Nonparametric measures of the impact of public research expenditures on Australian broadacre agriculture

Nonparametric methods are used to measure the impact of public research expenditures on Australian broadacre agriculture over the 1953–94 period. Results using both unrestricted and 30‐year lagged specifications of the research impacts on productivity suggest that while certain aspects of the nonparametric multi‐input/output technologies are quite robust to alternative specifications (in particular, the associated Malmquist total factor productivity indexes), other aspects are less stable (in particular, the indexes on input and, to a lesser extent, output biased technical change). Internal rates of return to research expenditures on Australian broadacre agriculture are estimated to be in the 12 per cent to 20 per cent range.