Competition’s Effect on Investment in Digital Infrastructure*

Competition has become a preferred means to promote investment in digital infrastructure. The U.S. Telecom Act, the WTO Reference Paper and EC Directives each adopt policies that encourage facilities-based entry into telecommunications markets to expand service coverage and to accelerate deployment of advanced technologies. Yet despite its popularity, empirical evidence that this policy stimulates infrastructure investment–either directly through entrants’ construction activity or indirectly through competitive pressures applied to incumbents–is lacking. This paper examines the effect of facilities-based competition on one specific type of digital infrastructure investment. A panel data set was collected on the deployment of digital fiber ring networks in largest U.S. cities over the 10-year period 1983-1992 by incumbent local exchange carriers (ILECs) and competitive local exchange carriers (CLECs). These networks provide highspeed access for large business customers. Controlling for state and federal regulation of rates and investment and for local economic conditions, standard exogeneity and causality tests reveal that CLEC entry leads to subsequent ILEC investment and that, with slightly less confidence, ILEC investment invites additional CLEC entry. The results suggest that incumbents and entrants match each others’ investments, and offer little evidence of preemptive behavior or strategic coordination. These findings recommend that policy makers intent on encouraging construction of the nextgeneration broadband network should eliminate artificial restrictions on facilities-based entry and ensure that their pricing of incumbent network services sold to entrants take account of the virtuous cycle of investment triggered by this entry. At the same time, the results argue for removing restrictions on incumbent investments, even those traditionally considered to be an effective means to deter competitive entry. _________________________________ * I am grateful to Simone Peart for superb research assistance, to the Center for Research on Telecommunications Policy (CRTP) for financial support, and to comments from participants at the 4 annual CRTP conference held at the University of Michigan and the 27 annual Telecommunications Policy Research Conference in Alexandria, VA.