Impact of ULL prices on the deployment of NGA (No. 404) © Photo Credit: Robert Kneschke - stock.adobe.com

Impact of ULL prices on the deployment of NGA (No. 404)

(full version only available in German)

Impact of ULL Prices on FTTH Investment empirically robust. Impact is non-linear: Positive Impact at low ULL Prices, negative Impact at high ULL Prices.

Summary

Scientific literature provides competing hypotheses on the impact of prices for Unbundled Local Loop (ULL) on investment in modern NGA networks. Theoretical models identify different effects on the investment incentives of incumbents and competing access seekers. We analyse in detail the replacement effect, the wholesale revenue effect, the business migration effect and the financing effect of ULL prices. 

This study first of all aims to describe the competing hypotheses in the literature in detail. The main contribution of the study is to test the impact of ULL prices on Fibre-to-the-Home (FTTH) investment by means of an econometric quantitative impact analysis. the econometric test is conducted on a EU-wide data set. 

The empirical analysis is based on an unbalanced panel data set which includes country-specific data for 27 EU member states for the period 2009-2014. Different to other studies we identify FTTH investments via the proxy "difference of FTTH homes passed" between two periods, which comes closer and more targeted to the relevant variable than those used in other studies. 

Our econometric analysis identifies a robust impact of ULL prices on FTTH investment. However, the structure of this impact remains complex. Simplistic impacts which always and under all circumstances hold, have not been supported by our analysis. One of our major results is that the context between ULL prices and FTTH investment is non-linear in an inverted u-shaped form. This result is robust and shows up in all econometric test model specifications. According to these findings there is a positive impact of ULL prices on investment at low ULL prices up to a turning point. From that point onwards the impact becomes negative. Our data do not allow to identify the exact location of this turning point in a robust way. Furthermore, it depends on the model specification. 

Discussion Paper is available for download.

Authors