University of Western Australia and Linda Chor Wing Yung Department of Economics
Chinese University of Hong Kong
DISCUSSION PAPER 12.27
MAIDS OR MENTORS? THE EFFECTS OF LIVE-IN FOREIGN DOMESTIC WORKERS ON SCHOOL CHILDREN’S EDUCATIONAL ACHIEVEMENT IN HONG KONG
Sam Hak Kan Tanga
University of Western Australia
Linda Chor Wing Yungb
Department of Economics
Chinese University of Hong Kong
DISCUSSION PAPER 12.27
Abstract: This paper studies the effects of live-in foreign domestic workers (FDWs) on children’s educational achievement using samples from two population censuses and a survey dataset. The census data shows that the incidence of express schooling is significantly higher for children who are under the care of an FDW when their mothers are at work. In the survey data, children scored higher for English if they had a Filipino FDW. The age of FDWs had a positive and significant relationship with children’s average scores for Chinese, English and Mathematics. These findings suggest that FDWs provide an important childrearing service, which is often unrecognised and undervalued.
JEL classification: J61, I21
Keywords: Foreign Domestic Workers, Educational Achievement, Hong Kong.
a. Corresponding author. Business School, the University of Western Australia, 35 Stirling Highway, Crawley WA 6009, Australia. Tel: +618-6488-2931. Fax: +618-6488-1016. E-mail: email@example.com
Many families in Hong Kong employ a live-in foreign domestic worker (FDW), who not only performs household chores, but also looks after children while both parents are working.1 Since FDWs are usually paid a relatively low wage compared to the wage the mothers can earn in their formal employment, FDWs provide an affordable solution to childrearing for many Hong Kong women who would otherwise not have been able to join the workforce.2 It is thus no coincidence that the demand for FDWs has been steadily increasing since the early 1980s and the total number of FDWs in Hong Kong has reached around 300,000 at the end of 2011. However, despite the dependence of many Hong Kong families on their FDW, the Hong Kong government has a series of discriminatory policies in place against FDWs working in Hong Kong. For example, FDWs are denied the right of abode even though other foreign workers can apply for permanent Hong Kong residency.3 The current paper does not directly evaluate these discriminatory policies per se but aims to show that the contributions of FDWs to Hong Kong’s economy may not be confined to providing domestic services alone. It is contended in this paper that given their role as children’s main carer in the households when both parents are working, FDWs can have important influences on children’s educational achievement.
Recent research indicates that educational achievement, as opposed to years of schooling, in primary and secondary schools is the most important aspect of human capital formation, which directly contributes to long-term economic growth (Hanushek and Woessmann, 2008, 2012). Factors that affect educational achievement such as financial constraints, family background and school quality have also been studied extensively in the literature.4 This paper builds on these insightful studies to ask whether and how live-in FDWs affect children’s educational achievement. The findings of this paper will not only have important implications for the long-term growth prospects of Hong Kong, but also for the long-term growth prospects of other countries where a substantial number of FDWs are present. In addition, this paper attempts to draw greater public attention and recognition to live-in FDWs whose services to the local families are often regarded as low skilled and low value-added.
In theory, FDWs can have either positive or negative effects on children’s educational achievement. On the positive side, FDWs can directly affect the schooling achievement of children under their care through informal learning via personal contacts and tutoring school works. This potential positive effect is related to the FDWs’ own educational attainment and experience in childrearing. One study by Tse et al. (2009) examines the English reading comprehension of a selected group of grade four students in Hong Kong. They found a statistically significant difference between the reading performance of students who had an English-speaking FDW and their peers who did not. They concluded that English-speaking FDWs may have a positive influence on how well students acquired English language skills in the primary school. They also add that FDWs tend to have a stronger influence on the students’ English learning than do the actual parents of the children. Another potentially positive channel is that FDWs can help improve children’s schooling achievement by giving parents more time to supervise their own children’s school homework. Chan (2005) finds that parents do not delegate certain tasks to FDWs because these tasks symbolise their status as parents. Supervision of children’s school homework is one of those important tasks which the parents reserve for themselves and can spend more time on when house chores are performed by the FDWs.
