Estimating the influence of extraction method and processing location on forest harvesting efficiency - A categorical DEA approach

Abstract

The increasing level of competition in the global forestry market demands that stakeholders continuously measure their performance with the aim of remaining competitive and profitable in the ever-changing wood market. This study applies categorical data envelopment analysis (DEA) methodology to the New Zealand forest harvesting sector. This methodology is able to account for ordinal non-discretionary variables in the DEA. The influence of log extraction method and processing location on the estimated efficiency scores were examined. To define the forest harvesting DEA production technology, three inputs (harvest area, average piece size, level of mechanization), one output (tons/scheduled hour) and one categorical non-discretionary variable with three levels were used. The categorical variables were defined by the level of difficulty as reported by harvest supervisors for specific forest harvesting operating environment. The study demonstrated the appropriateness of the categorical DEA approach in measuring performance in forest harvesting operations. It showed significant influence of timber extraction methods on the overall performance estimate, whereby grapple skidders at 58% had the highest mean efficiency score. While log processing locations showed no significant influence on the estimated performance, processing at the stump had the highest mean efficiency score.