Parental absence and child labor in vietnam

UNIVERSITY OF ECONOMICS ERASMUS UNVERSITY ROTTERDAM HO CHI MINH CITY INSTITUTE OF SOCIAL STUDIES VIETNAM THE NETHERLANDS VIETNAM – THE NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS PARENTAL ABSENCE AND CHILD LABOR IN VIETNAM BY NGUYEN NGOC MINH THU MASTER OF ARTS IN DEVELOPMENT ECONOMICS HO CHI MINH CITY, November 2017 UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES HO CHI MINH CITY THE HAGUE VIETNAM THE NETHERLANDS VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS PARENTAL ABSENCE AND CHILD LABOR IN VIETNAM A thesis submitted in partial fulfilment of the requirements for the degree of MASTER OF ARTS IN DEVELOPMENT ECONOMICS By NGUYEN NGOC MINH THU Academic Supervisor: NGUYEN HOANG BAO HO CHI MINH CITY, November 2017 ACKNOWLEDGEMENT I would first like to thank my thesis supervisor Dr Nguyen Hoang Bao of the Vietnam – The Netherlands Programme (VNP) at Ho Chi Minh City University of Economics I much appreciate for all his dedication, his guiding, supporting, and patience to my study He has set a great example of a teacher, researcher, especially for me I would like to express my gratitude to the VNP officers who were involved in my thesis process by updating thesis schedule and providing good condition for my research process Without their passionate participation, the thesis process could not have been successfully conducted Finally, thanks are also due to my classmates for providing me with unfailing support and continuous encouragement throughout my years of study and through the process of researching and writing this thesis This accomplishment would not have been possible without them Thank you Nguyen Ngoc Minh Thu Ho Chi Minh City, November 2017 ABSTRACT Child labor is defined as a task which had took out the childhood from children; have an ability to affect negatively on child physical and mental development These tasks also obstruct child schooling by disposing child from the chance of learning at school or force them to leave school early In a worst form, child labor contains child who serve as slaves, has been splitting with their parents, high opportunity to get serious illness and have to protect themselves on the street (International Programme on the Elimination of Child Labour (IPEC) ILO (2013) also estimated about one – third of the child labors in Vietnam, or around 569,000 children have to work around 42 hours per week in 2012 and this will have a bad influence on child schooling or children have to stop completely schooling time for work This paper uses a panel data built up from VHLSS 2012 with descriptive statistics of parents and children characteristics to determine whether parent’s absence is a main reason for child labor We apply Probit and OLS regression, as well as Heckman selection model, to examine the impacts of parents’ characteristics on the probability of child working and working hours The main finding of this study emphasize that parent’s absence is not a key factor lead to child labor It is economic factors and poverty that play an important role in leading child labor status and child working hours CONTENTS ACKNOWLEDGEMENT ABSTRACT CHAPTER 1: INTRODUCTION 1.1 Child labor 1.2 Parents time for children 11 1.3 The Impaction of Parents on child labor 12 1.4 Research objective 12 1.5 Structure of the thesis 13 CHAPTER 2: LITERATURE REVIEW 14 2.1 Review of theory 14 2.2 Effect of parents on children’s skill development .19 2.3 Effect of Parents on children education 20 2.4 Effect of parents on children’s labour 20 2.5 Effect of poverty on child’s labor 21 CHAPTER 3: RESEARCH METHODOLOGY 22 3.1 Conceptual framework: 22 3.2 Child labor status (working or not) – logit model 24 3.3 Child working hours – multiple linear regression model 25 3.4 The logit