The Contribution of ICT to Productivity in 18 OECD Countries: Focus on Australia
ANU Centre for Law and Economics Working Paper, No. 3, 2008
42 Pages Posted: 13 May 2017
Date Written: June 30, 2008
This study advances research on the contribution of Information and Communications Technology (ICT) to productivity. The research extends previous work reported in Roeller and Waverman (2001) and Fuss and Waverman (2005) to Australia. We use an econometric model that estimates the relationships that drive productivity so as to analyse the sources of productivity differences. Using the model we analyse the factors explaining observed differences in productivity between countries and how this changes over time. For this purpose the study uses a unique cross country data set including 18 OECD countries, and covering the period 1980 to 2003. The study examines not only the impact of a larger stock of ICT capital on productivity (ICT capital deepening), but also the potential spillover network and externality, effects of ICT (ICT spillovers). Recognising that the networking of computers, in for example the internet, rather than simply the growth of ICT capital may be the important characteristic to focus on. The study controls for the endogenous nature of ICT diffusion and adoption and therefore reverse causality, in that as GDP increases, demand for ICT increases to the extent demand for ICT is income elastic. Failure to control for this effect may lead to an overestimate of the effect of ICT on GDP.
Our comparison of US labour productivity (GDP per hour worked) with that in Australia shows the US was calculated to be 16.5% higher than in Australia in 2003. 17% of this labour productivity gap in 2003 was attributed to the US enjoying greater returns to scale than Australia. The combination of ICT capital deepening and ICT spillovers however accounted for 28% of the US’s labour productivity advantage. Production in the US in 2003 was more ICT capital intensive than production in Australia (capital deepening). In addition personal computer penetration, our indicator of the spread of ICT technology, was greater in the US than in Australia (ICT spillovers). Australia recorded a slight advantage in telecommunications penetration. Consistent with our research for other countries, it appears that ICT spillover effects play a greater role in explaining productivity differences than the direct ICT capital deepening effect. For example, ICT capital deepening accounted for 5% of the Australia-US labour productivity gap, whereas ICT spillovers accounted for 23% of the gap. Thus although both were important, we estimate that the Australia-US difference in the spread of ICT technology throughout the economy was a more important source of the labour productivity gap than the difference in ICT capital accumulation. Most of the impact of ICT spillovers is due to what we are calling IT penetration. This phenomenon is modelled as the penetration of personal computers plus the interaction of this spread with the digitalisation of the telecom network. We attribute 26% of the 16.5%Australian labour productivity disadvantage in 2003 to the fact that the US had a greater IT penetration than Australia. This disadvantage is offset to a small degree by Australia slight advantage in terms of telecommunications penetration
Turning to a comparison of labour productivity between Australia and Europe for the years 2000 and 2003, in 2000 Australia had a labour productivity disadvantage of 14.5% compared with Europe. By 2003 Australia had narrowed that gap to 9.4%. This improvement is associated with a substantial gain in the relative contribution of ICT. While the productivity gap relating to non-ICT capital deepening moved only slightly, Australia closed the gap by 2.5 percentage points through greater ICT deepening and ICT spillovers effects. Australia’s productivity gap with Europe is quite different to the gap with the United States. Compared with Europe Australia is behind in non-ICT capital deepening but significantly ahead in ICT capital deepening, and benefits substantially from ICT spillovers. In 2003, 2.6 percentage points of the 9.4% gap with the Europe can be attributed to greater non-ICT capital deepening in Europe. In contrast, Australia is ahead in terms of ICT contribution to labour productivity. In 2003, Australia had a 12.3 percentage point advantage in terms of ICT contribution to labour productivity over and above that of Europe.
We further examine, and report results on Total Factor Productivity (TFP) growth over a longer period 1998-2003, for Australia, US, Europe and Canada. Over the five year period 1998-2003, we estimate that TFP grew the fastest in Australia at 13.5% compared to 12.4% in the US, just over 10% in Canada, and just under 10% in Europe. In common with other researchers, we find that over this period Australia and the US had a particularly impressive TFP growth record. The higher TFP growth rate in Australia and US is attributed to a higher cumulative contribution of ICT spillovers: 14.9% in Australia, 13.8% for the US, compared with 12% in Canada and the 11.3% average for Europe.
Keywords: Productivity, Information and Communications Technologies, ICT, Economics, OECD Australia, applied econometrics
JEL Classification: D24, E24, J24, O47, L86, L96, N70
Suggested Citation: Suggested Citation