The determinants and spill-over effects of tractor adoption in Tanzania

Abstract

After decades of persistently low use of agricultural mechanisation, the proportion of African farmers using tractors has rapidly increased since 2005. The conventional perception is that tractor use in sub-Saharan Africa is confined mainly to large-scale commercial farms. However, there is mounting evidence of the rapidly increasing use of tractors by smallholder farmers, albeit through tractor rental markets in geographically concentrated areas. To date, evidence on the causes of rising use of mechanisation in the SSA region, and particularly on smallholder farms, remains limited. This has often attributed to state-led efforts to prematurely promote mechanisation, which is rarely considered to be a market-driven response to changing factor prices. However, the recent economic transformations in many regions in sub-Saharan Africa may be providing these incentives. This study contributes to the information gap by utilising four waves of data from the nationally representative National Panel Survey from 2009, 2011, 2013 and 2015 to identify the factors driving the use by Tanzanian smallholders of mechanisation, and by examining whether there is evidence of the potential role of larger-scale farmers in promoting a movement to more capital-intensive forms of farming. The study aimed to develop a conceptual framework that interprets the important interactions between the primary and secondary drivers of agricultural transformation change in sub-Saharan Africa and their respective implications for technology adoption, such as the use of mechanisation. Selective elements of these interactions will be empirically investigated in the context of agricultural households in Tanzania. For instance, the study examines the relevance of the Hayami-Ruttan hypothesis of induced innovation in Tanzania and explores whether it upholds the importance of changes in factor prices. The research focuses on the changes in land dynamics, including farm size and land size distribution, and the corresponding tractor use in Tanzania by developing descriptive statistics over the period from 2009 to 2015. Generalised Linear Models (GLM) and Mundlak-Chamberlain Correlated Random Effects (MC CRE) models for tractor demand by households were formulated to empirically investigate important relationships, such as changes in factor prices, the role of medium-scale farmers in promoting a shift towards capital-intensive input use, and changing farm size and land size distribution. The demand model applications, through post-estimation analysis, were utilised to develop key policy principles. The results of the household tractor rental model, derived from the pooled GLM and MC CRE Probit estimation, indicated that land size is coupled with tractor use through rental services, particularly in the 5-9.99 hectare group. As wage rates increase, a shift towards renting a tractor would be considered. The opposite is true when the cost to rent a tractor is considered. These results uphold the importance of relative changes in factor prices, which is consistent with the induced innovation hypothesis. The study also found that the demands for mechanisation services through participation in rental markets have increased in areas with a higher density of medium-scale farms, which then enabled small-scale producers to rent tractors. In these higher density areas, there is an increasing number of small-scale farmers, cultivating less than 5 hectares, who make use of tractor rental services. This positive spill-over effect has the potential to reduce labour input requirements and facilitate a shift by small-scale farmers to non-farming activities that generate higher returns to labour, while still deriving income from farming. In the 21 regions of mainland Tanzania that are representative of a lower concentration of medium-scale farmers, only 3% of small-scale farmers have rented tractors. However, in the 5 regions with the highest density of medium-scale farms, 23% of smallholders have rented mechanisation services.