model Error! Bookmark not defined 3.5 Heckman selection model 27 CHAPTER 4: EMPIRICAL RESULTS 28 4.1 Overview of child labor problem in Vietnam 28 4.2 Data 30 4.3 Empirical results 41 CHAPTER 5: CONCLUSION 47 5.1 Conclusion 47 5.2 Policy implications 47 REFERENCE 49 LIST OF TABLES Table 1: Variable description of child working or not regression 24 Table 2: Variable description of child working hour regression .26 Table 3: Descriptive Statistics of children’s variable 31 Table 4: Descriptive Statistics of father’s variables 32 Table 5: Descriptive Statistics of mother’s variable 33 Table 6: Frequency of social-demographic characteristics .34 Table 7: Child’s working general information 35 Table 8: Parents living in family characteristics by child labor status 39 Table 9: Parents income characteristics by child labor status 40 Table 10: Parents illness characteristics by child labor status 40 Table 11: Regression result .43 LIST OF FIGURES Figure 1: Child labor affects the nation, source: Srinivas Viswanathan, 2014 Figure 2: Global Incidence of Child Labor, 2000 to 2012, source: Our World in Data Figure 3: Incidence of Child Labor between SEA countries, 2000 to 2012, source: Our World in Data 10 Figure 4: Incidence of working children who work only in five SEA countries 11 Figure 5: The relationship of household, parent and child characteristics to child labor .23 Figure 6: Vietnam child population by age group and gender, source: Vietnam National Child Labor survey, 2012 29 Figure 7: Frequency of child labor status by gender 36 Figure 8: Frequency child labor status by ratio of educated child in household 37 Figure 9: Percentage working status of child by first child in household status 37 Figure 10: Frequency of child labor status by child’s reasoning to hospital 38 Figure 11: Percentage working status of child by father living together status 38 Figure 12: Percentage working status of child by mother living together status 39 Figure 13: Frequency of father education level by child labor status 41 Figure 14: Frequency of mother education level by child labor status 41 CHAPTER 1: INTRODUCTION 1.1 Child labor Cigno and Rosati (2005) has announced that children all over the world are related to a large number of activities which could be able classified as work These activities include a large range of activities, from less harm for children as home activities to the most damaging job as soldiering or prostitution ILO (2013) has estimated in World Report that 265 million cases of child labor have been found, which is nearly 17 per cent of children all over the world According to this report, the highest number of child labor happened in Sub-Saharan Africa (Ospina & Roser, 2014) Most of child labor cases has found in developing countries where children often work under manufacturing or agriculture factors According to ILO’s Statistical Information and Monitoring Program, agriculture is the sector in which child labor could be found the most Edmonds and Pavcnik (2005) has estimated that in Nepal, 85% of child labor is in agriculture, Cambodia 73% and in Morocco 84% Figure 1: Child labor affects the nation, source: Srinivas Viswanathan, 2014 Viswanathan (2014) presented the figure that child labor is undesirable because of its negative affect on the nation as a whole As indicated in Figure 1, child labor would disturb child schooling time and also their cognitive development This would affect child’s professional expertise and their soft skill which will lead to poor wage later in their working age Low wage caused bad working environment and poor labor standards These conditions will lately cause poor income or even unemployment when children grow up Figure presented child labor aged 5-17 which has been reduced from 23 per cent on 2000 to 17 per cent on 