Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 1598-7248 (Print)
ISSN : 2234-6473 (Online)
Industrial Engineering & Management Systems Vol.18 No.3 pp.511-519

Why Does a Child Labor Exist in Aceh Province, Indonesia?

T. Zulham, Said Muhammad, Diana Sapha, Fitriyani, Jumadil Saputra*
Department of Economics, Faculty of Economics and Business, University of Syiah Kuala, Indonesia
School of Social and Economic Development, Universiti Malaysia Terengganu, Kuala Nerus, Terengganu, Malaysia
Corresponding Author, E-mail:,
June 3, 2019 June 17, 2019 June 20, 2019


Child labor is a pervasive issue around the world, especially in developing countries. It is a wide phenomenon issue in which many countries have a similar problem related to child labor. The purpose of this study is to examine the factors determining child labor in Aceh Province. The data used is the primary data of the National Social Economic Survey (SUSENAS) 2016 conducted by the Central Bureau of Statistics. The dependent variable is children aged 10-17 years, while the independent variables are the educational level, the role of child as the head of the household, gender, residence area, single parent status, and age. The logistic regression model is used to describe the relationship between the independent variable (child labor and not child labor) with a number of explanatory variables. The result indicates that there is a relationship between child labor and all variables used in this model. Nonetheless, status as head of household and single parent woman are irrelevant as the significance value (α) is less than 5% by using the Wald Test Statistic. Thus, it is recommended to do some improvements in young people’s education system. Moreover, government policies that protect the human rights of child labors need to be done as a form of government intervention.



    Child labor is a wide phenomenon in which many countries have similar issues related to child labor especially, developing country like Indonesia.

    International Labour Organization (2015) reported that the number of child labors was significantly decreasing in these past ten years. In 2012, there were an expected 167.956 million child labors. It declined from the previous period which was about 215.209 million. Even though the number of child labors keeps falling, the number is fairly high. There are still 167.956 million children who are working and going through hardship even if they are not obligated toward it. Moreover, an estimate of 85.344 million was engaged in hazardous work. Mohamud (2016) argues that child labor occurs at an age where the child grew, resulting in losing nutrition, education, and leisure. Besides, International Labour Organization (2016) also states that employing children is one of the greatest acts of exploitation and violence in children. Children are required to work in the worst conditions and were being exploited in where they worked full time or part time with less salary.

    Some countries set a minimum working age of 12 until 13-years-old, below the standard set by UNICEF, which was 18 years old. Children are forced to work, part time or full time. They are required to help to support family income. Additionally, many people prefer to hire child labor because of cheap and less demanding salary. In fact, almost half work full time and the rest work parttime (Ahmad and Ahmad, 2018). In addition, Fateminasab (2014)Binder and Scrogin (1999), Betcherman et al. (2004) states that there are four reasons why child labor becomes an issue. First, child labor has less welfare than non-working children. Time used to play and study with friends is contrarywise used for working. Second, there are several child labors who do not continue their education because they are busy doing their job which cause higher dropout rates in children. Third, as more children withdraw from school, the rate of economic growth will slow down in long term. Yet, long-term economic growth depends heavily on human capital investment in children. Hence, it will reduce the level of educational participation in children which will give a negative impact on economic growth in the long term. Fourth, child labor will be forced to support their families. Working and earning an income, though it will motivate the children, it will affect their mental state because they have to carry a heavy burden that they should not have experienced yet.

    In addition, there are significant differences between working children and those not working in school attendance. The results of Demir et al. (2006), Lancaster and Ray (2003) suggest that the average grade in the classroom, as well as the attendance rate of working children, is lower than non-working children. This shows that by deciding to work, children do not have time to study and rarely attend a class which is the main job to do, thus affecting the length of the child’s learning level. Moreover, some studies believe that poverty is the most dominant factor that causes a child to decide to work. Grimsrud (2003), Binder and Scrogin (1999) believes that many children decide to work because children would benefit more than those who are only in school and do not work. These benefits are not only economic benefits, but also an opportunity to learn skills that do not acquire in school. In addition, many children are forced to work to pay for their schooling. Due to the lack of family economic resources, the child will be encouraged to work.

    Likewise, the complete family also influences a child’s decision to work. The divorced woman is mentally injured resulting in lack of resources and supports from the family causing the child to become the head of the family instead Scrogin (1999). This is similar to Ahmed (2018) research in which female as head of household tends to earn less income so that children have to work and help to support their parent. The absence of formal learning from parents during childhood also leads children to leave school and move to the labor market. Moreover, even if they combine both school and work, it will not be effective. It is difficult for the child to distinguish between ideal and nonideal conditions due to lack of basic education from parents (Burki et al., 1998;Ersado, 2003;Khan, 2003;Bakhyt et al., 2018).

