International Remittances and Human Capital Formation VANIA B. SALAS * The Pennsylvania State University, USA Summary. — This article investigates the effect of international migration on children left behind in Peru. The theoretical model is based on human capital theory and educational investment decision linked to remittances. The model analyzes the role of international remit- tances on the investment decision between sending children to a public school or to a private school. Using data for the period 2007–10, this study addressed the problem of a left-censored endogenous variable for panel data by using a two-step estimation, and found that international remittances have a positive effect on the likelihood to send children to private schools controlling for absenteeism of par- ents. Ó 2014 Elsevier Ltd. All rights reserved. Key words — international remittances, international migration, human capital 1. INTRODUCTION One of the main problems facing policy-makers in Peru is improving the quality of public education in order to enhance human capital in the country, and thus leave the bottom rank in the Programme for International Student Assessment (PISA) evaluation compiled by the Organisation for Economic Co- operation and Development (OECD). Although educational coverage has increased to almost the same level as in devel- oped countries (UNESCO-UIS/OECD, 2005), the quality of Peruvian education remains low. PISA (2003) results show that Peru has an average score of 327 points which is the low- est score in the Latin American region (Argentina, Chile, Mex- ico, Brazil, and Peru) and much lower than the average of the OECD countries (500 points) (UMC, 2004). As a consequence of PISA’s results the Peruvian education has been declared to be in a state of emergency by the national government. The educational system in Peru reflects the high inequality experienced within Peruvian society; private schools provide better education compared to that provided by public schools. According to the results on reading literacy skills re- ported in PISA (2003), the difference in student performance within Peru for the year 2000 between the highest and the lowest quintile is 314 points (UMC, 2004). 1 This gap is ex- plained by the differences among schools, and around 58% of student’s performance in PISA is explained by school fac- tors (UMC, 2004). This fact highlights that a better quality of education is provided by private schools compared to pub- lic schools. Attending school, especially private schools, has become a pathway to move up and improve socio-economic status (SES) which may have a larger impact on children from low- and middle-income households. Evidence from the US suggests that children from private schools have a higher prob- ability of accessing a better college education (Griffith & Roth- stein, 2009), which in turn increases their job opportunities. Children from low-income households are doomed to attend public schools which in turn make them less competitive than children educated in private schools. The main constraint to access not only private but also public education in Peru is its cost and the opportunity cost of children’s work. Public education in Peru is free and pro- vided by the government, however, parents need to pay for additional expenses such as uniforms, transportation, and school supplies that may pose a high economic burden on especially low income families. Saavedra and Suarez (2002) find that Peruvian families contribute to 32% and 33% of the total public spending for primary and secondary educa- tion, respectively. Hence, sending children to school becomes prohibitive for families in the lower group of Peruvian’s in- come distribution. International remittances might loosen this income constraint allowing families left behind by, firstly sending children to school, and secondly affording a private education for their children. However, there is still not a clear effect of remittances on schooling (Borraz, 2005; Hu, 2012) and to my knowledge all the studies on remittances’ effects have evaluated only quantity without including quality of schooling. Only one paper (Calero, Bedi, & Sparrow, 2008) approaches the quality decision for Ecuador but it uses one-year information and did not include absenteeism of parents. International remittances may be used to acquire more years of education and a higher quality of schooling but according to Rapoport and Docquier (2006) there are several motives for sending back money to the origin country. The empirical evidence is not clear; some studies find that remit- tances are used for consumption whereas other studies find that remittances are used to acquire investment goods such as education and health (Yang, 2006). Based on the perma- nent income hypothesis, if remittances are considered as a temporary income they will be invested rather than spent (McKenzie & Sasin, 2007). Hence, the allocation of remit- tances income is not perfectly fungible with other income sources of the household, and the expected effect is not straightforward. This paper focusses on the effect of remittances on human capital investment of children left behind in the home country. The potential positive effect of remittances on financial con- straints changes the opportunity cost of acquiring more schooling; thus, families may find it optimal to send children to school instead of sending them to the labor market. Researchers have explored remittances’ effects on different indicators of human capital. Cox and Ureta (2003) employed dropping out of school as an indicator to explain the large and positive effect of remittances on human capital in comparison to other income sources for El Salvador. Amuedo-Dorantes Final revision accepted: January 31, 2014 World Development Vol. 59, pp. 224–237, 2014 Ó 2014 Elsevier Ltd. All rights reserved. 0305-750X/$ - see front matter www.elsevier.com/locate/worlddev http://dx.doi.org/10.1016/j.worlddev.2014.01.035 224