The Returns From Research in Australian Broadacre Agriculture*

Abstract

Many people share the view that too little is invested in R & D in agriculture. The relationship between several measures of productivity and research expenditure was estimated using data from ABARE's surveys of broadacre industries and a new data series on R & D expenditure for the period 1953 to 1988. The internal rate of return to research was estimated to be in the range of 15 to 40 percent which does not provide strong evidence that Australia is either under-or over-investing in public research.

Environmentally adjusted productivity measurement: An Australian case study

“…Crops grew at 5.5 per cent, mixed crops and livestock at 2.4, sheep at 0.2, beef at 0.1 and sheep-beef at 2.4. Mullen and Cox (1995) measured productivity growth in broadacre agriculture between 1952-53 and 1987-88 using farm level data. They used the Australian Bureau of Agriculture and Resource Economic (ABARE) survey data that included producers with more than 200 sheep and found the average rate of growth for Australia was 2.3 per cent per year.…”

Agricultural research: Implications for productivity in New Zealand and Australia

“…Mullen was unable to distinguish between these models econometrically and hence any preference for say, the 16 year model, must be based on some subjective assessment about the contribution to a shorter research lag profile of the ongoing evolution of R&D institutional arrangements (management and funding) in Australia. Mullen (2007) concluded that the returns on investment are likely to have remained within the 15-40% per annum range estimated by Mullen and Cox (1995). Again the lower returns are associated with the 35 year lag model and the higher returns with a 16 year lag model estimated for the period since 1969.…”

“…Hall and Scobie (2006) tested some alternative ways of capturing the impact of a capital stock through time including the permanent inventory method (PIM) and the Koyck (1954) and Almon (1965) transformations which conserve degrees of freedom and reduce multicollinearity problems. Mullen and Cox (1995) used models similar to (1) with research lag profiles of 16 and 35 years estimated over the period 1953 -1994, which gave internal rates of return to research in the range of 15 -40%, with higher rates of return associated with the 16 year lag model. Mullen (2007) using data to 2003 found that both these models displayed evidence of structural change or misspecification over the longer observation period.…”

“…A key issue has been how best to impose structure on this model to construct the stock of knowledge variable associated with R&D. Typically it is constructed as a weighted sum of past expenditures on R&D. Mullen (2007) following Mullen and Cox (1995) imposed either an inverted V structure over 16 years or a trapezoidal structure over 35 years where the weights representing the impact of R&D in year t first increase and then depreciate to zero after 16 or 35 years. Hall and Scobie (2006) tested some alternative ways of capturing the impact of a capital stock through time including the permanent inventory method (PIM) and the Koyck (1954) and Almon (1965) transformations which conserve degrees of freedom and reduce multicollinearity problems.…”

Public investment in agricultural R&D and extension

“…In their study of Australian broadacre agriculture, Mullen and Cox (1995) estimated returns to have averaged 15 per cent between 1952-1953and 1987-1988. Mullen (2007 also found that similar rates of return had persisted over the period 1952-1953 to 2002-2003. Some commentators have questioned the credibility of these findings.…”