The Impact of Mobile Broadband Infrastructure on Technological Innovation: An Empirical Analysis*

Transcription

1 International Telecommunications Policy Review, Vol.22 No.2 June 2015, pp < Special Issue> The Impact of Mobile Broadband Infrastructure on Technological Innovation: An Empirical Analysis* Sangwon Lee ** Do Han Kim *** Hyunjung Son **** ABSTRACT Innovation is one of the drivers of national competitiveness and economic growth. Considering well-established information and communication technologies (ICT) infrastructure is essential for the national level of innovation, the ICT infrastructure such as mobile broadband has an impact on the innovation. Using longitudinal panel data, this study examines whether mobile broadband diffusion, R&D expenditure, education, income, and corruption perception index (CPI) have influenced technological innovation. The results of the data analysis suggest that mobile broadband infrastructure is one of the key drivers of technological innovation. This finding implies that mobile broadband diffusion stimulates knowledge-based innovation and contributes to economic growth. The study also finds that R&D expenditure is an influential factor of innovation, which implies that the improvement of Application received, March 14, 2015; Final version received, April 30, 2015; Accepted, May 18, * This work was supported by a grant from the Kyung Hee University in 2010 (KHU ). ** First author, Associate Professor, Department of Journalism and Communication, Kyung Hee University, 26 Kyungheedae-ro, Dongdaemun-gu, Seoul, , Republic of Korea; Tel , swlee2668@khu.ac.kr ***Corresponding author, Assistant Professor, Department of Public Administration, Kyung Hee University, 26 Kyungheedae-ro, Dongdaemun-gu, Seoul, , Republic of Korea; Tel , dohankim@khu.ac.kr **** Doctoral Student, Department of Journalism and Communication, Kyung Hee University 26 Kyungheedae-ro, Dongdaemun-gu, Seoul, , Republic of Korea; Tel , star1627@naver.com 93

2 International Telecommunications Policy Review, Vol.22 No.2 June 2015 knowledge capital by R&D investment can promote innovations. In addition, education and corruption perception index are positively associated with innovation, indicating that human capital and government efficiency can boost innovations. Key words: Mobile broadband infrastructure, Innovation, R&D expenditure, Corruption perception index Ⅰ. INTRODUCTION Technological innovation is one of the drivers of national competitiveness and economic growth. Considering well-established information and communication technologies (ICT) infrastructure is an influential factor of national level of innovation, ICT infrastructure such as broadband has an impact on the innovation. For instance, broadband infrastructure can contribute to business expansion, product innovation, and new business creation (ITU, 2012). Also, the collaborative innovation network formation through fixed and mobile broadband infrastructure can serve as platforms for sharing of ideas that are critical to innovation (Jerome, 2011). Broadband is defined as a network offering a combined speed of equal to, or greater than, 256 kbit/s for downstream connections and 64kbit/s for upstream connections, which may include more diverse broadband technologies such as mobile broadband and portable Internet (ITU, 2003; OECD, 2001). There has been a steady growth of broadband diffusion throughout the world. There were over 711 million active fixed broadband subscribers and 2,315 million active mobile broadband subscribers at the end of 2014 (ITU, 2014; see Figure 1). In terms of penetration rate, there were over 9.8 percent for fixed broadband and 32 percent for mobile broadband at the end of 2014 (ITU, 2014) Considering the potential economic impacts of broadband, it is necessary to examine the drivers of innovation including broadband infrastructure in knowledgebased economies. Especially mobile broadband infrastructure is an increasingly important component of the ICT infrastructure with diverse implications for innovations. 94

