《Broadband upgrade and firm performance in rural areas: Quasi-experimental evidence》

打印
作者
Giulia Canzian;Samuele Poy;Simone Schüller
来源
REGIONAL SCIENCE AND URBAN ECONOMICS,Vol.77,Issue1,P.87-103
语言
英文
关键字
Broadband Internet;Firm performance;Quasi-natural experiment;Regional economic development;Infrastructure policy;O33;J24;L25
作者单位
European Commission, Joint Research Centre, Ispra, Italy;FBK-IRVAPP, Research Institute for the Evaluation of Public Policies, Trento, Italy;Center for Applied Statistics in Business and Economics, Catholic University, Milan, Italy;Socio-Economic Research Institute of Piedmont (IRES Piemonte), Turin, Italy;University of Piemonte Orientale, Novara, Italy;Ifo Institute, Munich, Germany;LMU Munich, Germany;IZA, Institute for the Study of Labor, Bonn, Germany;CESifo, Munich, Germany;European Commission, Joint Research Centre, Ispra, Italy;FBK-IRVAPP, Research Institute for the Evaluation of Public Policies, Trento, Italy;Center for Applied Statistics in Business and Economics, Catholic University, Milan, Italy;Socio-Economic Research Institute of Piedmont (IRES Piemonte), Turin, Italy;University of Piemonte Orientale, Novara, Italy;Ifo Institute, Munich, Germany;LMU Munich, Germany;IZA, Institute for the Study of Labor, Bonn, Germany;CESifo, Munich, Germany
摘要
This paper analyzes the impact of advanced (up to 20 Mbps download, up to 1 Mbps upload) broadband accessibility on firm performance by exploiting a unique local policy of a staggered broadband infrastructure upgrade across rural municipalities. In a difference-in-differences framework based on longitudinal balance sheet data for limited companies, we show that ADSL2+ availability is associated with increases in firms’ revenue and total factor productivity (by on average 14.8 and 9.1 percent cumulatively over two years), but not with significant changes in personnel cost or employment. Additional analysis of aggregate administrative data shows no significant impact on the total number of active firms or local employment. Pre-trend evaluations support a causal interpretation.