    Several studies also show that an increasing number of family members leads to a greater chance for children to work and school than just school. This is because every child does not get the same learning opportunities. Oldest children are more likely to work and go to school or only work than younger children. This is in line with research (Khan, 2003;Khanam and Rahman, 2007;Mohamud, 2016). Gender and residence are also one of the factors that determine children to work. Mohamud (2016) expresses that children living in rural areas are more liable to work more than those living in urban areas and girls are more likely to go to school and work than boys. This is not consistent with research by Hamid and Ahmed (1994) in which boys are assumed to have a 1.3 greater chance of becoming child labors than girls. Likewise, Ersado (2003) concluded that in Peru, boys are more likely to work than women. Figure 1

    It is remarkable to address this issue due to the fact that many people still do not highly concern about this matter. Additionally, incomplete data regarding child labor and its determinants makes it hard to get the best view about the real number and real condition of child labor. There are still many child labors hidden and does not yet discover as the awareness of this issue is not as big as other macroeconomics and labors issues. Assuredly, eliminating child labors is not an easy task. However, Improvement of the quality of human resources to eliminate child labor in the future is required because child labor will become a new leader and has great potential to develop a country. Thus, to get an excellence human resources quality, it heavily depends on education. Lower education will make it challenging for kids to compete in the labor market. Furthermore, working at an early age will definitely affect the child’s education and other factors. It is undoubtedly will affect the children’s future as well. The circle of poverty will be hard to evade and it is even harder for a country to enhance their means (Afriyani et al., 2018;Kanashiro et al., 2018).

    This study is looking into another perspective on the child labor issue. The trade-off between work and school or a combination of school and work is interesting to study. Thus, in this study, it is necessary to stress the difference in the characteristic of working children and nonworking children. Furthermore, it is expected in 2020- 2030, there will be a “window of opportunity” for Indonesian demography as people who are in the labor force are increasing and has the highest percentage. However, it is necessary for people who are in the labor force to have higher skill and productivity. Without skill and ability, the labor will just be a burden to a country. Consequently, the unemployment rate will rise and will severely impact economic performance as well. Therefore, it is urgent to do research about this issue and examine which factors that are important to diminish the child labor since today’s child labor is the future of “window of opportunity” in the future. To the best of our knowledge, this research is a pioneer to determine factors affecting a child decides to work. It is also important to cover the most essential variables from those.

    Inhabitants are one of the main principals for regional development to achieve higher per capita income. However, the population can become a problem if not balanced with the excellent quality of human resources. Figure 2 portrays the number of percentages of Child Labor in Aceh. Aceh Province has a differing number of child labor widened in several regions. The percentage of child labors in Aceh Jaya is the smallest which are approximately 0.55% of the whole child in working age (10 and above). This number is still very small compared to the province of Aceh by 2.11%. Child Labors in Aceh Jaya scattered spread in various urban and rural areas where children work in various fields of work including Agriculture, Fisheries, Trade, and Industry. On Contrary, the percentage of child labors in Aceh Tenggara is the highest which are about 5.35% of the total population at work age (10+). This far exceeds the provincial average percentage at 2.11%.


    Variables used in this study are working children or non-working children, age, sex, the occupation of the head of household, job sector, depth of poverty, and residence location. The data are the raw data of the National Social Economics Survey (Susenas, Statistik, 2016) in 2016 at Aceh Province implemented by the Central Bureau of Statistics in September 2016. The selection of samples is done by two stages (two-stage sampling). In the first phase, 7,500 census blocks are selected by systematic sampling of 30,000 district blocks according to the allocation and distribution of samples per stratum at the district level. The second stage selects 10 households through systematic sampling with the implicit stratification of the highest education attained by the head of the household. For the Aceh region, the total samples of selected child labor are 7,263 households. Thus, Child laborers are defined as children aged from 10 to 17 which are working and earning an income no less than 1 hour during the past week repeatedly. The model used in this study is based on a simple model from Khandker et al. (1994) and Mason and Khandker (1997), namely:

    S i   =   S ( I i ,   H i ,   C i )

    W i   =   W ( I i ,   H i ,   C i )

    where S and W are the outcome variables (school or work of i-th children). i is a vector of individual characteristics (age, the age of squared, non-linearity, assuming a child is head of household or not, hypothesized by rarely are children as heads of households). Moreover, H is a vector of household characteristics (demand for labor from households and C is a vector of public characteristics (assuming that households live in rural or urban areas and how much the distance to primary and secondary schools. As a result, because the dependent variable is a child with dichotomy values 0 and 1, a logistic regression model is used (Howell: 2001). The logistic regression model in general for k variables is:

