The Returns to Research and Extension Services in Regions and Zones of South Australian Broadacre Agriculture: A Comparison of Two Total Factor Productivity Decomposition Models

This paper reports total factor productivity decomposition modelling results of the returns to research and extension in zones and constituent regions of broadacre agriculture in SA, with the ultimate objective of providing policy advice on future research and extension (R&E) resource allocation. A traditional TFP decomposition modelling approach provided unacceptable results at the smaller region level, but acceptable results at the larger zone level. Utilising systematic techniques to accommodate volatility in the raw statistics caused by seasonal rainfall and small sample sizes, it was possible to demonstrate that regional differences in productivity growth were due to differences in the level of research and extension resources applied and not differences in seasonal rainfall. When modelling regional activity, the results support the use of systematic techniques to counter data problems related to large inter-year fluctuations that are associated with exogenous factors and small sample sizes.