Migration, remittances and educational levels of household members left behind: evidence from rural Morocco.

• Author: Bouoiyour, Jamal
• Position: Report

1. Introduction

   Remittances have seen a considerable increase in recent years. For many developing countries, remittance inflows exceed the value of foreign direct investment and the amount of official development aid. These financial flows, needless to say, have already and will continue to have, profound effect on migrants' countries of origin. At the macro level, remittances seem to be more stable than other forms of external finance, and they also tend to be countercyclical, increasing during economic downturns (World Bank, 2012; Chami et al., 2005). Another important aspect of remittances is their impact on economic growth (Driffield and Jones, 2013; Benmamoun and Lehnert, 2013). At the micro level, it appears that migrants' remittances improve human welfare. For example, several studies have shown the beneficial effects of remittances on transient poverty reduction (Acosta et al., 2007; Esquivel and Huerta-Pineda, 2006). They can also affect the long term welfare of recipients by affecting human capital formation. There is empirical evidence for concluding that in some contexts international remittances have a positive impact on education and child labor reduction, thereby increasing the demand for child schooling (Mansuri, 2006; Acosta, 2006; Yang, 2008; Lopez Cordovas, 2006). This is particularly true for recipient households who are from disadvantaged groups.

   While international remittances can apparently help to increase educational attainment of children in receiving households by lifting liquidity constraints, migration of a family member may, however, have a deleterious impact on children' educational success in a given local area. For example, by implying that unskilled work can be an important source of wealth whereby additional income from remittances can be earned independently of schooling, migration may become a source of disruption in households. Further, the division of household labour will be done at the expense of the non-migrant members so reducing the time allotted to their education. Other potential negative influence of migration is that it may increase the likelihood that family members migrate in the future and reduces the desired schooling and, hence, decreases the educational attainment. Furthermore, migrants invest less in education of their own children if the return of education is low in their host countries (as noted by McKenzie and Rapoport (2011), this is the case of Mexican migrants living in United States). In sum, emigration may increase or decrease household investments in children's schooling, depending on whether the income effects from remittances offset the effects of household disruptions.

   Remittance flows to Morocco, which originate primarily from France, have increased remarkably over time. According to World Bank data, during the period 2001-2011 the amount of international remittances increased from 2.8 to 7.2 billions US$. Together with tourism, migrants' remittances represent the country's major source of foreign currency receipts. At over 6.5 billion US$ in 2012, they placed Morocco as the 15th largest recipient of remittances in developing countries. In the case of this recipient country, migrants' remittances are an important resource that can be a means to promote its development. However, beyond their quantitative importance, the possible impact of these remittances should be viewed in terms of their use. In this respect, the African Development Bank report (2007) found that in Morocco the priority is given to household consumption (essential goods and services), health and education. Unfortunately, empirical studies have not documented any significant work on the intra-family benefits of remittances to Morocco. For example, there is no evidence that international remittances can increase the educational attainment of children and of adults. This relation between remittance receipts and child education is of particular interest due to the constraints faced by families with respect to educational decisions. Our paper attempts to show how international remittances can play a significant role in shaping recipient family decisions on child education in Morocco (in particular on completion of non-compulsory grades). Do they have a chance to make a difference? Do they affect children at distinct educational stages differently? This paper is trying to effectively answer these questions. It examines the relation between the migrant inflow and educational attainment, measured by the level of education completed by children inside the typical age of graduation. This research is the first to empirically evaluate the effect of remittances on the schooling of children and to gauge whether or not being a girl is disadvantageous, by estimating probit models of higher secondary and higher education levels for male and female children separately. To accomplish this, we use data from an original Moroccan household survey, which provides information on household structure, education, income, housing and health and identifies all sources of income to Moroccan households. The main results is that when we consider the probability of completing secondary and higher education as measure of academic success, a strong statistically significant relationship emerges between remittances and educational attainment. More precisely, we find that remittances promote children's school attainment in rural Morocco, particularly among secondary school-age children. However, the impact of remittances on education attainment levels is heterogeneous with regard to child's gender and educational level.

   Beyond the conventional education impact of remittances, this study also tries to identify the effect of the migration itself on children's educational outcomes. It is thus the first to explore whether aggregation of data at the household level can explain the true impact of international migration -i.e., considering the sample of migrant households who have members abroad but do not receive remittances. The ability to identify migrant and not just recipient households allows us to directly analyze the relationship between international migration and educational attainment, and to compare our results with the ones that we obtain when we follow the more standard approach in the literature. In fact, the implicit assumption in most of the existing studies of remittances is that migration only affects children's educational success through remittances. However, the overall effect of migration on child education could be negative which was supported by results of this research. In fact, we find evidence of a negative impact of migration on educational attainment among adult children aged 21-24 whereas the positive effect of remittance receipt effectively disappears when we consider these school-age children. Further, the results show that the households' socioeconomic status may determine to what extent the household mitigates the negative effects of migration on their children's educational outcomes. In sum, the present study acknowledges a complex and ambiguous relationship between migration and children's outcomes.

   The remainder of the paper is as follows. Section 2 presents the theoretical framework. Section 3 describes the data and explains our methodological approach. Section 4 presents and discusses the empirical results. Section 5 concludes.

2. Results of past research

   Researchers have conducted several studies of associations between remittances and child schooling. These studies turn to regression-based analysis and consider a number of basic assumptions:

   Parents' level of schooling is one of the major factors behind the academic success and achievement of children. More-educated parents or heads are better informed about the benefits of education and the wage enjoyed by those who have higher levels of education (Cox and Ureta, 2003). These parents will encourage their children to pursue their studies, instead of working (Hanson and Woodruff, 2003; Gang et al, 2008). By the same token, several empirical studies on the development impact of remittances have shown the existence of a strong correlation between the educational level of parents and the academic success of their children. For example, McKenzi and Rapoport (2007) studied the impact of Mexican migration on the academic success of rural children aged between 12 and 18. They highlighted the important role of maternal education in the schooling of children living in migrant households. Indeed, the authors found that international migration reduced the educational level of girls aged between 16 and 18 if their mothers were educated, and reduced that of boys whatever the mother's level of education. However, their results show that migration has no effect on the education of girls aged between 12 and 15. In contrast, Hanson and Woodruff (2003) note an improvement in the education of girls whose parents have a low level of education. They justify this by the fact that Mexican migration, by reducing financial constraints on households, can be involved in financing the education of their children.

   Parents from disadvantaged backgrounds will compare the return on educational investment to alternative investments which could also increase the human capital of a child such as food, medical services and clothing. These families often find it difficult to send their children to school, if the costs of their education are too high. It appears therefore that there is a link between family income or wealth and the education of children (3). For Mansuri (2006), prosperous parents invest more in human capital because in return they can receive the lifetime earnings of children deriving from their educational attainment. Cox and Ureta (2003) present separately the effect of both income remitted by international migrants and income from other sources on the academic success of children in El Salvador. Their results show that in urban...