    π ( x ) = exp ( β 0 + β 1 x 1 + β 2 x 2 + ... + β k x k ) 1 + exp ( β 0 + β 1 x 1 + β 2 x 2 + ... + β k x k )

    where π(x) is the probability of a successful event, whereas β 0 , β 1 , β 2 , ... , β k are the parameter values. The function π(x) is a nonlinear function so that a logit transform is required to attain a linear function in order to see the relationship between the dependent variable and the explanatory variables. The logit transformation from π(x) becomes into:

    β 0 + β 1 x 1 + β 2 x 2 + ... + β k x k = g ( x )

    g(x) is a linear function of its parameters.

    Additionally, the child labor logistic regression function in this study can be written as follows

    W i = β 0 + β 1 D S E K i + β 2 D K R T i + β 3 D W I L i + β 4 D K E L i + β 5 D O R i + β 6 D U M U i + u i

    where Wi is a child aged 10-17 years, 1 if working and 0 if not working. DSEKi is the school status if the child is not in school 1, and if school 0. DKRTi is the child as head of household, 1 if the child as head of household and 0 if not. DWILi is a residential area, 1 if rural and 0 if urban. DKELi is gender, 1 if female and 0 if male. DORTi is 1 if they live with single parent women 1 and 0 if not. DUMUi is the age of the child, expressed in years. u is an error term. i is the child to i = 1, 2, 3, ..., n


    3.1 Child Labor Characteristics

    From the National Social Economics Survey (Susenas) data (see Table 1), there are approximately 414 children who are classified as child labor and 5.70 percent of them work and most (15.94 percent) of them are out of school. Meanwhile, children who did not go to school and also did not work reached, 85 people. Subsequently, the dropout rate has reached 1.17 percent. Furthermore, 97 percent of children are still in school and another 2.08 percent or 151 children do not go to school anymore. Poverty is still regarded as the main cause. Likewise, child-ren who are not attending school are forced to work and gain income. Children entering this group reached 43.71 percent. In contrast, the majority of children who are still attending the school choose not to work. Additionally, child labor who drops out from school generally work for a long period of time compared to another and most of them work within normal working hours (> 35 hours).

    Only a small number of children are responsible as head of the household and involuntary to work. On the contrary, children who still have parents are rarely working. They earn about IDR533,590. per month and capable to work at least 28 hours per week. Alternatively, single female parent status also affects the child’s decision to work and even some of them have to abandon their education. Approximately, 7.99 percent of children who only have a female parent is child labor. Risk of a single-parent child to quit from school is greater than those who still have both of their parents and male single parent. Small average income per capita is one of the sole reasons for the parent to force their children in assisting their work and earn higher pay. Moreover, about 22.22 percent of this group has to work at least 35 hours per week. In addition, rural areas are home to the majority of children or about 72.44 percent and the number of child labor is about twice as child labor in urban areas. Around 88.08 percent of children who are not attending school live in suburban areas and trigger the growth of child labor. The relatively low income per capita which is approximately IDR658,595 per month result in the children being forced to work to increase their family’s income and most have worked at normal working hours or even 35 hours per week.

    Gender percentage for child labor is higher for male than female. Boys who are working have 67.15%. This is higher than girls which are only approximately 32.85%. At the same time, the proportion of children who are not working evenly balanced between men and women with 51.23% and 48.77% respectively. For boys about 7.34 of them work and only 3.91 percent of girls. For female labors, about 38.24 percent of them is working under 15 hours per week. It is greater than boys which are 29.14 percent. Hence, the average per capita income per month for girls is above average, which is IDR782, 976, compared to boys which earn IDR765,183. Furthermore, a higher the age, the higher the probability to be child labor. Children who are 17 years old and 16 years old have a higher chance to work. It means that the proportion of children as child labor tends to enlarge with age and the time spent in working. This is due to the amount of awareness and responsibility to the family. In addition, physical and non-physical abilities are also better than younger child labor. Consequently, the income given is also getting higher (see Table 2).