3 The Impact of Mobile Broadband Infrastructure on Technological Innovation: An Empirical Analysis The number of patent applications has been widely used to measure technological innovation. According to the OECD (2013), the number of total patent filed grew from 88,304 in 1999 to 187,858 in OECD countries made up about percent of world patent (OECD, 2013). In spite of a growing body of scholarship on economic impact of broadband, few empirical studies have examined the nexus between the diffusion of mobile broadband and technological innovation at the macro-level. Also, very few empirical studies have investigated whether institutional factor such as corruption perception index (CPI) affects innovation. In addition, there is also insufficient research on whether the level of education influences technological innovation. Using longitudinal panel data from 44 countries, this study examines whether mobile broadband diffusion, R&D expenditure, education, income, and CPI have influenced technological innovation. The data for the estimation of technological innovation cover the period from 2002 to This study employs fixed effects regression analysis for the data analysis. The study provides policy implications for countries seeking to promote innovation. <Figure 1> Global mobile broadband diffusion, World Developed Developing Source: International Telecommunication Union (2014) 95

4 International Telecommunications Policy Review, Vol.22 No.2 June 2015 Ⅱ. SURVEYING THE LITERATURE 1. The Impact of Broadband Infrastructure on the Economy Broadband infrastructure is recognized as a contributor to economic growth, productivity, employment, and national competitiveness (ITU, 2003). Also, broadband infrastructure stimulates innovation and attracts foreign investment (ITU, 2011). Well-established broadband infrastructure is essential to achieve countries social, economic and scientific goals (ITU, 2012). There is a growing body of scholarship on economic impact of broadband. Previous studies mainly focused on the impact of broadband infrastructure on economic growth, productivity gains, and job creation. For instance, previous empirical studies measured the economic impacts of the broadband infrastructure on growth (Koutroumpis, 2009; Lehr et al., 2006). By incorporating a simultaneous approach methodology that endogenizes supply, demand and output, Koutroumpis (2009) estimated the economic impact of broadband infrastructure on growth in OECD countries and found that there are increasing returns to broadband telecommunications investments. Some empirical studies found that broadband diffusion has an impact on GDP growth (Koutroumpis, 2009; Qiang & Rossotto, 2009). However, it appears that the impact of broadband varies widely, from 0.25 to 1.38 percent for every increase in 10 percent of broadband penetration (ITU, 2012). Utilizing multivariate regression analysis, Lehr et al. (2006) also found broadband diffusion enhanced economic growth and performance, and that the economic impact of broadband was measurable. Crandall & Jackson (2001) reported that a $63 billion investment in broadband in the United States resulted in a cumulative increase of $179 billion to the GDP, allowing the maximization of consumer surplus generated by new services, savings in transportation time, and new computer applications. Ford & Koutsky (2005) found that broadband infrastructure led to a 28 percent improvement in economic growth. Recently, Thompson & Garbacz (2011) found that mobile broadband had an important direct impact on GDP. Also, previous studies on the impact of broadband on productivity found positive effects of broadband on productivity. For instance, Waverman (2009) examined the economic effect of broadband infrastructure on the GDP of 15 OECD nations between 1980 and It was estimated that for every 1 percent increase 96

5 The Impact of Mobile Broadband Infrastructure on Technological Innovation: An Empirical Analysis in broadband penetration in high and medium income countries, productivity grows by 0.13 percent (Waverman, 2009). In addition, previous studies examined the impact of broadband infrastructure on job creation. For example, Atkinson et al. (2009) estimated that a $10 billion investment in broadband in the United States resulted in a total 180,000 jobs per year. Furthermore, broadband infrastructure and technological innovations are closely interlinked. Broadband infrastructure can be a key driver of knowledge-based innovation at the macro-level. Diffusion of broadband infrastructure accelerates innovation by introducing new consumer applications and services (ITU, 2012). Ubiquitous broadband Internet access has promoted networked innovation and collaboration among universities, industries, and governments. With the diffusion of broadband networks, collaborative innovation networks can serve as platforms for the incubation and sharing of ideas that are critical to innovation (Jerome, 2011). In spite of a growing body of scholarship on economic impact of broadband, there is no empirical work examining the impact of broadband infrastructure on technological innovation. Especially many studies focused on economic impact of fixed broadband infrastructure. Considering there has been a steady growth of mobile broadband adoption, it is necessary to examine whether mobile broadband infrastructure affects technological innovations. Based on the literature review, the following research question is proposed: RQ 1: Does mobile broadband infrastructure affect technological innovations? 2. R&D and Education The size of R&D spending and employment is considered as one of important determinants of national innovative capacity (Furman & Hayes, 2004; Furman, Porter & Stern, 2002). Also, R&D plays a key role in building the knowledge-based innovation system through the interactions of academia, industry, and government (Leydesdorff, 2006). Therefore, R&D expenditures have generally been used as a measure of inputs in the national innovation system (Castellacci & Natera, 2013; Daniell & Persson, 2003; Krammer, 2009; Leydesdorff, 2006). For example, Castellacci & Natera (2013) included a total R&D expenditure as one of the innovative capabilities and examined relationships between innovative capability and absorptive capacity. Also, Daniell & Persson (2003) analyzed the Swedish regional R&D activities to evaluate the national innovation system. 97