2012 (Our World in Data, n.d.) From 2000 till 2012, the number of child labor around 78 million which has reduced almost one-third The ratio of reducing in girl labor as 40 per cent while the ratio falling in boy labor is 25 per cent From 2008-2012, the reducing in child labor is greater than 2000-2008, among of that, the Asia and Pacific region has contributed the large number of reducing on fours year 2008-2012 (Making progress against child labor- Global estimates and trends 2000-2012, 2013) Figure 2: Global Incidence of Child Labor, 2000 to 2012, source: Our World in Data Together with South Sudan and Suriname, Vietnam has been listed in 74 countries which have serious problem of child labor by the US labor Department (List of goods produced by child labor and forced labor, 2016) In the same report, most of Southeast Asian countries such as Malaysia, Indonesia, Cambodia, and Thailand are included in the list Our World in Data (n.d.) has shown as Figure the Incidence of Child labor figure a trend of child labor which aged from 7-14, however the figure not show all the result at the same 10 time so that the dot line presents as: Cambodia has the rapid reducing from the beginning of around 60 per cent to 11.5 per cent of child labor in 2012, Vietnam also have a large decreasing from more than 20 per cent on 2006 to 10.9 per cent in 2012, Thailand: 15.1% of child labor in 2005-2006, Philippines has a small different from 2001 to 2011 as their number of child labor in beginning is already low around 12% on 2001 to 9% on 2011 Lastly, Indonesia have a stable trend on child labor from 2000 to 2009 and drop incredibly on 2010 as 3.7 per cent Figure 3: Incidence of Child Labor between SEA countries, 2000 to 2012, source: Our World in Data Even the overall cases of child labor reduce; the percentage of children who work only increased in Vietnam and Indonesia in Figure Vietnam has the percentage of children who work only increased from 15.9 per cent in 2011 to 19 per cent in 2012 For Indonesia the figure is 11 per cent in 2009 and increases to 44.4 per cent in 2010 Child labor in others countries such as Thailand and Philippines remain unchanged, or slightly increased as Cambodia -0.566 -591.364*** (140.681) 0.096 37.112 (36.470)  Parent characteristics Father Age (year) Mother Age (year) Father living In-house (month/year) Mother living In-house (month/year)  Child characteristics Child Gender -0.014* (0.008) 0.014* -0.004 (0.008) 0.034 0.004 (0.056) 0.004 (0.095) 0.011 0.001 -34.055 (63.885) 98.736 (73.442) -0.002 67.932 (52.652) -0.008 (0.056) Child age 47.019*** (14.676) 68.864*** Child wage (thousand VND/ hour) Subsidy for Child Education (million/year) Out-patient Times (per year) In-patient Times (per year)  Parent Working Father Income (million/year) Mother Income (million/year) Father Working Hours (hours/month) Mother Working Hours (hours/month)  Cost Father In Patient Cost (million/year) Mother In Patient Cost (million/year) Child School Cost (million/year) 0.242*** (0.048) -0.110*** (11.673) 0.080 (0.026) -12.085 (26.231) -23.299 (87.296) 0.102 (0.102) -0.008*** (0.002) -0.002 -0.187 (2.474) -0.019*** (0.004) -0.0006 -0.006 -3.125 (4.458) (0.0005) 0.0002 (0.0004) -0.000 0.000 -0.004 (0.000) -0.009 (0.017) -0.309*** (0.038) -0.001 -0.003 2.087*** (0.536) 1.101** (0.497) - 4.015 (13.003) 28.189 (21.386) -86.870* (48.263) It is also stated in the research of Saga (1975) that the parent’s characteristics such as education, occupation, income status would have related to children labor In this research, we look further some laws or programs have a strong significant to child labor such as: subsidy with the opposite impact on child labor status Table 12 presents the estimation results from the Heckman selection model Hypothesis testing results indicate