Working Paper: CEPR ID: DP13910
Authors: Jonathan Haskel; Harald Edquist; Peter Goodridge
Abstract: Did the huge investment in telecommunications networks in the 1990s affect subsequent total factorproductivity? Using data from 13 European countries and the US, 1995-2013, we document the substan-tial growth and then slowdown in “telecommunications” capital and ask if this is related to the growthand slowdown in TFP. We explore this by disaggregating ICT equipment investment into “IT” and “CT”equipment investment. We test for distinct effects from each using a simple framework where CT cap-ital has network externalities and so potentially impacts TFP, with the marginal impact of CT capitalgrowth being higher in countries spending more on renting CT capital. We find: a) evidence of a robustcorrelation between (lagged) growth in (rental share-weighted) CT capital services and TFP growth; b)the estimated externality from CT capital potentially explains around 30-40% of TFP growth in NorthEuropean countries, 60% in Scandinavia and around 90% in the US; c) CT capital has a social returnaround five times its private return; and d) a slowdown in the accumulation of CT capital accounts forjust over half of the post-2003 TFP slowdown in the US but only one-tenth of the TFP slowdown in theEU
Keywords: spillovers; network effects; telecommunications; ICT; RD; externalities; growth; TFP
JEL Codes: O47; O38; O32
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| CT capital services (L84) | TFP growth (O49) |
| CT capital (F38) | TFP growth (O49) |
| CT capital accumulation slowdown (O49) | TFP growth slowdown (O49) |
| CT capital (F38) | private return (Y60) |