Productive Efficiency of Small Scale Sawmilling Industries in Mufindi District, Tanzania

  • JM Abdallah
  • DA Woiso
  • GC Monela
  • R Phillip


This study was carried out to examine the productive efficiency of small-scale sawmills in Mufindi District. The specific objectives were: (i) to assess the relative efficiency of small-scale sawmills in Mufindi, (ii) to identify factors for variation in the small scale sawmill’s relative efficiency and (iii) to provide policy recommendations for efficiency improvement in utilization of forest resources. A structured questionnaire was used to collect data from 80 small-scale sawmills in Mufindi District. Data were analysed using descriptive as well as quantitative methods. Technical, scale and allocative efficiency score of sawmills were computed using data envelopment analysis programme developed by Coelli. Censored regression models were estimated to identify factors for inefficiency of smallscale sawmills. Results showed a mean technical efficiency of 84% (CRS model) and 92% (VRS model), allocative efficiency of 84% (CRS model) and 89% (VRS model) and cost efficiency of 70% (CRS model) and 81 % (VRS model). Furthermore, results from the censored regression model revealed that Owners/manager’s education, experience, the size of the sawmill timber yard and partnership ownership had positive effects on sawmill’s efficiency while machine age had a negative effect on sawmill’s efficiency. Recommendations for enhancing small-scale sawmills production efficiency are: Strengthening extension services to increase sawmilling experience, and insistence on partnership ownership of sawmills. Since mill size positively enhanced sawmills’ relative efficiency, an
increase of the size of mills must receive priority. Squeezing sawmill area as a result of increased number of sawmills leads to inefficiency. Lastly but not least, the use of sawmilling by products particularly chips, saw dusts and slabs should be promoted as it is a way of increasing efficiency in sawmilling.

Journal Identifiers

eISSN: 2408-8137
print ISSN: 2408-8129