6 International Telecommunications Policy Review, Vol.22 No.2 June 2015 Previous empirical studies found that R&D expenditure is one of drivers of innovation. For instance, using 17 OECD countries data from 1973 to 1996, Furman et al. (2002) found that the aggregate R&D spending is one of the determinants of national innovative capacity. Employing Stochastic Frontier Approach (SFA), Fu & Yang (2009) found that a gross expenditure on R&D in GDP is associated with innovation measured by the number of patents granted 3 years later. In addition, using a firm-level data, Horbach (2008) found that the improvement of the technological capabilities (knowledge capital) by R&D triggered environmental innovations. Also, education is one of the determinants of innovation. Often, a role of higher education is assessed in the context of the national innovation system. In the triple helix model, higher education plays a key role as a provider of trained persons and basic knowledge (Etzkowitz, 2003). As Etzkowitz & Dzisah (2007) suggested, although the university is the highest educational institution that fulfills its mission of research and education to motivate the process of innovation, mass primary and secondary education should precede the extensive development of tertiary educational capabilities (Etzkowitz & Dzisah, 2007). Previous empirical studies found that education is one of the drivers of innovation. For instance, Castellaci & Natera (2013) found human capital measured by the level of education is one of the drivers of scientific and technological output. Furman et al. (2002) found a positive relationship between education resources and new patents. Also, utilizing panel-data of 29 countries, Varsakelis (2006) found that the investment of a society in the quality of education affected the output of innovation activity. Based on the literature review, the following research questions are proposed: RQ 2: Does R&D expenditure affect technological innovations? RQ 3: Does level of education affect technological innovations? 3. Income and Corruption Perception Index Previous studies examined the relationship between socio-economic factor such as income and innovation (Allred & Park, 2007; Castellacci & Natera, 2013; Fu & Yang, 2009). For example, Allred & Park (2007) used population and per capita GDP to test firm innovation investment. Some empirical studies supported the positive influence of income on the innovation (Castellacci & Natera, 2013; Fu & 98

7 The Impact of Mobile Broadband Infrastructure on Technological Innovation: An Empirical Analysis Yang, 2009). For instance, Castellacci & Natera (2013) found that income measured by the GDP per capita affected scientific and technological output. Also, using Stochastic Frontier Approach (SFA), Fu & Yang (2009) found that gross GDP per capita is one of the determinants of patenting efficiency. National innovation system theory suggests that institutional factors are determinants of national level of innovation (Edquist, 1997; Lundvall et al., 2002). As Mauro (1995) suggests, corruption and bureaucratic inefficiency are negatively associated with high levels of investment and innovation. Varsakelis (2006) examined the relationship between institutional factor such as CPI and innovation. The study found that CPI affected innovation. Based on the literature review, the following research questions are proposed: RQ 4: Does income affect technological innovations? RQ 5: Does CPI affect technological innovations? Ⅲ. RESEARCH METHOD 1. The Empirical Model To examine influential factors of technological innovation, this study employs a regression model. The study formulated the following fixed-effects regression model : LnT it = β 0it + β 1it *lnmobilebroadband + β 2it *RD + β 3it *EDU + β 4it *INC + β 5it *CPI + δ i Z i + ε it (1) In the empirical model, the dependent variable (T it ) is technological innovation in country i by time t. Since the distribution of the dependent variable in this regression model is rightly skewed, data transformation with logarithm was employed. The independent variables included in the proposed research model were Based on the previous studies, fixed broadband penetration and the total number of researchers in a country were considered as independent variables. However, fixed broadband penetration is highly correlated with mobile broadband, and the total number of researchers was highly correlated with R&D expenditure. Therefore these factors were not included in the empirical model. 99

