Econometric Modelling of R&D and Australia's Productivity

The objective of this study was to undertake a series of modelling exercises that took advantage of the additional available data to explore the effects of R&D on Australian productivity. However, despite the advances in data and methods, our research was unable to find a consistent robust measure of the impact of R&D on productivity. In addition to core data measurement issues, the most likely explanation is that the extra data period includes disruptions or ‘shocks’ to the relationship between R&D and productivity performance in Australia. This has frustrated attempts to clearly determine the magnitude of any long-term relationship between R&D and Australian productivity. The study has covered a large expanse of modelling territory and has tested the limits of standard models to explain the effect of R&D on Australia’s productivity. It has also explored a promising approach that deals with influences on R&D and on productivity in separate models, but within a simple related system. However, whilst it has provided some clear and plausible results, uncertainties about magnitudes of some effects remain. A major message from all the analysis is that, at least for the time being, empirical estimates of the effects of R&D on Australian productivity are unreliable. Any assessment therefore requires a high degree of judgment. The views expressed in this paper are those of the staff involved and do not necessarily reflect those of the Productivity Commission.