Smallholder farm management under evolving land institutions: Evidence from Zambia

In Zambia, clearing land for cultivation is a reported mechanism for making land claims in contexts of uncertain rights and expropriation risk, and leaving land fallow is less common for those with weaker land rights. When only some farmers in a village obtain land title, the land security of untitled landholders, particularly women, can become less secure in contexts where there are fears that chiefs can reallocate land away from untitled smallholders to obtain financial gains from seekers of titled land. On the other hand, if village-level titling signals a general transition towards more transparent regulation and/or more favorable developmental prospects for the area, then local trends in titling may contribute to an increased shared sense of tenure security, even for non-title-holders. Motivated by these hypothesized effects, we examine whether neighborhood titling rates are positively or negatively associated with smallholders leaving land virgin or fallow, after controlling for own-titling status. We show that both own-title status and neighboring titling rates are positively associated with having farmland in virgin and fallow states, i.e., reduced incidence of land clearing, indicating that community titling rates are associated with increased security and expected future returns on uncultivated land. However, these results vary by gender: community titling rates are less strongly associated with decreased land conversion by female household heads (except in the case of leasehold titling) who often have weaker bargaining rights in customary systems. Additionally, we show that, after controlling for one’s own land titling status, neighbors’ titling is associated with greater rental market participation and more longer-term productivity investments. These indirect effects are stronger for leasehold titling as compared with customary land certification, and for titles which have been finalized (compared with titles still being processed). Our analysis indicates that tenure formalization policies may have both direct and indirect effects which differ by type of actor.