Family and Hired Labour in Tanzanian Smallholder Crop Production: Evidence on Labour Input Interaction

Abstract

This study examined whether family and hired labour function as 
substitutes, complements, or independent inputs in Tanzanian 
smallholder crop production. Using data from 2,768 households in 
the 2022/23 Annual Agricultural Sample Survey, a translog 
production function with region fixed effects and an inverse 
hyperbolic sine (IHS) transformation was estimated. Results show 
hired labour (β = 0.0533, P < 0.01) and family labour (β = 0.0380, 
P < 0.05) contribute positively to output, but their interaction is 
insignificant (β = −0.0065, P = 0.571), indicating that the two 
labour inputs operate independently rather than as substitutes or 
complements. Land is the strongest driver (β = 0.7031, P < 0.01), 
followed by inorganic fertiliser (β = 0.6377, P < 0.01) and 
pesticide use (β = 0.0919, P < 0.10). Female-headed households (β 
= −0.1706, P < 0.01) and older household heads (β = −0.0868, P < 
0.01) are associated with lower output, while capital, organic 
inputs, irrigation, education, and agricultural training are 
insignificant. Labour-hiring households record higher average 
output (ln Y = 6.707) than non-hiring households (ln Y = 5.930). 
The findings remain robust under a Cobb–Douglas specification 
and the labour-hiring subsample. The study concludes that 
improving access to labour markets and reducing seasonal labour 
constraints may enhance smallholder agricultural productivity in 
Tanzania.