that the two equations (selection and regression equations) are not correlated (see Appendix) and as a result, estimating the two equations separately does not result in biased estimates The results of Heckman selection model presented in Table 12 is very consistent and close to those from the Probit and OLS regression presented in Table 11 CHAPTER 5: CONCLUSION 5.1 Conclusion This study examines the leading role of parent’s absence on child labor in Vietnam based on dataset of VHLSS 2012 The regression analysis emphasizes three main findings as: Parents’ absence not effect on child labor even on the decision of pushing child to the labor market or not and child working hours Instead of parents’ absence, parents’ income are factors play a leading role on the child labor with father or mother income have a same highest significant to child labor status This result also has been presented by Cigno and Rosati (2005) as child labor elasticity with poverty because all the child activities as schooling, free time would be trade off with a cost Beside parents’ characteristic, children’s characteristics such as age, health condition (out-patient time) and household characteristic such as ratio of educated children, no of child in household also have a strong significant with child working hour This finding has been stated in the same research as young as children have to work, as high as percentage of children would have an bad effect on health both in present or in later- their adult life Although parents’ absence is not the leading role cause child labor but these are definite a key factor on children development All decisions related to children consumptions, daily activities as schooling or working are taken by their parents since young children are dependent agents (Cigno & Rosati, 2005) On the other hand, this study might not have a full dataset and this study just only consider the situation on one year 2012 Even 2012 is the year have full information or child labor report from ILO and some organization but the analysis based on a long period might help to predict a trend of child labor better 5.2 Policy implications: Child labor rate of Vietnam in 2012 is lower than the average point of the world (ILO, 2014) since all the policies on poverty alleviation help to control the poverty – key factor caused child labor Policies for a broad-based poverty decrease include the expansion of income earning opportunities, develop human capital and reduce risk… (Vietnam Academy of Social Sciences – VASS, 2011) Beside policies for poverty reducing, programs target to children such as easy school access will help to keep maintaining and increase child education time These program might include also schooling fee, education materialize and easy way to go to school in daily A program to secure child education against the risk of household poverty is needed In addition, programs and projects which increase the awareness of community about the negative of child labor will help parents understand and take decision better for children long term development REFERENCE International Programme on the Elimination of Child Labour (IPEC) (n.d.) What is child labor Retrieved from http://ilo.org/ipec/facts/lang en/index.htm Esteban Ortiz-Ospina and Max Roser “Child labor” Retrieved from https://ourworldindata.org/child-labor/ Srinivas Viswanathan Child labor presentation Retrieved from https://www.slideshare.net/srinisrg/chiild-labour Family undergoing major shifts in Viet Nam, shows first-ever nationwide survey on the family (2008, June 26) Retrieved from: https://www.unicef.org/vietnam/media_16537.html Children the losers in modernisation (2008, June 27) Retrieved from Vietnam News: http://vietnamnews.vn/society/177910/children-the-losers-in-modernisation.html Speech at the launch of the Survey of the Family in Viet Nam by Ms Maniza Zaman, Deputy Representative of UNICEF (2008, June 26) Retrieved from: http://www.un.org.vn/en/unicef-speeches/549-speech-at-the-launch-of-the-survey-of-thefamily-in-viet-nam-by-ms-maniza-zaman-deputy-representative-of-unicef.html List of goods produced by child labor or forced labor (2016) Bureau of international labor affairs United States department of labor Marking progress against child labor- Global estimates and trends 2000-2012.