Agricultural mechanization often accompanies agricultural transformation. In some countries in Africa
south of Sahara (SSA), such as Nigeria, the mechanization process appears slow, in spite of the declining
share of the agricultural sector in the economy and employment. Knowledge gaps exist regarding this
slow mechanization process, and filling this knowledge gap is important in identifying appropriate
policies on agricultural mechanization in Nigeria.
In Nigeria, average horsepower and prices of tractors appear high, despite the scarcity of tractors.
These patterns are different from the experiences in other parts of the world where initially tractor
horsepower was often smaller (for example Asia), or farmers were better endowed with land and wealth
(the Americas). In Nigeria, joint ownership of tractors is rare, and formal loans are often unavailable due
to high transactions costs. IFPRI’s survey in Kaduna and Nasarawa states in 2013 also suggested that the
spatial mobility of tractors is generally low, and uses are highly seasonal (Takeshima et al. 2014). There
do not seem to be plausible explanations for the seeming dominance of large tractor use, based on the
available information of the prices and soils. Nevertheless, these patterns seem driven by private sector’s
own initiative rather than by governments’ policies.
Indivisibility of large tractors and limited mobility of supplies may cause the imperfection in the
custom hiring market. In order to distinguish the impacts of technology adoption at the extensive margin
from those at the intensive margin, in the empirical analyses we tested these hypotheses focusing on the
differences among marginal adopters of tractor hiring services and nonadopters of similar characteristics.
The results are two-fold: (1) adoptions patterns of tractor services are partly explained by basic factor
endowments, suggesting that the market for custom hiring is in some way functioning efficiently in
response to economic conditions; (2) adoptions are, however, affected by supply-side factors including
the presence of large farm households (and thus potential tractor owners) within the district, and (3) per
capita household expenditure level differs significantly between the marginal adopters and nonadopters of
similar characteristics. This difference seems to arise from the adoption per se, rather than the intensity of
adoption, which is consistent with the hypothesis of the imperfection of the custom hiring market.