One of the most frequently used methods of measuring production performance is to construct a total factor productivity (TFP) index to measure productivity growth. Increases in productivity as measured by such an index will generally be considered as ‘beneficial’ to the producer in the sense that gains in productivity will be expected to contribute positively to producer welfare. If we focus solely on the effect of productivity in the form of technical change on expected output, this will generally hold. However, technical change can also affect production risk, which will in turn affect welfare if producers are not risk-neutral. Consider, for example, the adoption of new technologies. A change in production technology which yields productivity gains through positive technical change may at the same time lead to higher production risk. Risk-averse producers will weigh both these aspects of the new technology before taking a decision as welfare may be adversely affected despite positive technical change. Similarly, technologies which may not have a significant effect on expected output may substantially reduce risk. As an example, the adoption of a new pesticide may not affect expected output but may lead to a significant decrease in output variability, thereby increasing producer welfare. Note here, moreover, that observed productivity, measured by a ratio of output to inputs, may actually fall in this case if pesticide use increases.