Heterogeneity in returns to agricultural technologies with incomplete diffusion: Evidence from Ghana

• Published date: 01 February 2022
• Author: Sadick Mohammed,Awudu Abdulai
• Date: 01 February 2022
• DOI: http://doi.org/10.1111/rode.12837

Rev Dev Econ. 2022;26:323–353.

|

323

wileyonlinelibrary.com/journal/rode

Received: 25 February 2021

|

Revised: 7 September 2021

|

Accepted: 8 September 2021

DOI: 10.1111/rode.12837

REGULAR ARTICLE

Heterogeneity in returns to agricultural

technologies with incomplete diffusion:

Evidence from Ghana

SadickMohammed

|

AwuduAbdulai

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and

reproduction in any medium, provided the original work is properly cited.

© 2021 The Authors. Review of Development Economics published by John Wiley & Sons Ltd.

Department of Food Economics and

Consumption Studies, University of Kiel,

Kiel, Germany

Correspondence

Sadick Mohammed, Department of Food

Economics and Consumption Studies,

University of Kiel, Kiel, Germany.

Email: smohammed@food-econ.uni-kiel.

de; msadick@uds.edu.gh

Funding information

The first author acknowledges funding

from the Deutscher Akademischer

Austauschdienst (DAAD) as part of

the Ghanaian- German Postgraduate

Training Programme, 2017 (funding ID

no. 57344816)

Abstract

In this study, we employ a dynamic treatment effect

approach to analyze heterogeneity in returns to farm-

ers at different stages of adoption of a newly introduced

inoculant technology, using a recent survey data of 600

soybean farmers from northern Ghana. Although farm-

ers differ in their returns to adoption of new technolo-

gies, many empirical studies often fail to account for this

heterogeneity. The empirical results reveal that farmers

who are at advanced stages of adoption appear to, on

average, more than double their yields and farm net re-

turns, suggesting that the inoculant technology may be

a game changer in the fight against extreme poverty in

the region, where poverty is endemic and crop yields

are persistently below the average potential yield tar-

get. Our findings further reveal that extension services

as well as efficient input and output markets are key to

the adoption process, by influencing knowledge acqui-

sition, adoption, and continued adoption. Our findings

also show significant impact heterogeneity at each adop-

tion stage, with the long- term benefits of the inoculant

technology outweighing its short- term benefits.

324

|

MOHAMMED and ABDULAI



|

INTRODUCTION

Low agricultural productivity and perennial food insecurity are major global concerns facing

low- income countries, particularly countries in sub- Saharan Africa (SSA). Central to tackling the

problem are increasing crop yields and sustaining gains through the adoption of improved agri-

cultural technologies (Takahashi et al.,2020). Yet the technology adoption rate among farmers

in these countries appears to be very low (Macours,2019; Sheahan & Barrett,2017; Suri,2011).

While some analysts partly attribute the phenomenon to factors such as lack of information, low

education, and credit constraints, others question the empirical and theoretical adoption models

used to analyze farmers’ adoption decisions (Besley & Case, 1993; Feder etal., 1985; Lindner

et al., 1982). In particular, Besley and Case(1993) note that technology adoption is a dynamic

process in which farmers make a series of decisions over multiple stages or seasons. Lindner

etal. (1982) succinctly summarized the adoption process into three broad categories: discovery

stage, evaluation stage, and trial stage. Each stage in the adoption process collects different sets

of vital information for the farmer to update subsequent decisions. However, classical studies on

technology adoption mostly consider farmers’ adoption decisions as static, ignoring the dynamic

processes embedded in farmers’ decision- making. As a result, important information on farmers’

adoption behavior relevant to policy formulation is lost, and their decisions are misinterpreted.

Thus, it is not uncommon for analysts to find farmers’ adoption decisions at odds with rationality

and sometimes counterintuitive (Besley & Case, 1993).

This study departs from the classical approach and analyzes farmers’ adoption decisions in

a dynamic framework. Previous studies that examined farmers’ technology adoption decision-

making in a dynamic framework mainly focused on adoption determinants, patterns of diffu-

sion, and intensity of adoption (e.g., Abdulai & Huffman,2005; Feder & Slade,1984; Lambrecht

etal., 2014; Simtowe etal., 2016), while some studies employed it to explain farmers’ learning

behavior, risk preferences, and uncertainties (Ghadim & Pannell, 1999). The missing link in the

dynamic adoption literature is the impact of adoption on output levels and other welfare indica-

tors, such as yields and farm net returns, which underlie farmers’ adoption and continued adop-

tion decisions. These indicators also drive adoption patterns and clarify risks and uncertainties

that may surround a given technology (Besley & Case, 1993; Feder etal.,1985). We contribute to

the literature by analyzing farmers’ adoption decision- making process as a multi- stage decision

problem and how adoption impacts on farm outcomes. One in which each stage of adoption is

characterized by different margins of payoffs or gains that accrue to farmers at that stage. We

apply this approach to analyze farmers’ adoption decisions of a new Rhizobia inoculant tech-

nology among 600 soybean farmers in northern Ghana, considering that farmers’ returns from

adoption may be heterogeneous and stage dependent.

Few studies in the technology adoption literature have analyzed heterogeneity in returns to

adoption of agricultural technologies (Abdul Mumin & Abdulai,2021; Shahzad & Abdulai,2021).

KEYWORDS

dynamic treatment effect, impact heterogeneity,

inoculanttechnology adoption, multistage decision- making

JEL CLASSIFICATION

C32; D83; O33; Q10; Q16