8 International Telecommunications Policy Review, Vol.22 No.2 June 2015 mobile broadband penetration (MobileBroadband), R&D expenditure (RD), education (EDU), income (INC), and corruption perception index (CPI). Mobile broadband penetration was also log-transformed to normalize right-skewed data. In the empirical model, β 0it is constant, Z i represents country dummies, and ε it is the error term. For the regression model, 367 observations from 44 countries were used. To control for the unobserved heterogeneity among countries, this study utilized the fixed-effects regression model. For the empirical model, the data covers the period from This study also conducted the data analysis for 34 OECD countries only. For the 34 OECD countries data analysis, 297 observations were used. 2. Measurement and Data Sources Table 1 shows the variables and data sources for the regression model. Since patents are acknowledged to provide a reliable and unbiased indication of national innovative effort (Huang et al., 2011), technological innovation was measured by the number of patent applications per one million inhabitants. Technological innovation was measured by number of patent applications in previous studies (Hipp & Grupp, 2005; Vasakelis, 2006). <Table 1> Description of variables Variables Measurement Data sources Technological innovation Mobile broadband penetration R&D expenditure Number of patent applications per 1,000,000 OECD inhabitants Number of mobile broadband subscribers per 100 ITU inhabitants Research and development expenditure (percent World Bank of GDP) Education School enrollment, secondary (percent of GDP) World Bank Income GDP per capita World Bank CPI Corruption Perception Index Transparency International 100

9 The Impact of Mobile Broadband Infrastructure on Technological Innovation: An Empirical Analysis The number of mobile broadband subscribers per 100 inhabitants was employed to measure mobile broadband penetration. R&D expenditure variable was measured by the R&D expenditure as a percent of GDP. In the model, the level of education was measured by the gross enrollment ratio at secondary schools. Income was measured by GDP per capita. Finally, corruption perception index (CPI) from Transparency International (TI) was included in the empirical model. Data were collected from the OECD, ITU (International Telecommunication Union), World Bank, and Transparency International (TI). Ⅳ. EMPIRICAL RESULTS Table 2 presents the descriptive statistics of our variables. On average, 44 countries applied 83.4 innovation patents per one million inhabitants between 2002 and 2012, ranged from 0.0 to The mean number of innovation patents per one million inhabitants for only OECD countries was about 107. The average number of mobile broadband penetration was 23.7, and 1.6 percent of their GDP was invested on R&D expenditures, on average. <Table 2> Descriptive statistics Variables Obs. Mean Std. Dev. Min. Max. Technological innovation Mobile broadband penetration R&D expenditure Education Income CPI Table 3 shows a correlation matrix of the variables. Most of our independent variables were not highly correlated referring to.70 benchmark for the strength of correlation. While the correlation between income and CPI is which is higher than benchmark, the multicollinearity check among independent variables indicates 101

10 International Telecommunications Policy Review, Vol.22 No.2 June 2015 that none of their variance inflation factors (VIFs) are higher than Thus, all independent variables are included in the model. <Table 3> Correlation matrix Variables Technological innovation Mobile broadband penetration R&D expenditure Education Income CPI Technological 1 Mobile 0.402*** 1 broadband penetration R&D expenditure 0.884*** 0.478*** 1 Education 0.398*** 0.174*** 0.405*** 1 Income 0.560*** 0.224*** 0.456*** 0.501*** 1 CPI 0.727*** 0.239*** 0.655*** 0.655*** 0.723*** 1 Note: *** p<0.01, ** p<0.05, * p<0.1 Table 4 reports the results from our estimation of the relationships among mobile broadband penetration, R&D expenditures, education, income, CPI and technological innovation. In all specifications, the directions of the coefficients of mobile broadband penetration are positive and statistically significant at 0.05 level. The regression coefficient for mobile broadband penetration in a complete model, specification (3), was 0.013, which indicated that one percent increase in mobile broadband penetration would increase innovation by percent when holding other factors constant. The regression coefficients for R&D expenditure are also positive and statistically significant in specification (2) and (3). In specification (3), the regression coefficient for R&D expenditure was 0.145, which indicated that, as R&D expenditure per GDP increase 1 percent, the innovation would increase 15.6 percent. The R-squared for the complete model (3) was In addition, the signs of coefficients for education, income, and CPI are all positive and statistically significant which are consistent with our expectations. Additionally, we ran a Hausman test for a complete model (3) using STATA s HAUMAN, and the result indicated that a random-effect model is resoundingly rejected at 0.01 level (Chisquared: 58.09). It appears the results of fixed effects model are robust. We also ran 102

