This paper investigates the impact of internet on economic growth and this impact during the recession. The data are drawn from World Bank in a panel of 201 countries from 1988 to 2010. Results from an OLS model reveal that a 10 percentage point increase in internet penetration rate raises real GDP per capita by 0.57 to 0.63 percentage points. During the recession relative to expansion, the coefficient of internet reduces but it still remains statistically positive. This suggests that internet provides a way to solve the problem of economic recession. Our results are robust to the inclusion of time and country fixed effects.
In macroeconomics, the recession refers to the period of stagnation or negative growth in an economy’s gross domestic product (GDP). In order to get rid of the recession, many economists suggest policies of stimulating aggregate demand following Keynes’s macroeconomics theory or some other monetary policies. However, in addition to these policies, internet may play a useful role in leading the growth of economy or productivity which is evident in the literature but rarely studied on the topic of recession. Accordingly, this paper investigates the relationship between internet and economic growth and further examines it, especially during the recession. Our examination suggests that internet provides stimulus to economic growth and hence solves the problem of economic recession.
This paper uses data from World Bank in a panel of 201 countries from 1998 to 2010 with 1226 observations. Following Czernich et al., (2011) [
Internet positively impacts the economy since it spreads information, stimulates innovation, builds up network, fosters business, deepens capital, improves labor market, strengthens market competition, and helps firms to profit from emerging markets. Therefore, internet may help to fight against economic downturn. We review the literature mentioning the causality between internet and economic growth as follows.
Internet facilitates access to information and reduces search costs. Firms adopting internet are able to communicate better, faster and at lower costs. This reduces internal as well as external transaction costs and thus lowers production costs and enhances productivity and generates economic growth (Harris, 1998) [
Internet deepens innovative capacities and capital, which is crucial for economic growth. Internet may affect the innovative capacities of the economy through development of new products, processes and business models to promote growth (Benhabib and Spiegel, 2005) [
Internet builds up network which enhances economic growth. It is a form of network capital with the ability to improve overall productivity across different sectors in the economy (Moshiri and Nikpour, 2010) [
Internet fosters business and hence results in economic growth. Internet is utilized to reach new markets, learn new techniques through shared experiences and develop more resilient supply chains (Parikh et al., 2007) [
Internet improves labor market and economic outcomes. High-speed internet helps better job matching (Bloom et al., 2011) [
Internet plays a great role in spreading information. Endogenous growth proposed by Romer (1990) [
A simple macroeconomic production function with constant returns to scale and aforementioned variables is formulated by a reduced form as
Here, subscript i stands for country while t means year. Yit means real gross domestic product (GDP) per capita, Ait reveals the technology state, Lit represents labor, Kit is physical capital and Hit means human capital. We take a logarithm form to transform Equation (1) into Equation (2) as below.
(2)
where α2 to α4 are parameters to be estimated. εit is an error term. Assume that technology state evolves along an exponential growth path over time with a growth parameter, λi, which is specific to a country i’s technological progress. Therefore, we can rewrite Ait as
Since internet helps to generate and spread information, λi is expected to be affected by internet. The technological growth parameter can be specified a s a function of internet as
where Iit represents internet usage for country i in year t and αt captures time specific effects. We integrate Equations (2)-(4). In addition, in order to investigate the relationship between economic growth with internet as well as recession, an interaction term between internet and recession dummy, R, is set. Subsequently, we build up a complete regression framework as
Here Yit representing real GDP per capita is expressed in 1998 purchasing power parity. In order to capture country-specific and year-specific effects, αi and αt which are separately composed of a set of country and year dummies are integrated into the model.
The term Iit capturing technology state in an economy is measured by internet penetration rate which is the number of internet users per 100 population. Lit representing labor inputs is measured by the labor participation rate. Kit meaning physical capital is proxied by the ratio of real fixed capital formation to real GDP. Hit standing for human capital is captured by secondary education gross enrollment rate. Rit is a dummy variable with value equal to 1 representing recession and 0 otherwise.
The effect of internet on real GDP per capita can be estimated by α1 and α5 while Rit =1 is specific to the recession. We can realize how internet affects economic growth during the recession by summing up α1 and α5.