FDWs can also have negative effects on children’s schooling achievement. Some studies investigated the social and psychological impacts of FDWs on children. FDWs may cause behavioural problems because they have responsibility for the children, but have no authority and thus are forced to appease the children in order to gain control (Yeoh et al., 1999). Cheuk and Wong (2005) found that children less than five years old and cared for by FDWs are associated with an increased risk and severity of specific language impairment. Tam (1999) discusses the role of FDWs in providing an affordable childcare service for Hong Kong families with young children. However, heavy reliance on FDWs to fill the childcare service gap is said to incur social and psychological costs for the FDWs themselves and the local families.5
Clearly, the questions of whether and how FDWs affect children’s educational achievement cannot be settled by appealing to theory alone. Using sample data from the Hong Kong 2001 Population Census and the 2006 Population Bi-census, this paper examines the incidence of express schooling among school children in Hong Kong. Express schooling suggests exceptional schooling achievement, indicating a child studying in a higher grade for their age than the rest of the class. The key results from the Census data show that the probabilities of express schooling are significantly higher for those school children who are under the care of an FDW than those who are not while their mothers are working. For children whose mothers are not working, however, the effects of FDWs on children’s educational achievement are weak but still present, indicating that FDWs’ effects are not as important when mothers are the main carers of the children rather than being in the workforce.
This paper exploits the exogenous differences in age, experience and education between the two largest ethnic FDW groups, Filipino and Indonesian, to show that the effects of FDWs on children’s educational achievement are not simply due to reverse causation. Specifically, if families employ an FDW because their children excel in school, they will naturally employ a Filipino FDW rather than an Indonesian FDW because Filipino FDWs are older, more experienced and better educated. This well-known behaviour among Hong Kong parents allows us to expect Filipino FDWs to have a greater effect on children’s educational achievement, some of which could have been the result of reverse causation. Indonesian FDWs, in contrast, are not expected to be subject to the same potential bias. As will be shown later, both Filipino and Indonesian FDWs produce positive estimated effects on children’s express schooling which are broadly similar, confirming our expectation that reverse causation is not a serious issue because we excluded high-income households from the samples and only included middle-income households in our regression analysis.
Because Census data does not provide direct measures of educational achievement of school children, we further investigated the effects of FDWs on children’s educational achievement by conducting a survey of school children in selected schools in Hong Kong. The results based on detailed data returned from 151 families with school children show strong evidence of higher school marks in English for children with a live-in FDW, especially if Filipino. The survey data also reveals that the age of FDWs is an important factor in children’s schooling achievement: children scored a higher average mark for the three core academic subjects (Chinese, English and Mathematics) when an older live-in FDW is employed in the household. Older FDWs, many of whom are mothers themselves, have the experience and training to be children’s quality carers. The survey results thus corroborate the findings of the Census data that live-in FDWs contribute positively to children’s academic achievement.
Previous studies on the effects of FDWs on school children’s educational achievement are rare. Economists are drawn into studying FDWs (or domestic workers in general) mainly by the question of how wage is determined for FDWs (Suen, 2000), or of the effects of legislating a minimum wage law on the wage and employment of domestic workers (Dinkelman and Ranchhod, 2012), or of the effect of tax reductions on the employment of domestic workers (Flipo et al., 2007). To the best of our knowledge, the paper by Cheo and Quah (2006) is the closest to the present study which investigates the effects of maternal employment, FDWs, and private tutoring, among other factors, on school children’s academic results. Their empirical findings, based on a survey of 429 grade eight students from three premier schools in Singapore, find that students with working mothers, spending less time with a private tutor and less traveling time to school, produce better academic results. However, FDWs are found to have little effect on school children’s academic achievement. One important difference between theirs’ and the present study is that we uses a much larger random sample set which consists of more than 20,000 school children from middle-income families. Our random sample set generated from the Census data contains detailed family characteristics of school children and thus provides more comprehensive controls for the covariates of academic achievement. Also, excluding school children from high-income families in our study ensures that school quality is relatively homogeneous. Moreover, our survey sample is drawn from typical government schools which are not premier schools in Hong Kong. Finally, the effects of FDWs on school children’s academic performance should decline with children’s age, and by the time the children are admitted to grade eight when they are already 13 or 14 years old, the effects of FDWs may not be easily detectable with a small sample set.
This paper is organised into five sections. Section 2 discusses sample statistics obtained from the Hong Kong 2001 Population Census and the 2006 Population Bi-census, and the measures of educational achievement using “express” and “late” schoolings. Section 3 reports and interprets the regression results based on samples taken from the Census data. Section 4 discusses the design and results of the school survey. Section 5 summarises and concludes the paper by drawing attention to the policy implications of our findings.