the correlation between the number of total telephone lines and capital stock of telecommunications sector as high as 0.99. Thus, total number of telephone lines is a reliable indicator for telecommunications development according to past research. So this study follows the same definition to extend the study with the variable of mobile phone subscriber density and Internet subscriber density because there are now an estimated 298 million Internet subscribers in China with a year-on-year increase of 41.9%
published by China Internet Network Information Center in 2009. China's Internet popularity rate reached 22.6%, which exceeded the world's average rate of 21.9% for the first time. The number of broadband Internet subscribers in China was 270 million and the number of the national .CN domain names increased to 13.572 million, a year-on-year increase of 50.8%. Hence, this study conducts the number of Internet subscriber density as a variable in this research.
27 http://en.wikipedia.org/wiki/Urbanization_in_the_People's_Republic_of_China
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
Another type of infrastructure included in this study is transportation percentage. This study uses the total lengths of length of railways in operation, length of highways and length of navigable inland waterways as measurement of transportation infrastructure for regions in China. Length of railways in operation refers to the total length of the trunk line for passenger and freight transportation. The calculation is based on the actual length of the first line if this line has a full or partial double. Length of highways refers to the length of highways which are built in conformity with the grades specified by the highway engineering standard formulated by the Ministry of Transport, and have been formally checked and accepted by the departments of highways.
Length of navigable inland waterways is an indicator reflecting the size and development of inland water network. It refers to the length of the natural rivers, lakes, reservoirs, canals, and ditches open to navigation during a given period, which enables transportation by ships and rafts.
Data Sources
Data for 31 regions of China for the span of time from 2003 to 2008 is utilized in this research. The aggregate regional economic data of real GDP per capita, employment, population, fixed investment, urbanization,
transportation percentage, foreign direct investment, total stated owned enterprise industrial output, telecommunications infrastructure for different regions is collected and then calculated from China Statistics Yearbook and Statistical Yearbook of China Telecommunications.
In this study, we should collect historical data from following:
China Statistics Yearbook (various years)
Statistical Yearbook of China Telecommunications (in Chinese,
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
various years)
World Bank, World Development Indictors (http://www.worldbank.org/)
International Telecommunication Union Yearbook
Research Limitation
Data quality is a significant issue for any quantitative analyses. As a developing country, the aggregate levels of economic data are gathered by different levels of statistical agencies. The data of study mainly come from National Bureau of Statistics and Ministry of Industry and Information Technology’s annual statistical data. However, the data is not the first-hand information; reliability and accuracy of data within the panel dataset used in this study may not be as high as those of developed countries. Furthermore, according to the announcement expressed by National Bureau of Statistics of China, regional data in Chinese Statistical Yearbook are provided by the statistical bureaus of the provinces, autonomous regions and municipalities.
The sum of the regional data is not equal to the national total due to the decentralized accounting approach. Therefore, it is inevitable that some of these data will have some imperfections.
‧
國立 政 治 大 學