(2013) International Labor Office Result of nation-wide survey on the family in Viet Nam 2006.(2008) Ministry of culture, sports and tourism Nguyen Thi Lan Huong “Vietnam national child labor survey 2012- Main findings.” ILO Cataloguing in Publication Data (2014) Alessandro Cigno and Furio Camillo Rosati "The Economics of Child Labour." Oxford Scholarship Online (2005) Claus C Poărtner "Effects of parental absence on child labor and school attendance in the Philippines." Springer Science +Business Media New York (2014) 50 Akhtar Abdul Hai Ambreen Fatima Mahpara Sadaqat " Socio-economic conditions of child labor" International Journal of Social Economics (2010) Shamma Adeeb Alam “Parental health shocks, child labor and educational outcomes: Evidence from Tanzania” Journal of Health Economics 44 (2015) Carlo Alcaraz, Daniel Chiquiar , Alejandrina Salcedo “Remittances, schooling, and child labor in Mexico” Journal of Development Economics 97 (2012) Amarakoon Bandara, Rajeev Dehejia and Shaheen Lavie-Rouse “The Impact of Income and Non-Income Shocks on Child Labor: Evidence from a Panel Survey of Tanzania” World Development Vol 67 (2015) Kaushik Basu and Pham Hoang Van “The Economics of Child Labor” American Economic Association (1998) Christelle Dumas “ Why Do Parents Make Their Children Work? A Test of the Poverty Hypothesis in RuralAreas of Burkina Faso” Oxford Economic Papers (2007) Huong Thu Le, Ross Homel “The impact of child labor on children’s educational performance: Evidence from rural Vietnam” Journal of Asian Economics 36 (2015) Michele Binci and Gianna Claudia Giannelli “Internal versus International Migration: Impacts of Remittances on Child Labor and Schooling in Vietnam” International Migration Review (2016) Ralitza Dimova, Gil S Epstein, and Ira N Gang “Migration, Transfers and Child Labor” Review of Development Economics (2015) Elizabeth Kaletskia and Nishith Prakash “Does Political Reservation for Minorities Affect Child Labor? Evidence from India” World Development (2016) Diana I Kruger “Coffee production effects on child labor and schooling in rural Brazil” Journal of Development Economics 82 (2007) Carolyn M Moehling “Family structure, school attendance, and child labor in the American South in 1900 and 1910” Explorations in Economic History 41 (2004) 51 Guy B Nkamleu, Anne Kielland “Modeling farmers’ decisions on child labor and schooling in the cocoa sector: a multinomial logit analysis in Cˆote d’Ivoire” Agricultural Economics 35 (2006) Iliana Reggio “The influence of the mother's power on her child's labor in Mexico” Journal of Development Economics 96 (2011) M Najeeb Shafiq “Household schooling and child labor decisions in rural Bangladesh” Journal of Asian Economics 18 (2007) Ellen Webbink, Jeroen Smits, Eelke de Jong “Household and Context Determinants of Child Labor in 221 Districts of 18 Developing Countries” Soc Indic Res (2013) Francesca Francavilla, Gianna Claudia Giannelli “The Relation between Child Labour and Mothers’ Work: The Case of India” Iza (2007) Franc¸ois Charles Wolff , Maliki “ Evidence on the impact of child labor on child health in Indonesia, 1993–2000” Economics and Human Biology (2008) AnokhiParikh and Elisabeth Sadoulet “The Effect of Parents’ Occupation on Child Labor and School Attendance in Brazil” University