Impact of risk on cost and revenue efficiencies

Abstract

This study focused on the inclusion of risk in efficiency measures to determine its impact on traditional efficiency scores. Previous research and theory suggests efficiency scores will be lower under risk and for risk averse individuals. Risk aversion may deter use of new production technologies and production levels may not be as high as under other risk preferences. Two data sets were used in the analysis. A panel data set of 256 farms from 1993-2010 was used to address the impact of risk measured as variability in outputs and downside risk on efficiency. A separate data set of 258 farms for 2008 was used with a corresponding risk preference score to determine the impact of risk preference on efficiency. The risk preference scores in the sample ranged from 5 to 86 where a smaller value represents stronger risk aversion. Data envelopment analysis was used to construct a nonparametric efficiency frontier and calculate cost- and revenue-based economic, overall, technical, allocative, and scale efficiency measures. Five inputs: labor, crop input, fuel, livestock input, and capital; and two outputs: crops and livestock were used in the analysis. The results focused on cost- and revenue-based economic efficiency. They showed that risk did affect average efficiency scores and is necessary to include in efficiency analysis. The average cost efficiency without risk was 0.6763. It increased to 0.7200 and 0.7018 respectively when cost efficiency was adjusted to recognize variability in outputs and downside risk. The average portion of cost inefficiency explained by variability in outputs was 28.06 percent. Downside risk explained 22.66 percent of cost inefficiency. The average revenue efficiency without risk was 0.7611 and increased to 0.8372 and 0.7811 when revenue efficiency was adjusted for variability in outputs and downside risk, respectively. Variability in outputs explained 42.53 percent and downside risk explained 30.58 percent of revenue inefficiency. The average cost efficiency for the 258 farms was 0.5691 and increased to 0.6043 with the consideration of risk preference scores. The average revenue efficiency was 0.6735 and increased to 0.6987 with risk preference scores. The efficient farms varied across cost and revenue efficiency, and the risk measures used. This lends support to the use of both input-oriented (cost) and output-oriented (revenue) efficiency measures as well as the use of multiple measures of risk.