    3.2 The Result of the Dependency Test among Variables

    The null hypothesis (H0) is that there is no dependence between child labors and education, child labors and status as head of household, child labors and gender, child labors and residence area, child labors and single female parent, and child labors and age, otherwise the alternative hypothesis (H1) is that there is a dependence. With a 95 percent confidence level (α = 5%), it is assumed that H0 is rejected if the value of Asymp. Sig. (2- sided) in column (5) is smaller than 0.05. Table 3

    The correlation between child labor and other determinants is demonstrated in Table 2. This shows that the relationship between child labor and school year are higher compared to other variables. While the relationship between child labor and status as head of household is the lowest which is only about 2.40 percent. Table 4

    3.3 Logistic Regression Analysis

    The result shows that as the p-value of α is less than 0.05, thus we reject the null hypothesis (Ho). It can be inferred that the logistic regression model is significant and can be used in this model. In addition, the sample data used can also explain this logistic model by 94.45 percent. Therefore, the household income of the logistic regression model in Aceh Province can be written as:

    g(x) = -9.452+1.806DSEK+1.456DKRT +0.522DWIL-0.68DKEL + 0.217DORT + 0.442XUMU

    The table shows the probability of the children to be child labor based on the dummy variables and age characteristics. The interpretation number of 0.00651 on DSEK = DKRT = DWIL = DKEL = DORT = 0, illustrates a child who is 10 years old, educate; not as head of house-hold; lives in an urban area; male sex; has parents instead of a single parent woman; and thus has 0.65 percent chance to do work. The number of 0.12670 on dummy variables and is 17 years old, indicates that a child with the same characteristics has a probability to be child labor by 12.67 percent. On the other hand, for a child who is 10 years old and have contrary characteristics (DSEK = DKRT = DWIL = DKEL = DORT = 1) describes a child who is not in school, bears responsibility as head of household, lives in rural area, female, has a single parent woman, and has a chance to become a child labor by 15.32 percentage. Whereas, the chance of someone with the same characteristics but already 17 years old is very high which is nearly 80.02 percent. Table 5

    3.4 The Result of the Wald Test Statistic

    The regression model asserts a substantial effect. However, it is necessary to test each of the parameters by using the Wald Statistic test following the distribution of χ2 (0.05; 1). The Wald test resulted indicates that most of the variables significantly influence children decided to work. Nonetheless, status as head of household and single parent woman is irrelevant as the significance value (α) is less than 5%. Table 6

    3.5 Odds Ratio

    Odds Ratio is the ratio of the tendency of a successful event to the failure event. The Exp value (β) for the school years variable of 6,088 means that a child who is not going to school has a higher opportunity to become child labor by 6 times greater than a child who is attending the class, assuming other variables are constant. Moreover, a child who is responsible as a head of household has a tendency to be child labor four times more than those that are not the head of the household. Responsibility as head of the household will push a child to join the labor market and earn additional income. Furthermore, a child living in rural areas tends to be child labor by 1,7. Cultural differences and improvements in knowledge and technology preserve the children to join the labor market earlier. They also have better revenue than rural communities, so they can provide a better education for their children. Growth disparities between these areas also cause the increasing of child labor in rural areas.

    In addition, the possibility for a girl to be child labor is slightly smaller than the boy by 0.5. This is because a boy is more obliged to raise the household’s income. Girl incline to work at home to do the household chores, such as washing, cooking, cleaning the house, which does not provide income. Likewise, a child who lives with only one parent is expected to be child labor than a child who has parents. Single parent woman is likely to send their children to work and assist them in working to get a higher wage. Finally, age strongly influences the propensity of a child to become a worker. A child with a one-year-older is 1.56 more likely to join the workplace than a one-yearyounger. This is due to the fact that their responsibility is bigger as they grow old. Moreover, their capability and skill in the workplace are also better. Accordingly, employers prefer to cast older children than another.


    There are several reasons that can be inferred for why child labor exists. Poverty is one of the main problems that influence children’s decision to work. Living in poverty obliges children to join the labor market earlier. Children are more likely to be in low paying jobs and some join for unpaid family jobs. Due to the financial problem, some child labors even drop out of school as they cannot access education anymore. Moreover, the fact that they are not in school shows that poverty has been going on from one generation to the next generation. It is convinced that child labor is part of a cruel-poverty cycle. Additionally, the importance of this issue is related to the fulfillment of human rights. Children are being exploited and some even work in a hazardous place and abuse in their workplace. Government policies that protect the human rights of child labors need to be done as a form of government intervention. That is by applying school participation for children through compulsory education (12 years) and continued with compulsory education with free charge or no cost. Provision of allowances for poor households with children, households with single female heads of household enables children to learn and attend the school without worrying about their family conditions. Equal educational programs and quality between urban and rural areas and gender should be enforced across agencies and institutions, as well as socializing equal rights and obligations of boys and girls. Likewise, complete data about child labor is still limited. There are so many child labors hidden and cannot be identified. Research and assessment of child labor problems also can be developed by looking at other variables, such causes of drop out, educational performance, and student’s award. Thus, improvement in young people’s education system is necessary to boost the welfare of the community in the long term. It is due to the fact that children are the investment for the future’s development. However, not only does child labor affect children’s future but better children’s prospect it also will affect child labor as well. More people educated will obviously reduce and even diminish child labor in the future.