of Oxford, Department of Economics (2005) APPENDIX Appendix 1: The summary statistics of numerical variables Variable Obs Mean Std Dev Min Max Edu 6026 4.32 2.85 11 NoofChild 6026 1.72 0.8 Edu1 5354 6.98 3.73 12 WorkDays1 5352 20.34 7.2 30 YIncome1 5344 15218.15 27687.78 394097 PTimes1 5301 0.95 2.95 81 Edu2 5887 6.12 3.92 12 WorkDays2 5886 19.39 8.86 30 YIncome2 5883 7296.5 18477.4 328413 PTimes2 5878 1.37 2.98 48 Appendix 2: The summary statistics of category and dummy variables Labor LivingF Freq Percent Cum 5,392 89.48 89.48 634 10.52 100 Total 6,026 100 749 12.43 12.43 5,277 87.57 100 Total 6,026 100 LivingM Firstchild RatioofEC 154 2.56 2.56 5,872 97.44 100 Total 6,026 100 1,929 32.01 32.01 4,097 67.99 100 Total 6,026 100 266 4.41 4.41 1/4 0.07 4.48 1/3 23 0.38 4.86 1/2 185 3.07 7.93 2/3 111 1.84 9.77 3/4 37 0.61 10.39 4/5 0.08 10.47 5,395 89.53 100 Total 6,026 100 STATA RESULTS The regression results of OLS reg labor Edu LivingF LivingM NoofChild RatioofEC firstchild Edu1 WorkDays1 YIncome1 PTimes1 Edu2 W > orkDays2 YIncome2 PTimes2 note: LivingM omitted because of collinearity Source SS df MS Model Residual 103.336817 370.57238 13 5140 7.94898593 072095794 Total 473.909197 5153 09196763 labor Coef Edu LivingF LivingM NoofChild RatioofEC firstchild Edu1 WorkDays1 YIncome1 PTimes1 Edu2 WorkDays2 YIncome2 PTimes2 _cons 0174438 0650914 0373695 -.4400898 0743853 -.0032023 -.0003837 -3.05e-07 -.0013888 -.0121073 0002469 -2.34e-07 -.0008629 3671005 Std Err .0014702 0604138 (omitted) 0058423 0174882 0103951 0014304 0006152 1.68e-07 0013567 0014249 0005073 2.44e-07 0014426 0653533 t Number of obs F( 13, 5140) Prob > F R-squared Adj R-squared Root MSE P>|t| = = = = = = 5154 110.26 0.0000 0.2181 0.2161 26851 [95% Conf Interval] 11.86 1.08 0.000 0.281 0145615 -.0533455 020326 1835282 6.40 -25.16 7.16 -2.24 -0.62 -1.82 -1.02 -8.50 0.49 -0.96 -0.60 5.62 0.000 0.000 0.000 0.025 0.533 0.069 0.306 0.000 0.626 0.337 0.550 0.000 0259161 -.4743742 0540065 -.0060065 -.0015898 -6.35e-07 -.0040486 -.0149007 -.0007476 -7.13e-07 -.003691 2389802 048823 -.4058054 0947641 -.0003981 0008223 2.44e-08 0012709 -.0093139 0012415 2.44e-07 0019651 4952208 The regression results of Probit * WORKING OR NOT: THE LOGIT MODEL logit labor FMonths1 FMonths2 YIncome1 YIncome2 Age1 Age2 IPCost1 IPCost2 head1 /// > Monthhours1 Monthhours2 Subsidy firstchild RatioofEC NoofChild gender OPTimes IPTimes SchCost,r Iteration 0: log pseudolikelihood = -1701.4775 Iteration 1: log pseudolikelihood = -1477.4399 Iteration 2: Iteration 3: log pseudolikelihood = -1307.4596 log pseudolikelihood = -1235.7539 Iteration 4: log pseudolikelihood = -1230.5131 Iteration 5: log pseudolikelihood = -1230.4843 Iteration 6: log pseudolikelihood = -1230.4843 Logistic regression Log pseudolikelihood = -1230.4843 Number of obs = 5140 Wald chi2(19) = 590.32 Prob > chi2 = 0.0000 Pseudo R2 = 0.2768 Robust labor Coef Std Err z P>|z| FMonths1 FMonths2 YIncome1 0764581 -.0146663 -.0143955 1253734 1556533 0043613 0.61 -0.09 -3.30 0.542 0.925 0.001 -.1692692 -.3197411 -.0229435 3221855 2904086 -.0058476 YIncome2 -.0425462 0112466 -3.78 0.000 -.0645892 -.0205032 Age1 -.0228134 0151183 -1.51 0.131 -.0524448 006818 Age2 IPCost1 0222973 -.0085472 0155094 0152994 1.44 -0.56 0.151 0.576 -.0081006 -.0385335 0526952 0214391 IPCost2 -.0345084 0442163 -0.78 0.435 -.1211708 052154 head1 4403358 2355231 1.87 0.062 -.021281 9019526 Monthhours1 Monthhours2 -.0012132 0005022 0008612 0007073 -1.41 0.71 0.159 0.478 -.0029011 -.000884 0004747 0018884 Subsidy firstchild 3908649 1.571144 0882219 1394234 4.43 11.27 0.000 0.000 2179532 1.297879 5637765 1.844408 RatioofEC -2.923389 1988984 -14.70 0.000 -3.313222 -2.533555 NoofChild 5282748 0628953 8.40 0.000 4050023 6515473 gender OPTimes -.023959 -.2217683 1066443 0705196 -0.22 -3.14 0.822 