The empirical application is the Norwegian salmon aquaculture industry for which we
have a large farm level panel data set for the period 1985–93. Salmon aquaculture is a
good candidate for this type of analysis since it is characterized by substantial production
risk due to largely uncontrollable shocks in terms of fish diseases, sea temperature changes,
and frequent occurrences of toxic algaes and adverse weather conditions. Although the
occurrence of these shocks can not be prevented by fish farmers, it is believed that their
effect on output can be influenced through choice of input levels. The empirical analysis
in this paper will provide estimates of the risk properties of inputs in salmon farming, thus
giving us empirical knowledge on which inputs reduce the level of output risk and which
inputs increase risk