Stuck exchange: Can cash transfers push smallholders out of autarky?

ABSTRACT

This paper focuses on the role of unconditional cash transfers in helping smallholders’ commercialization by overcoming barriers to trade from transaction costs. We use data from a controlled experiment for the evaluation of the Child Grant model in Zambia. We employ a Heckman model that allows us to capture the effects of the program on the propensity to engage in trade in both inputs and outputs markets as well as on the value of trade. The cash transfer program contributes significantly to increase farmers’ commercialization. The program produced greater benefits for those households that face more binding transaction costs from transportation and information gathering.

KEYWORDS:

• Cash transfers
• transaction costs
• commercialization
• selection model

JEL Classifications:

• C21
• Q13
• H24

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Since proportional transaction costs are a change in the unitary price, their impact on the threshold price is modeled by adding explicitly $t_p^s$ to the shadow price. For fixed transaction costs, the impact on the threshold price is modeled implicitly through the higher price ${\underset{\_}{p}}^s$ (for sellers) that is needed to compensate them for the reduced utility from fixed transaction costs. Formally, the new threshold condition for sellers, is $p^m>{\underset{\_}{p}}^s+t_p^s>\tilde{p}+t_p^s$.

2 Between baseline and follow-up data collection, the local currency, the kwacha, has been rebased at a rate of 1000 old kwacha = 1 new kwacha

3 A detailed description of the evaluation design can be found in American Institutes for Research (2011).

4 In fact, more than half of the differences come from price variables which are measured at the community level. The reason for the differences may be that the number of randomized communities me not be large enough to balance community-level covariates.

5 Although in the estimation we use the monetary values of marketed quantities of seeds and crops, for consistency with the exposition in the theoretical framework we keep the terminology of quantities here.

6 The marginal effects in the first stage represent changes in the probability of participation, while those in the marketed surplus equation express changes in the actual amount of sales for a unitary change in the regressor.