ICT, Productivity and Economic Growth – Empirical Results on Country Level

This dissertation examines the relationship between Information and Communication Technology (ICT), productivity and economic growth. ICT, as one of the driving forces for productivity development and thus for economic growth, is of considerable interest to economics. Although the impact of ICT on productivity and economic growth has already been examined in a wide range of national studies (particularly for the U.S.), comparative studies for the rest of the world are scarce. For this reason, this thesis extends the existing available cross-country literature on this topic by three empirical studies which investigate the economic impact of ICT for a broad sample of countries at all development stages. The work focuses on the question of whether ICT contributes significantly and positively to productivity and economic growth globally or whether this only applies to individual countries or country groups. Chapter 2 provides an overview of the current state of research on the contribution of ICT to productivity and economic growth. First, the economic impact of ICT is discussed from a theoretical perspective and reflected on from a productivity standpoint. Furthermore, the empirical approaches of growth accounting and estimating production functions, which are commonly used in the literature to investigate the research subject, are presented and discussed. Subsequently, an overview of the subject-related literature is given. Chapter 3 revolves around the definition and quantitative measurement of ICT. In the literature, a quantitative measurement of ICT is mostly given monetarily in the form of capital stocks. The disadvantages of using capital stocks to measure ICT are pointed out and the use of non-monetary penetration rates as adequate proxy variables for ICT is motivated. By performing a statistical principal component analysis, the penetration rates are merged into a single variable that comprises most of the information. The thus constructed ICT infrastructure variable serves as a dependent or descriptive proxy variable for ICT in the respective empirical analyses. Chapter 4 examines the economic and institutional determinants of ICT infrastructure, which explain its diffusion over a broad cross-section of more than 100 countries for the period 2002-2012. This follows the well-known fact that developed countries possess a higher level of ICT than developing countries. These differences in the distribution are examined in the literature under the term “global digital divide”. The chapter follows an approach based on variable selection methods originating from machine learning research. This approach considers a broad set of candidate explanatory variables simultaneously and selects the most relevant ones. The ICT infrastructure variable is subsequently regressed to these selected variables. The results show that the identified determinants have a high degree of explanatory power to describe the diffusion of the ICT infrastructure. Chapter 5 investigates the role of ICT in economic growth for the long-term period of 30 years (1980-2010). It is examined whether there is a positive and significant relationship between ICT and per capita income growth over a sample of more than 95 countries. The ICT infrastructure variable as constructed in chapter 3 is included in a commonly used cross-country regression model. To avoid the problem of endogeneity due to reverse causality between per capita income and ICT, an instrumental variable approach is applied. The results show that ICT infrastructure explains the per capita income growth during the investigation period positively and significantly. Chapter 6 examines the role of ICT in productivity growth for more than 120 countries in the 2001-2012 period. There is particular interest in the research question of whether developing countries have also been able to obtain substantial productivity increases through the use of ICT. This is achieved by applying an extension of the non-parametric Multi-directional Efficiency Analysis (MEA) approach. The results show that ICT productivity increased worldwide over the investigation period. The results further reveal that developing countries benefit to a lesser extent from the productivity-enhancing effects of ICT in comparison to developed countries. A regression analysis also identifies factors that explain the differences in ICT productivity between countries