The health consequences of hazardous and nonhazardous child labor

• Published date: 01 May 2019
• Date: 01 May 2019
• Author: Alberto Posso
• DOI: http://doi.org/10.1111/rode.12571

REGULAR ARTICLE

The health consequences of hazardous and

nonhazardous child labor

Alberto Posso

Royal Melbourne Institute of

Technology, Melbourne, Victoria,

Australia

Correspondence

Alberto Posso, International Development

and Trade Research Group, Royal

Melbourne Institute of Technology,

Melbourne, Australia.

Email: alberto.posso@rmit.edu.au

Abstract

The health consequences of child labor are ambiguous. On

the one hand, heavy lifting, using dangerous tools and han-

dling fertilizers may impact negatively on health. On the

other, child labor could be used to achieve a minimal subsis-

tence standard, without which the child could experience

deteriorating health. Previous empirical studies are inconclu-

sive because, until now, existing data sources could not dis-

entangle between different activities performed by child

workers. To establish how work is related to health, it is

essential to know what activities are being performed. This

study fills this gap with a unique child labor survey con-

ducted in Peru in 2015. Child labor is classified into haz-

ardous and nonhazardous activities. The econometric results

suggest that childrendoing nonhazardous work are less likely

to have health problems than nonworking children. Con-

versely, thosethat work in hazardous activitiesare potentially

more likely to exhibit healthconcerns. A series of robustness

tests confirm these findings. The results give impetus to the

notion that child work increases total resources available for

households, which can improve child health. Therefore, it is

over and above this potential increment in resources that

some hazardousactivities worsen working children's health.

KEYWORDS

Child labour, hazardous activities, child health, Latin America, Peru

1

|

INTRODUCTION

Child labor is usually thought of as a set of dangerous activities, such as lifting heavy bags with agri-

cultural goods, which should have significantly negative health consequences. However, theoretically ,

DOI: 10.1111/rode.12571

Rev Dev Econ. 2019;23:619–639. wileyonlinelibrary.com/journal/rode © 2018 John Wiley & Sons Ltd

|

619

the health effects of child work are not obvious. On the one hand, the use of dangerous tools and

machinery, contact with fertilizers and pesticides, exposure to excessive water, and lifting and carrying

heavy materials should impact negatively on health. On the other hand, in very poor communities,

child labor could be used to obtain a minimal subsistence standard and without it the child could expe-

rience deteriorating health (O'Donnell, Rosati, & Van Doorslaer, 2005). Moreover, households may

decide to allocate more resources to their working household‐members to ensure their productivity and

health (Pitt, Rosenzweig, & Hassan, 1990; O'Donnell et al., 2005; Adhvaryu & Nyshadham, 2012).

It is because of this conjectural ambiguity that to understand the effect of child labor on chil-

dren's health economists have turned to empirics. There is only limited multivariate quantitative

research on the child health consequences of the phenomenon.

1

Roggero, Mangiaterra, Bustreo,

and Rosati (2007) use macroeconomic data for 83 countries to find that child labor, defined as the

percentage of working children, is significantly and positively associated with mortality rates of

10‐to 14‐year‐old children. Using macroeconomic data for 18 countries and a bivariate measure of

child labor, O'Donnell, Van Doorslaer, and Rosati (2002) find an undetermined relationship

between the frequency of children reporting health problems and child work. In some instances,

child labor is positively associated with child health, while in others the opposite is true.

Using Vietnamese panel data from 1992/1993 and 1997/1998, O'Donnell et al. (2005) find no

significant difference in weight between children participating in work and those that do not.

2

However, children in paid work were found to be significantly heavier than children working in

the household or not at all. They conclude that this may be owed to households allocating addi-

tional nutritional resources to income‐earning children. However, O'Donnell et al. (2005) also find

evidence of negative health consequences of child labor in the long run—work participation raises

the risk of illness 5 years later. Using the same panel, Beegle, Dehejia, and Gatti (2009) confirm

that child labor raises vulnerability to subsequent illness, although the negative repercussions do

not seem large enough to affect height growth. Taking a very similar approach, Wolff and Maliki

(2008) use Indonesian household data to show a positive relationship between child labor (defined

as either a binary variable or number of hours) and the probability of having a fever, a cough or

other health complaints (influenza, diarrhoea, breathing difficulty, etc.).

Overall, existing empirical evidence is mixed—both macro‐and micro‐level studies show vary-

ing evidence on the relationship between child labor and child health. This may be owed to previ-

ous empirical studies being unable to disentangle between the different activities performed by

child workers. To establish the mechanisms through which work affects health, it is important to

know what activities are being performed at work. Simple binary variables, the number of hours,

or even the approach taken in O'Donnell et al. (2005), who divide child labor into paid and non-

paid work, cannot disentangle between various activities performed by children. That is, previous

studies cannot differentiate between children working at a store after school selling goods and

those working in a construction site carrying bricks.

The premise in this paper is that not all child work is necessarily bad, it is, therefore, the activi-

ties and the conditions under which work is performed that can have damaging health conse-

quences (Rodgers & Standing, 1981; Patrinos & Psacharopoulos, 1997). This paper uses data that

categorizes work activities performed by children as either hazardous or nonhazardous. Hazardous

work occurs when children lift heavy things, work with dangerous machinery, or are exposed to

hazardous conditions, such as height, noise, and water. The data is sourced from a unique survey

conducted in Peru by its government and the International Labour Organization (ILO) in late 2015.

Peru is a relevant case study because it has a high prevalence of child labor, offering a unique

opportunity to study the health consequences of this phenomenon. It is important to note that this

study does not consider child work activities that concentrate purely on household chores. This

620

|

POSSO