"Economists have long assumed that all information and communication technologies affect the performance of firms in similar ways. Enterprise and Resource Planning Software connecting sales, marketing, inventory and accounts, affected performance in the same way as a Virtual Private Network that allowed working from home or a website that created opportunities for e-commerce. Newer theoretical evidence has shown that this is not the case: technologies that affect the communication of information affect the management and organisation of firms in ways that are distinct from technologies that make it easier to store and process information. The next step was to find empirical evidence that either supported or refuted this new view.



In this GEP working paper GEP internal fellow Richard Kneller along with two former Nottingham PhD students, Tim De Stefano and Jonathan Timmis, both now at the OECD, provide for the first time evidence for the ways that communication-ICT affect the performance of firms. To do so free from concerns about endogeneity bias they use an instrumental variable approach that relies on differences in firms' access to one particular communication-ICT, namely broadband internet. These differences arise because this technology is delivered using the telephone network and different telephone exchanges were enabled for broadband at different times and the speed of connection slows the further one is located away from the exchange. They show that, after placing various restrictions on the sample, instruments based on the timing of ADSL broadband enablement and the cable distance to the local telephone exchange satisfy the conditions for instrument relevancy and validity for some types of ICT. They find in turn, that communication-ICT causally affects firm size (captured by either sales or employment) but not productivity."

Support: Digital