Complementarity and substitutability between farm and nonfarm activities: Evidence from agricultural households in Tanzania

• Published date: 01 February 2023
• Author: Jehovaness Aikaeli,Martin Julius Chegere,John Rand
• Date: 01 February 2023
• DOI: http://doi.org/10.1111/rode.12942

REGULAR ARTICLE

Complementarity and substitutability between

farm and nonfarm activities: Evidence from

agricultural households in Tanzania

Jehovaness Aikaeli

1

| Martin Julius Chegere

1

| John Rand

2

1

School of Economics, University of

Dar es Salaam, Dar es Salaam, Tanzania

2

Department of Economics, University of

Copenhagen, Copenhagen, Denmark

Correspondence

Martin Julius Chegere, School of

Economics, University of Dar es Salaam,

Dar es Salaam, Tanzania.

Email: chegeremartin@gmail.com;

chegere.martin@udsm.ac.tz

Funding information

Danish International Development

Agency; Copenhagen University in

Denmark; University of Dar es Salaam;

Growth and Development Research

Project (GDRP)

Abstract

Increasing evidence shows that farm households in

developing countries maintain a portfolio of income-

generating activities. Concerns have been raised that

household income diversification to include nonfarm

activities will stifle farm works, and thus reduce agricul-

tural productivity. On the contrary, some evidence of

complementarity between the farm and nonfarm activi-

ties in rural areas has been found. This paper uses data

from the fourth and the fifth waves of Tanzania's

National Panel Surveys to examine whether there exists

complementarity or substitutability between farm and

nonfarm activities in rural Tanzania. The findings show

that participation in off-farm work and nonfarm business

activities has no clear relationship to the adoption and

use of improved agricultural practices, namely, irrigation,

fertilizers, pesticides, and herbicides. Participation in off-

farm work and nonfarm business activities does not also

seem to directly influence agricultural investments, produc-

tion, and productivity nor overall household income. These

results suggest that the complementary and substitution

effects of nonfarm activities offset each other in Tanzania,

implying that proper policies are required to unleash the

potential impact of nonfarm activities on agriculture.

KEYWORDS

agricultural assets, agricultural investment, farm household,

household income, nonfarm activities, panel data

Received: 7 August 2021 Revised: 6 August 2022 Accepted: 1 September 2022

DOI: 10.1111/rode.12942

Rev Dev Econ. 2023;27:89–111. wileyonlinelibrary.com/journal/rode © 2022 John Wiley & Sons Ltd. 89

JEL CLASSIFICATION

C23, O13, Q12

1|INTRODUCTION

The majority of sub-Saharan Africa (SSA) households engage in smallholder agriculture, and

they still depend on agriculture as their main income source (World Bank, 2015). This remains

true even after several decades of ongoing structural transformation and growth in urban

populations (Babatunde & Qaim, 2016; Gollin, 2014). Around 53% of the total SSA workforce is

currently registered within agriculture (World Bank, 2021). In such predominantly agricultural

economies, there has been a perception that farm households depend exclusively on agriculture

as their income source and undertake little or no nonfarm work. This idea dates back to the

early literature on development, viewing farm households as depending solely on their own

farming activities for their food security and overall livelihood. If farms did not meet the needs

of the household, members could resort to work for wages on other farms or migrate, which in

turn could bring remittances. While there were early theoretical work and some scattered

case studies on linkages between farm and nonfarm work (e.g., Hymer & Resnick, 1969), rural

nonfarm employment in local manufacturing and services (both as wage-employment and as

self-employment) features little in the literature.

In SSA, only a few rural households actually depend solely on agriculture for their livelihoods

(Ellis, 2007; Reardon, Berdegué, Barrett, & Stamoulis, 2006). Estimates show that about 52% of

households in SSA participate in nonfarm activities (Davis, Di Giuseppe, & Zezza, 2017)and

obtain their income by participating in both agricultural and nonagricultural activities (Nagler &

Naudé, 2017). Nonfarm income is an important contributor to the total household income in

SSA (Haggblade, Hazell, & Reardon, 2010; Van den Broeck & Kilic, 2019), and its share in

the total household income has been increasing and is expected to increase even further

in the coming years due to rapid population growth, limited improvements in agricultural pro-

ductivity, and climate change (Davis et al., 2017;Davis,Winters,Reardon,&Stamoulis,2009;

Haggblade, Hazell, & Reardon, 2007).

Researchers have also found that smallholders often maintain a portfolio of off-farm

income-generating activities (Barrett, Reardon, & Webb, 2001; Reardon, Crawford, & Kelly,

1994). But, some concern has been raised about whether diversification of household income

sources may stifle farm works and reduce agricultural productivity. On the contrary, there has

also been some evidence of complementarity between the farm and nonfarm activities. In the

case where income from nonfarm works acts as collateral for agricultural loans or where

income from nonfarm activities is being spent on agricultural inputs, nonfarm income can play

an important role in cushioning the liquidity constraint caused by little access to credit in devel-

oping countries (Collier & Lal, 1986; Deininger & Olinto, 2001). Further, seasonality in farming

allows some off-farm work to be undertaken during periods when labor demand in farming is

low and thus not necessarily substituting farm activities.

Most studies analyzing the effects of diversification of economic activities have used the ana-

lytical framework of “push”and/or “pull”factors, a framework that this paper also draws upon

(De Janvry & Sadoulet, 2001; Kilic, Carletto, Miluka, & Savastano, 2009). However, while impli-

cations of nonfarm activities on poverty and inequality have been studied in various developing

countries using this framework, much less research has been conducted on the effect of

90 AIKAELI ET AL.