Technical efficiency drivers of farmer-led restoration strategies, and how substantial is the unrealised potential for farm output?

In Ethiopia, similar to other parts of Sub-Saharan Africa, the effects of farmland degradation on ecosystem services and well-being are especially prominent as 85% of the population lives in rural areas and depends on agriculture for their livelihoods. Existing farmer-led restoration activities involve integrating trees on farms (ToF) through traditional agroforestry or woodlots systems. Although ToF practices can potentially bring economic and ecological gains, their production systems and efficiency in terms of input use remain overlooked in the current literature. The study analyses the technical efficiency levels in Ethiopia's home gardens and woodlot systems and examines the drivers contributing to technical efficiency. The study estimates the technical efficiency of farmers who implement the ToF strategies, i.e. the extent to which the farmers can achieve the highest output level given their set of inputs. Deviation from the total technical efficiency is then referred to as unrealised potential. We also analyse the drivers of technical inefficiency by controlling for geographical, biophysical, and production system attributes of woodlots and agroforestry patches, as well as household characteristics and institutional settings. We use a one-step Latent Class Stochastic Frontier Modelling approach based on the cross-sectional household data of 543 home gardens and 426 woodlot systems patches in Ethiopia. According to the results, in woodlots, extensive farms lose 38% and intensive farms 22% of the potential output. In home-gardens, intensive farms lose 39%, and extensive farms lose 49% of the output due to technical inefficiency. Significant determinants of technical efficiency in both systems are ecological zones, access to markets, communal carbon payments, forest user groups, training, land title factors, rotation age and product diversification, among others. Our results help identify priority policy and practical action fields for improving farmer-led restoration strategies' technical efficiency and management in Ethiopia and the global south. Future national reforestation policies could reduce inefficiencies in farmer-led systems across Sub-Saharan Africa by targeting the factors exacerbating technical inefficiency per class in a given strategy. Thus, handling these drivers increases the success of farmer-led restoration strategies and contributes to achieving restoration pledges and AFR100 commitments in mosaic landscapes. [Display omitted] • We examined technical efficiency (TE) of farmer-led restoration strategies, Woodlots and Home-gardens. • To control heterogeneity, we used a Latent Class Stochastic Frontier Model to estimate TE and its drivers. • We obtain mainly two classes in both systems, labour-intensive and labour-extensive farms. • Unrealised potential ranges between 22 and 49% and differs among classes within a restoration system. • Geographical, household, institutional and product system drivers contribute to model estimation and affect class TE. [ABSTRACT FROM AUTHOR]