0.002 -.232978 -.3599841 1850601 -.0835525 IPTimes SchCost 2024208 -.7282065 2097731 1223696 0.96 -5.95 0.335 0.000 -.208727 -.9680465 6135685 -.4883664 _cons -1.739144 2.049114 -0.85 0.396 -5.755333 2.277045 Note: 18 failures and successes completely determined [95% Conf Interval] 56 Probit regression – Marginal effect * Marginal effects mfx Marginal effects after logit y = Pr(labor) (predict) = variable 03521325 dy/dx Std Err z P>|z| [ 95% C.I ] X FMonths1 0025975 00426 0.61 0.542 -.005761 010956 11.894 FMonths2 YIncome1 -.0004983 00529 -0.09 0.925 -.010858 009862 11.9691 -.0004891 -.0014454 00016 00036 -3.08 -4.03 0.002 0.000 -.0008 -.000178 -.002149 -.000742 14.896 7.41303 -.000775 00053 -1.47 0.141 -.001808 000258 42.3623 0007575 00054 1.41 0.159 -.000297 001812 39.5152 -.0002904 -.0011724 00052 00148 -0.56 -0.79 0.578 0.428 -.001314 -.004071 000734 001727 281068 248845 YIncome2 Age1 Age2 IPCost1 IPCost2 head1* 0127448 00596 2.14 0.032 001064 024426 905253 Monthh~1 Monthh~2 -.0000412 0000171 00003 00002 -1.40 0.71 0.162 0.477 -.000099 -.00003 000017 000064 151.925 139.143 Subsidy firstc~d* 013279 0036 3.68 0.000 006216 020342 157017 0447298 00576 7.77 0.000 033444 056016 676848 Ratioo~C NoofCh~d -.0993171 0179472 01398 00299 -7.11 6.01 0.000 0.000 -.12671 -.071924 012094 023801 932101 1.73074 gender OPTimes IPTimes SchCost -.000814 00362 -0.22 0.822 -.007916 006288 1.48074 -.0075342 0068769 0025 00714 -3.02 0.96 0.003 0.336 -.012425 -.002643 -.007126 020879 659728 038327 -.0247396 00285 -8.67 0.000 -.030334 -.019145 1.51521 (*) dy/dx is for discrete change of dummy variable from to Heckman regression * HECKMAN TWO STEPS *heckman hour wage Edu Age NoofChild RatioofEC firstchild FMonths1 FMonths2, /// > *select(labor = Edu FMonths1 FMonths2 NoofChild RatioofEC firstchild Edu1 YIncome1 Edu2 YIncome2) t > wostep * HECKMAN MAXIMUM LIKELIHOOD heckman hour wage Age NoofChild RatioofEC FMonths1 FMonths2 YIncome1 YIncome2 /// > IPCost1 IPCost2 head1 gender OPTimes IPTimes SchCost Monthhours1 Monthhours2, /// > select(labor = FMonths1 FMonths2 YIncome1 YIncome2 Age1 Age2 IPCost1 IPCost2 head1 /// > Monthhours1 Monthhours2 Subsidy firstchild RatioofEC NoofChild gender OPTimes IPTimes SchCost) Iteration 0: log likelihood = -5343.4507 Iteration 1: log likelihood = -5343.3488 Iteration 2: log likelihood = -5343.3309 Iteration 3: log likelihood = -5343.3308 Heckman selection model Number of obs = 5140 (regression model with sample selection) Censored obs = 4612 Uncensored obs = 528 Wald chi2(17) = 195.24 Prob > chi2 = 0.0000 Log likelihood = -5343.331 Coef Std Err z P>|z| [95% Conf Interval] hour wage 68.86435 11.67312 5.90 0.000 45.98546 91.74323 Age 47.01981 14.67699 3.20 0.001 18.25344 75.78619 NoofChild 37.11222 36.47017 1.02 0.309 -34.36801 108.5924 RatioofEC -591.3648 140.6817 -4.20 0.000 -867.0958 -315.6338 FMonths1 -34.05514 63.88521 -0.53 0.594 -159.2679 91.15756 FMonths2 98.73628 73.44211 1.34 0.179 -45.20762 242.6802 YIncome1 187771 2.474505 0.08 0.940 -4.66217 5.037712 YIncome2 -3.12588 4.458748 -0.70 0.483 -11.86487 5.613104 IPCost1 -4.015432 13.00344 -0.31 0.757 -29.50171 21.47084 IPCost2 28.18058 21.38496 1.32 0.188 -13.73317 70.09433 head1 -134.2243 120.5015 -1.11 0.265 -370.403 101.9543 gender 67.93223 52.65287 1.29 0.197 -35.26551 171.13 OPTimes -12.08578 26.33132 -0.46 0.646 -63.69421 39.52266 IPTimes -23.29997 87.29647 -0.27 0.790 -194.3979 147.798 SchCost -86.8703 48.26382 -1.80 0.072 -181.4657 7.725053 Monthhours1 2.087788 5365496 3.89 0.000 1.03617 3.139406 Monthhours2 1.101699 4978577 2.21 0.027 1259157 2.077482 _cons -357.978 1177.211 -0.30 0.761 -2665.27 1949.314 FMonths1 0348146 0565642 0.62 0.538 -.0760492 1456785 FMonths2 0047545 0957421 0.05 0.960 -.1828965 1924056 YIncome1 -.0082402 002237 -3.68 0.000 -.0126248 -.0038557 YIncome2 