11 The Impact of Mobile Broadband Infrastructure on Technological Innovation: An Empirical Analysis the same specifications using a one-year lagged mobile broadband penetration to see lagged influence on technological innovation, and found that the sign and significance level of mobile broadband penetration did not change. <Table 4> Regressions of technological innovation (Fixed effect model): All counties Variables (1) (2) (3) Mobile broadband penetration 0.033*** 0.027*** 0.013** (0.005) (0.005) (0.005) R&D expenditure 0.182*** 0.145*** (0.059) (0.052) Education 0.006** 0.005** (0.002) (0.002) Income 0.000*** (0.000) CPI 0.177*** (0.030) Constant 3.478*** 2.585*** (0.014) (0.245) (0.319) Observations R-squared Number of countries Notes: 1. The number of innovation and mobile broadband penetration are log-transformed to normalize right-skewed data. 2. We also ran the same specifications using a one-year lagged mobile broadband penetration to see lagged influence on innovation, and found that the sign and significance level of mobile broadband penetration did not change. 3. We checked a multicollinearity of independent variables and their VIFs are less than We ran a Hausman test for a complete model (3) using STATA s HAUMAN, and the result indicated that a random-effect model is resoundingly rejected at 0.01 level (Chi-squared: 58.09) 5. Standard errors in parentheses. 6. *** p<0.01, ** p<0.05, * p<0.1 Table 5 indicates the results of the reduced sample using 34 OECD countries. The regression coefficients for mobile broadband penetration are still positive and statistically significant at 0.05 level in all three specifications. The regression coefficient of broadband penetration in a complete model (3) was which was 103

12 International Telecommunications Policy Review, Vol.22 No.2 June 2015 <Table 5> Regressions of technological innovation (Fixed effect model): OECD counties Variables (1) (2) (3) Mobile broadband penetration 0.029*** 0.023*** 0.019*** (0.005) (0.005) (0.005) R&D expenditure 0.176*** 0.142*** (0.054) (0.048) Education (0.002) (0.002) Income 0.000*** (0.000) CPI 0.197*** (0.029) Constant 4.135*** 3.805*** 1.593*** (0.015) (0.257) (0.381) Observations R-squared Number of countries Notes: 1. The number of innovation and mobile broadband penetration are log-transformed to normalize right-skewed data. 2. We also ran the same specifications using a one-year lagged mobile broadband penetration to see lagged influence on innovation, and found that the sign and significance level of mobile broadband penetration did not change. 3. We checked a multicollinearity of independent variables and their VIFs are less than We ran a Hausman test for a complete model (3) using STATA s HAUMAN, and the result indicated that a random-effect model is resoundingly rejected at 0.01 level (Chi-squared: 51.22) 5. Standard errors in parentheses. 6. *** p<0.01, ** p<0.05, * p<0.1 relatively larger than in Table 4 with all countries included. The regression coefficients for R&D expenditure are also positive and statistically significant in specification (2) and (3). In specification (3), the regression coefficient for R&D expenditure was 0.142, which was slightly smaller than in Table 4 with all countries included. The R-squared for the complete model (3) was In addition, the signs of coefficients for income and CPI are all positive and statistically significant which are consistent with our expectations. The regression coefficient for education was also positive but not statistically significant at 0.05 level. In 104

