Education and Social Mobility in Latin America United States

 Education and Social Mobility in Latin America United States

Latin America has witnessed an important expansion in educational coverage over the last two decades. On average, enrollment rates in primary (net) and secondary education (gross) increased from 85.9 percent and 49.6 percent in 1980 to around 94.0 percent and 89.7 percent in 2011, respectively. Many countries in the region are on track to meet or have already attained the Millennium Development Goal (MDG) in this dimension. Furthermore, this increase in educational coverage has been identified as an important driver of the observed reduction in earning inequality in the region (López-Calva and Lustig 2010), at a time when income inequality is rising within many developed and developing countries all over the world (see OECD 2011a and 2011b).

Despite this important progress, many challenges remain. In this note, I highlight two interrelated issues: 1) intergenerational mobility continues to be low; and 2) the quality of education is low, with significant differences across social classes in the opportunities of accessing high-quality education. Next, I will briefly discuss the empirical evidence relevant to each issue and outline some policies that could address these challenges.

Low intergenerational Educational Mobility

Social mobility is a multidimensional concept. Economists (e.g., Solon 1992) have traditionally focused on intergenerational income mobility, i.e., the link between a person’s permanent income level and that of his or her parents, based on the stream of income a person or household receives stripped of short-term fluctuations. Yet it is clear that other dimensions such as social status, often related to type of job or level of formal education, are also relevant. (For an interesting case study of Chile,see Torche 2005.) In what follows, I focus on educational mobility in terms of how parental education and family background affect educational attainment and achievement. This focus has an important practical advantage: relatively good-quality and comparable data for these variables are available for a significant number of countries in Latin America.

Several studies have addressed the issue of how family background affects educational outcomes of the next generation in Latin America. Behrman, Gaviria, and Székely (2001) present several alternative measures of educational and social mobility in Latin America and find that intergenerational mobility is much lower in Latin America compared to the United States. This finding is also confirmed by Daude (2011; 2012) and Gaviria (2007), considering the correlation between parental and child education outcomes using alternative data sources. Furthermore, in terms of correlation—that is, how much of the variation in the child’s education is explained by the variation in parental education—there are no significant changes over time (Daude 2011). In addition, a

study of a large number of developing countries also shows that Latin American countries rank poorly compared to other regions in terms of these measures of intergenerational mobility in education (Hertz et al. 2007). However, other studies find a recent improvement in mobility using alternative measures (Conconi et al. 2008). For example, they find that the importance of family background in explaining the “schooling gap”—defined as the difference between the years of education completed and the hypothetical years the child should have completed in the absence of repeating grades or dropping out—for children currently of compulsory enrollment age has declined between the mid-1990s and mid-2000s in several Latin American countries. In particular, they rely on a Fields decomposition (1996, 2003) proposed by Andersen (2001) that takes into account the household’s income per capita and the highest level of education between the mother and the father. However, these measures of mobility do not take into account differences in the quality of education, which are large in Latin America.

Low Quality of Education

Considering education achievement measures such as the OECD’s Program for International Student Assessment (PISA) scores, Latin American education systems tend to score poorly on two dimensions. First, the average achievement in terms of testable knowledge is low. For example, while on average within the OECD less than 20 percent of 14- to 15-year-old students do not reach a minimum level of  reading comprehension, in Latin America it is almost 50 percent. Second, the relationship between performance and aocioeconomic background is much stronger in Latin America than in the OECD (Figure 1). This shows that external circumstances—such as the household’s income level, parental education, gender, or geographical location—are key to student performance (Brunori, Ferreira, and Peragine 2013). In part, the lower levels of average achievement are due to a composition effect: as more poor students

with lower performance reach higher levels of schooling and stay for more time within the system (expanding coverage), in the short run performance falls. Some countries such as Brazil, Chile, and Mexico have shown significant improvements over the past decade in reducing the importance of family background in educational achievement outcomes. However, problems in terms of performance and inequality of opportunities are still prevalent. (See also Ferreira et al. 2012 regarding the evolution of social mobility and inequality of opportunities in Latin America.)Another interesting empirical fact from the PISA surveys is related to the social mix within schools. If one decomposes the variation in the index of socioeconomic status indicator and separates variation between and within schools, some OECD school systems (such as Finland) show that most of the variance comes from within schools, while for Latin America most of the variance is between schools. This means that schools in Latin America are socioeconomically very omogeneous. The poor go to the same schools asthe poor, while the rich gather only with the rich. Interestingly, the aggregate evidence shows that there is no trade-off between having socially mixed education systems and attaining high levels of educational achievement (OECD 2010a). Thus, education systems in the region continue to be an instrument for reproducing the current social order rather than a way to facilitate social mobility and opportunities for the poor