The Economics of Broadband in Ireland: Country Endowments, Telecommunications Capital Stock, and Household Adoption Decisions

Abstract

Economic growth is central to improved living standards, in turn; technological change and the spread of economically useful knowledge are essential inputs for sustainable growth. Broadband matters for long-run growth because it reduces the cost of knowledge search and increases the rate of knowledge production. Ireland lags the OECD across a range of broadband indicators including, penetration; availability; price; and speed. What explains these disappointing broadband outcomes and why is Ireland different? Are the differences rooted in policy, or are there specific geographic and demographic reasons? Is privatisation to blame, or do the causes predate privatisation? Do consumer characteristics (income, education, preferences) explain Ireland's outcomes, or are the differences driven more by investment and supply-side constraints? The different answers to these questions will have distinct implications for policy. To investigate these questions I adopt a methodologically pluralistic approach conducting three empirical studies with different data sets and methodologies. The first study uses international data and OLS regression techniques to examine the various effects of a set of country endowments. These are each expected to provide differential advantages to countries, to be unrelated to policy, and to help explain international variation in broadband subscription. The derived model explains over 85 per cent of international variation. The results suggest broadband penetration is positively influenced by population density, the diffusion of vintage technologies, and the population's educational attainment. Using the model estimates I construct an international league table of broadband efficiency and confirm Ireland is indeed underperforming compared to the OECD. The second study uses the Perpetual Inventory Method and data from government agencies to investigate the development of Ireland's pre broadband telecommunications infrastructure and the year-on-year growth in the telephone capital stock. The telephone capital stock is the infrastructure enabling the majority of fixed-line broadband connections in Ireland. I find investment was stop-start in the twentieth century. There is little evidence telecommunications infrastructure was a consistent priority of government. The third study uses sample data from the 2006 Census of Population to estimate a logit model of household broadband adoption decisions in Ireland. The derived estimates show that odds of broadband adoption are influenced by variables related to wealth and variables related to location. The results suggest geographic differences in broadband adoption are driven more by differences in availability, than by differences in consumer preference and awareness. Overall I find little evidence that the personal characteristics of Irish consumers are responsible for Ireland's poor broadband outcomes. Ireland's underperformance is best understood as a function of weak competition and low levels of investment, especially in low density areas where the commercial case for network infrastructure provision is particularly unattractive.