13 The Impact of Mobile Broadband Infrastructure on Technological Innovation: An Empirical Analysis addition, we ran a Hausman test for a complete model (3) using STATA s HAUMAN, and the result indicated that a random-effect model is resoundingly rejected at 0.01 level (Chi-squared: 51.22). It seems the results of fixed effects model are appropriate. Utilizing the reduced sample, we also ran the same specifications using a one-year lagged mobile broadband penetration to see lagged influence on technological innovation, and found that the sign and significance level of mobile broadband penetration did not change. Ⅴ. DISCUSSION AND CONCLUSION Employing fixed effect regression models, this study examined the influential factors of technological innovation in 44 countries. One of the main purposes of this study was to examine whether mobile broadband infrastructure contributes to technological innovation. The results of the data analysis show that mobile broadband infrastructure is a key driver of technological innovation. Table 4 and Table 5 show that mobile broadband infrastructure variable is statistically significant in explaining innovation in 34 OECD countries as well as in 44 countries. This finding implies that mobile broadband diffusion stimulates knowledge-based innovation and contributes to economic growth in these countries. Also, considering that a one-year lagged mobile broadband diffusion variable is also significant, the impact of mobile broadband infrastructure on innovation may not be discontinued in a short period of time. In terms of the economic impact of broadband infrastructure, the significance of mobile broadband diffusion in the model is consistent with previous empirical studies, which found the economic impacts of broadband infrastructure (Thompson & Garbacz, 2011; Koutroumpis, 2009; Lehr et al., 2006; Qiang & Rossotto, 2009). It appears that mobile broadband infrastructure promotes innovation by introducing new consumer applications and services (ITU, 2012). Also, mobile broadband infrastructure may stimulate networked innovation and collaboration among universities, industries, and governments, which is critical to innovation. In addition, mobile broadband is essential infrastructure for promoting open innovation, which can stimulate creativity and collective intelligence across the entire innovation process. 105

14 International Telecommunications Policy Review, Vol.22 No.2 June 2015 The results of the data analysis also suggest that R&D expenditure is an influential factor of innovation. This finding is consistent with previous empirical studies on drivers of national innovation (Castellaci & Natera, 2013; Fu & Yang, 2009; Furman et al., 2002; Horbach, 2008). As Leydesdorff (2006) suggested, it appears that R&D plays an important role in building the knowledge-based innovation system through the interactions and triple helix collaboration among academia, industry, and government. Also, this finding implies that the improvement of knowledge capital by R&D investment promotes innovations. The findings of this study also indicate a significant positive relationship between education and innovation. This finding suggests that education plays a key role by providing human capital and knowledge, which is essential for innovation in a country. This finding is consistent with previous empirical studies, which examined the relationship between the level of education and innovation (Castellaci & Natera, 2013; Furman et al., 2002; Varsakelis, 2006). In addition, the results of the data analysis also indicate that income is a statistically significant factor in explaining innovation. The finding implies that financial status of a country as a socio-economic factor is an economic environmental factor, which affects national level of innovation. This result is consistent with previous empirical studies (Castellaci & Natera, 2013; Fu & Yang, 2009). Furthermore, the findings of this study indicate a significant positive relationship between corruption perception index and innovation. As Mauro (1995) suggested, corruption could be as a proxy for the relative performance of the governmental institutions. Considering Mauro s (1995) study, this finding implies that the high levels of governmental institutions performance promote innovation. In other words, as an institutional factor, the governmental institutions efficiency affects innovative productivity (Varsakelis, 2006) This study is limited by the relatively small number of observations. Utilizing more observations, future studies may include more diverse independent variables in the empirical model. For instance, future studies may include stock of international patent as an independent in the empirical model. Also, future studies could examine the effects of other factors (e.g., social media penetration, smartphone diffusion) on innovation. 106

