Veterinary expenditure has almost no effect on expected output but has a negative effect on variance, implying that increases in this input will tend to reduce production risk. Cows increase variance but this effect is smaller than that on the mean function, so increases in the number of cows reduce relative risk. Increases in the other two inputs – feed and fertiliser – on the other hand would tend to reduce welfare for risk-averse producers as they increase relative risk.
The final column of Table 4 shows the overall average effects of a scale increase. An increase in scale will lead to a greater proportional increase in the mean function compared with the variance function, so the riskiness of output will decrease. From this we can draw some policy implications. For example, in the EU there are voluntary abandonment schemes and similar type measures to encourage increases in farm size. These measures have been promoted with the aim of increasing average productivity. To evaluate such policies, not only the effect on average productivity (which is expected to be moderate) should be taken into account, but also their effect on risk. In the case of the sector we study, increases in farm size will lead to additional gains in the form of a reduction in output risk, thus providing further justification for promoting policies to increase scale.
The TFP growth rates corresponding to equation (8) and the changes in production risk as per equation (14) for the five annual periods are presented in Table 5, which also presents the changes in inputs. The first block of Table 5 shows how inputs have changed over the period, and it can be seen that farms have expanded over time in that, with the exception of 1996/1997 and the use of fertiliser in 1994/1995, all inputs have risen for all years. Turning to the evolution of observed TFP growth, two aspects are worth highlighting. First, the scale effect, which captures the effect of input growth on observed TFP growth, is relatively small. This is to be expected given that we have overall constant returns to scale. The exception is the final year, where the scale effect is much larger. This latter result is influenced by large changes in input use by a few farms which had significantly strong increasing returns to scale. Second, the impact of technical change is generally much stronger than the scale effect. Finally, observed TFP growth was positive in all years except 1995/1996, where there was a strong negative technical change effect.