labor -.0194015 0041409 -4.69 0.000 -.0275174 -.0112856 Age1 -.0141481 0080905 -1.75 0.080 -.0300051 0017089 Age2 014396 0083643 1.72 0.085 -.0019976 0307897 IPCost1 -.0042819 0091667 -0.47 0.640 -.0222482 0136844 IPCost2 -.0092398 0170884 -0.54 0.589 -.0427324 0242527 head1 2481089 1281336 1.94 0.053 -.0030284 4992461 Monthhours1 -.0006424 0005191 -1.24 0.216 -.0016598 0003749 Monthhours2 0002099 000462 0.45 0.650 -.0006957 0011155 Subsidy 2423634 04898 4.95 0.000 1463644 3383625 firstchild 8284861 0698555 11.86 0.000 6915719 9654004 RatioofEC -1.704879 1008426 -16.91 0.000 -1.902526 -1.507231 3581572 NoofChild 289697 0349293 8.29 0.000 2212368 gender -.0089399 0563454 -0.16 0.874 -.1193749 1014952 OPTimes -.1103979 0267785 -4.12 0.000 -.1628828 -.0579129 IPTimes 101797 1012589 1.01 0.315 -.0966668 3002608 SchCost -.309821 0382082 -8.11 0.000 -.3847077 -.2349342 _cons -1.050929 1.27805 -0.82 0.411 -3.555861 1.454002 /athrho /lnsigma -.194902 6.372894 202084 0401657 -0.96 158.67 0.335 0.000 -.5909793 6.29417 2011754 6.451617 rho -.192471 1945978 -.5305996 1985047 sigma 585.7503 23.52705 541.4065 633.7261 lambda -112.7399 116.9489 -341.9556 116.4757 LR test of indep eqns (rho = 0): chi2(1) = 0.76 Prob > chi2 = 0.3846 58 Heckman regression- Marginal effects * Marginal effects, the selection (labor) margins, dydx(FMonths1 FMonths2 YIncome1 YIncome2 Age1 Age2 IPCost1 IPCost2 head1 /// > Monthhours1 Monthhours2 Subsidy firstchild RatioofEC NoofChild gender) predict(psel) atmean Conditional marginal effects Model VCE : OIM Expression Number of obs = 528 : Pr(labor), predict(psel) dy/dx w.r.t : NoofChild RatioofEC FMonths1 FMonths2 YIncome1 YIncome2 IPCost1 IPCost2 head1 gender Monthhours1 Monthhours2 Age1 Age2 Subsidy firstchild at : wage = 8132828 (mean) Age = 13.1875 (mean) NoofChild = 1.929924 (mean) RatioofEC FMonths1 FMonths2 = = = 6964331 (mean) 11.95455 (mean) 11.97538 (mean) YIncome1 YIncome2 = = 4.840604 (mean) 1.55983 (mean) IPCost1 = 2464621 (mean) IPCost2 head1 gender = = = 1497803 (mean) 9469697 (mean) 1.458333 (mean) OPTimes IPTimes = = 3333333 (mean) 0435606 (mean) SchCost = 4912841 (mean) Monthhours1 Monthhours2 Age1 = = = 137.1136 (mean) 127.6326 (mean) 42.53598 (mean) Age2 Subsidy = = 40.51894 (mean) 320089 (mean) firstchild = 7727273 (mean) dy/dx Delta-method Std Err z P>|z| [95% Conf Interval] NoofChild RatioofEC FMonths1 FMonths2 0963421 -.5669768 011578 0015812 0120884 0396257 0188143 0318397 7.97 -14.31 0.62 0.05 0.000 0.000 0.538 0.960 0726492 -.6446419 -.0252973 -.0608235 1200349 -.4893118 0484533 0639858 YIncome1 YIncome2 -.0027404 -.0064522 0007485 0013938 -3.66 -4.63 0.000 0.000 -.0042074 -.0091839 -.0012733 -.0037205 IPCost1 -.001424 0030491 -0.47 0.640 -.0074001 0045521 IPCost2 head1 gender -.0030728 0825114 -.0029731 005684 0426972 0187393 -0.54 1.93 -0.16 0.589 0.053 0.874 -.0142133 -.0011736 -.0397014 0080677 1661964 0337552 Monthhours1 Monthhours2 -.0002137 0000698 0001727 0001537 -1.24 0.45 0.216 0.650 -.0005522 -.0002314 0001249 000371 Age1 -.0047051 002695 -1.75 0.081 -.0099873 000577 Age2 Subsidy firstchild 0047876 0806007 2755225 0027867 0164842 0252304 1.72 4.89 10.92 0.086 0.000 0.000 -.0006743 0482923 2260718 0102494 1129091 3249732 ... street and children working in unhealthy situation and help 70% of those return back and living with their families 4.1 Overview of child labor problem in Vietnam Child laborers defined by Vietnam. .. ECONOMICS INSTITUTE OF SOCIAL STUDIES HO CHI MINH CITY THE HAGUE VIETNAM THE NETHERLANDS VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS PARENTAL ABSENCE AND CHILD LABOR IN VIETNAM. .. of child school cost on both child working status and child working hours If child? ??s school cost increase one billion per year, the probability of child labor decrease 2.4% and child working