移动技术与经济增长

Working paperNovember2020 Authors: Kalvin BahiaPrincipal Economist, GSMA Intelligencekbahia@gsma.com Pau CastellsHead of Economic Analysis, GSMA Intelligencepcastells@gsma.comXavier PedrosSenior Economist, GSMA Intelligencexpedros@gsma.com Contents 1. Introduction...........................................................................................................................................32. Research questions................................................................................................................................53. Data.......................................................................................................................................................74. Empirical strategy..................................................................................................................................85. Results.................................................................................................................................................136. Conclusions..........................................................................................................................................25References...............................................................................................................................................27Appendices..............................................................................................................................................30 1. Introduction A significant body of empirical research hasstudied the economic impacts of telecommunications overthe last two decades. The impact of fixed communications infrastructure has been coveredextensively, mostly looking at technology deployments until 2010.1Meanwhile, the literatureanalysing the impacts of mobile technology is more limited, particularly when it comes to the rolloutof more recent network technologies. While some studies have considered the rollout of mobileconnectivity until 2010, these miss the recent acceleration in the use of mobile broadband throughthe extension of 3G and 4G.2Recently, some papers have specifically addressed the impact of mobilebroadband globally for the last decade, for example Edquist et al. (2018) and ITU (2018). We contribute to this more recent wave of research in four areas. First, we evaluate the impact ofmobile technology from 2000 to 2017 across more than 160 countries–this covers the earlydeployment of 2G to the 4G cycle, and represents one of the most comprehensive panels in theliterature (particularly in the context of the most recent research looking at mobile broadband).Second, while related literature tends to look at the overall impact of mobile technology, we apply aframework to unpick specific impacts of 2G, 3G and 4G technology–addressing the question ofwhether there are additional impacts with newer infrastructure. Third, we evaluate heterogeneities inthe impact of mobile, including network effects from take-up, skills and economic structure–mechanisms that can shift mobile’s impact and that have been largely unexplored globally in recentperiods. Finally, using the same panel, this is a novel paper in applying several methods to addressendogeneity, including Structural Equations Models (SEM), Dynamic Panel Data (DPD) andInstrumental Variable (IV) approaches–which have been used in the existing literature in differentsettings. Firstly, we find that mobile technology has significantly driven GDP in the 2000–2017 period–a 10%increase in mobile adoption increased GDP by 0.5% to 1.2%, with these effects remaining broadlystable over the period analysed and materialising over andabove fixed infrastructure. Second, wefound the subsequent rollout of 2G, 3G and 4G networks has driven increasing impacts.Specifically,mobile’s baseline economic impact increases by about 15% when connections are upgraded to 3G. Forconnections upgrading from 2G to 4G, the economic impact of mobile increases by approximately25%. Third, our analysis for determinants of the impact of mobile reveal important conclusions. We havefound that the macroeconomic impacts of mobile increase with adoption, suggesting strong networkand learning effects. We also noted countries with more education experience stronger effects,providing evidence of complementarities with human capital accumulation. We also found someevidence of higher impact in economies where services and manufacturing represent a moreimportant part of the economy, further suggesting complementarities with capital and labour in thesesectors. The remainder of this paper is organised as follows. Section 2 explains the research questions weexplore, framing them in the relevant empirical literature. Following this, Section 3 describes the dataand Section 4 details the empirical strategy that we propose for our OLS fixed effects, SEM, IV and DPDframeworks. Section 5 provides results, and Section6 summarises the key conclusions andimplications, in addition to