15 The Impact of Mobile Broadband Infrastructure on Technological Innovation: An Empirical Analysis REFERENCES Allred, B. B., & Park, W. G. (2007). The influence of patent protection on firm innovation investment in manufacturing industries, Journal of International Management, 13, Atkinson, R., Castro, D., & Ezell, S. J. (2009). The digital road to recovery: A stimulus plan to create jobs, boost productivity and revitalize America. The Information Technology and Innovation Foundation, Washington, DC. Castellacci, F., & Natera, J. M. (2013). The dynamics of national innovation systems: A panel cointegration analysis of the coevolution between innovative capability and absorptive capacity, Research Policy, 42, Crandall, R. W., & Jackson, C. L. (2001). The $500 billion opportunity: The potential economic benefit of widespread diffusion of broadband Internet access. Criterion Economics, L.L.C. Danieli, R., & Persson, O. (2003). Regional R&D activities and interactions in the Swedish Triple Helix, Scientometrics, 58(2), Edquist, C. (1997). Systems of Innovation: Technologies, Institutions and Organizations. Pinter, London. Etzkowitz, H. (2003). Innovation in innovation: the Triple Helix of university-industrygovernment relations, Social Science Information, 42, Etzkowitz, H., & Dzisah, J. (2007). The Triple Helix of Innovation: Towards a University-Led Development Strategy for Africa, ATDF Journal, 4(2), Ford, G. D., & Koutsky, T. M. (2005). Broadband and economic development: A municipal case study from Florida. Review of Urban and Regional Development Studies, 17(3), Fu, X., & Yang, Q. G. (2009). Exploring the cross-country gap in patenting: A stochastic frontier approach, Research Policy, 38, Furman, J., & Hayes, R. (2004). Catching up or standing still? National innovative productivity among follower countries, Research Policy, 33, Furman, J., Porter, M. E., & Stern, S. (2002). The determinants of national innovative capacity, Research Policy, 31, Hipp, C., & Grupp, H. (2005). Innovation in the service sector, Research Policy, 34, Horbach, J. (2008). Determinants of environmental innovation: New evidence from German panel data sources, Research Policy, 38,

16 International Telecommunications Policy Review, Vol.22 No.2 June 2015 Huang, H., Shih, H., & Wu, Y. (2011). Contagion effects of national innovative capacity: Comparing structural equivalence and cohesion models, Technological Forecasting & Social Change, 78, ITU (2003). Birth of broadband. Geneva: ITU (2011). Broadband: A platform for progress. Geneva: ITU (2012). Impact of broadband on the economy. Geneva: ITU (2014). ITU key ICT data. Geneva: ITU Jerome, L. (2011). Triple helix knowledge clusters, In: M. Saad, G. Zawdie (Eds.), Theory and Practice of Triple Helix Model in Developing Countries, Taylor & Francis. Koutroumpis, P. (2009). The economic impact of broadband on growth: A simultaneous approach, Telecommunications Policy, 33, Krammer, S. (2009). Drivers of national innovation in transition: Evidence from a panel of Eastern European countries, Research Policy, 38, Lehr, W., Gillett, S., Osorio, C., & Sirbu, M. (2006). Measuring broadband s economic impact. Broadband Properties, Leydesdorff, L. (2006). The knowledge-based economy: Modeled, measured, simulated. Boca Raton, Florida: Universal-Publishers. Lundvall, B. A., Johnson, B., Andersen, E. S., & Dalum, B. (2002). National systems of production, innovation and competence building, Research Policy, 31, Mauro, P. (1995). Corruption and growth, Quarterly Journal of Economics, 110, OECD (2001). The development of broadband access in OECD countries, Paris, OECD. (2013). OECD statistics, OECD, Paris. Qiang, C. Z., & Rossotto, C. M. (2009). Economic Impacts of Broadband. Information and Communications for Development 2009: Extending Reach and Increasing Impact, Washington, DC: World Bank. Thompson, Jr., H. G., & Garbacz, C. (2011). Economic impacts of mobile versus fixed broadband, Telecommunications Policy, 35, Varsakelis, N. C. (2006). Education, political institutions and innovative activity, Research Policy, 35, Waverman, L. (2009). Economic Impact of Broadband: An Empirical Study. London: LECG. 108