Our data collected form World Bank are unbalanced panel data composed of 201 countries and 13 years from 1988 to 2010. GDP per capita is gross domestic product divided by midyear population. It is the sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products. It is calculated in US dollars. Countries under consideration are composed of four different levels of income: low, lower-middle, upper-middle and high. The mean GDP per capita for each level separately are 1560, 5844, 12,876, and 20,533. There are 1226 observations in our study and 18% of them are in the recession. Summary statistics of variables are tabulated in
Empirical model represented by Equation (5) in the above Section is estimated based on an ordinary least squares (OLS) technique. Estimation results are demonstrated in
Model 1 uses the whole sample of 1266 observations. A positive relationship between internet and economic growth is evidently consistent with findings in the literature. Results from an OLS model reveals that a 10-percen
Note: *, **, *** separately represent that the coefficient is significant at 10%, 5%, 1% significance level.
tage point increase in internet penetration rate raises real GDP per capita by 0.63 percentage points when the economy is not in the recession. The coefficient of the interaction term between internet penetration rate and recession dummy is negative (−0.011) and, at a 10% significance level, it rejects the null hypothesis that the impact of internet on economic growth keeps identical regardless of the economy is in the recession or not. That is, during the recession, the marginal effect of log of internet penetration rate reduces by 0.011. However, adding the coefficient of log of internet penetration rate (0.063) and that of the interaction term (−0.011), we still obtain a positive relationship (0.063 − 0.011 = 0.052) between internet and economic growth during the recession. In other words, a 10-percentage point increase in internet penetration rate raises real GDP per capita by 0.52 percentage points when the economy is in the recession. Therefore, internet still provides a possible way to stimulate real GDP per capita to grow during the recession. We accordingly suggest that some policies enhancing internet penetration rate should be taken in order to solve the problem of the recession.
Since Model 1 supports that the impact of internet on economic growth varies across economy cycle, we subgroup data into two parts which separately produce estimation results of Model 2 for non-recession and Model 3 for recession. R-squared coefficients across Model 1 to Model 3 are all around 0.95 that means the explanation power of each model reaches 95%. Year-specific and country-specific dummies are included in all models to control for time and country fixed effects. Almost all these coefficients are significant.
Model 2 utilizes 1036 observations when the economy is not in the recession while Model 3 utilizes 230 observations during the recession. Comparing estimation results, the coefficients of internet are both significantly positive in Models 2 and 3. However, it is smaller and less significant during the recession. This is likely due to network effects. Our data reveal that the internet penetration rate grows by 54% when the economy boosts; however, it increases only by 49% during the recession. Slower growth in the number of internet users during the recession makes the network increase slowly. Subsequently it results in slower information spreading across the network and hence a smaller impact of internet on economic growth is produced.
Moreover, the slower growth of internet use during the recession is possibly arisen from less sufficient ICT infrastructure. For instance, the broadband infrastructure grows by 247% when the economy boosts while it increases only by 75% during the recession.
Getting rid of recession is one of the most important policies of the government. In addition to traditional policy advices like stimulating aggregate demand, this paper proposes the recommendation of spreading the usage of internet, which has started quickly developing in recent two decades. We hence investigate the relationship between internet and economic growth and further examine this relationship during the recession. 1266 observations are drawn from World Bank in a panel of 201 countries from 1988 to 2010. Among them, 230 observations belong to the period of the recession. Results from an OLS model reveal that a 10 percentage point increase in internet penetration rate raises real GDP per capita by 59 percentage points when the economy boosts and by 57 percentage points during the recession, respectively. This suggests that internet is a useful tool to deal with the problem of economic recession. Our results are robust to the inclusion of time and country fixed effects.
Since a significantly positive impact of internet on economic growth during the recession is evident, we suggest that a country should increase internet penetration rate to get rid of economy recession. The government should stimulate ICT infrastructure, which can lead to more internet users and then increase real GDP per capita via network effects.
National Science Council supports this research. The project number is NSC 102-2918-I-033-001-.