    Number of child labors and children doing hazardous work.


    The percentage of child labors in aceh.


    Child labor in Aceh statistic (Susenas data)

    Child labor and age

    Result of dependency test

    The relationship between variables

    Result of probability: Child labor based on dummy variables and age characteristics

    The result of wald statistic test


    1. Afriyani, D. , Sa’dijah, C. , Subanji, S. , and Muksar, M. (2018), Characteristics of students’ mathematical understanding in solving multiple representation task based on solo taxonomy, International Electronic Journal of Mathematics Education, 13(3), 281-287.
    2. Ahmad, I. , & Ahmad, S. (2018) . Multiple Skills and Medium Enterprises’ Performance in Punjab Pakistan: A Pilot Study. The Journal of Social Sciences Research, Special (4), 44-49.
    3. Bakhyt, S. , Kalimbetov, B. , and Khabibullayev, Z. (2018), Possibilities of mathematical problems in logical thinking: Development of secondary education pupils, Opción, 34(85-2), 441-457.
    4. Betcherman, G. , Fares, J. , Luinstra, A. , and Prouty, R. (2004), Child labor, education, and children’s rights, Social Protection Discussion Paper Series No.0412.
    5. Binder, M. and Scrogin, D. (1999), Labor force participation and household work of urban schoolchildren in Mexico: Characteristics and consequences, Economic Development and Cultural Change, 48(1), 123-154.
    6. Burki, A. A. , Fasih, T. , and Din, M. (1998), Households’ non-leisure time allocation for children and determinants of child labour in Punjab, Pakistan, The Pakistan Development Review, 37(4), 899-914.
    7. Demir, C. E. , Demir, E. , and Uygur, S. (2006), The relationship between work, school performance and school attendance of primary school children in Turkey, Proceedings of the European Conference on Educational Research, University of Geneva, Geneva.
    8. Ersado, L. (2003), Child labor and schooling decisions in urban and rural areas: Cross-country evidence, Paper prepared for presentation at the American Agricultural Economics Associatilahiion Annual Meeting, Montreal, Canada.
    9. Fateminasab, A. (2014), Investigating the challenges and barriers of convergence between Iran and republic of Azerbaijan, UCT Journal of Social Sciences and Humanities Research, 4(1), 18-24.
    10. Grimsrud, B. (2003), Millennium development goals and child labour, understanding the child labor in developing countries: A challenge children’s work an inter agency, Research Cooperation project of ILO, UNICEF and World Bank Group, Project. Available from:
    11. Hamid, S. and Ahmed, K. (1994), A Micro analysis of urban child labour: Some determinants of labour and its conditions, The Pakistan Development Review, 33(4), 1249-1271
    12. International Labour Organization (2015), World report on child labour 2015: paving the way to decent work for young people, International Labour Organization.
    13. Kanashiro, L. , Ribeiro, A. , Silva, D. , Meirelles, P. , and Terceiro, A. (2018), Predicting software flaws with low complexity models based on static analysis data, Journal of Information Systems Engineering & Management, 3(2), 17.
    14. Khan, R. E. (2003), Children in different activities: Child schooling and child labor, The Pakistan Development Review, 42(2), 137-160
    15. Khanam, R. and Rahman, M. M. (2007), Child work and schooling in Bangladesh: The role of birth order, Journal of Biosocial Science, 39(5), 641-656.
    16. Khandker, S. R. , Lavy, V. , and Filmer, D. (1994), Schooling and cognitive achievements of children in Morocco: Can the government improve outcomes?, The World Bank.
    17. Lancaster, G. and Ray, R. (2003), Does child labour affect school attendance and school performance? Multi country evidence on SIMPOC data, Unpublished Report, ILO/IPEC.
    18. Mason, A. and Khandker, S. (1997), Household schooling decisions in Tanzania, World Bank (mimeo).
    19. Mohamud, A. H. (2016), Child labor and school attendance in Somalia, A Research Paper Submitted in Partial Fulfillment of the Requirements for the Award of the Degree of Masters of Arts in Economic Policy Management of the University of Nairobi.
    20. Statistik, B. P. (2016), Data Susenas 2016, Jakarta (ID): Badan Pusat Statistik Indonesia.
    Do not open for a day Close