Alcance

Progresa-Oportunidades has led to higher enrollment and continuation rates in both rural and urban areas which has increased the difference in education levels between recipients and their children (Schultz, 2004; Parker, 2003; González de la Rocha, 2008; Parker and Berhman, 2008; Parker, Rubalcava and Teruel, 2009). The program has also led to a reduction in the probability of child labour (Skoufias and Parker, 2001; Parker, Berhman and Todd, 2005) and delayed the reproductive age by extending the years of schooling mainly among girls (González de la Rocha, 2008).

Participation in the program has also been linked to narrowing of gender and ethnic gaps in years of education (Agudo, 2008). Impacts are particularly notable at the secondary and high school level as enrollment rates in elementary school were very high before the implementation of Progresa-Oportunidades (Schultz, 2004; Parker, 2005).

Program impacts on school enrollment and continuation rates have been established, but; the few studies on achievement tests show limited learning results (Berhman, Parker and Todd, 2007; Parker, 2008; Mancera, Serna and Priede, 2008;

2012). This is attributed to the poor quality of education provided to recipients. Even the positive impacts on education are relatively modest as argued by Yaschine (2012) and Escobar (2012). Yaschine questions the efficacy of the program in the long-term if only an additional year of education has been achieved after more than ten years of program operation, and learning levels are low; and questions if the magnitude of the impacts is sufficient to break the intergenerational transmission of poverty. These effects are modest relative to the potential impact of the program if the services provided were of good quality (Yaschine, 2012: 63). Numerous studies have emphasized the need for more attention to the quality of education provided to program recipients (Skoufias, 2005; Agudo, 2008; Behrman, Parker and Todd, 2009; Yaschine, 2012; Mancera, Priede and Serna, 2012). Although program recipients are completing more years of schooling, Mexico’s highly segmented educational system means that the existing gap in quality of

services between urban and rural and indigenous and non-indigenous remains (Agudo, 2008: 126). In Mexico, the low quality of education offered to the poorest households operates as a mechanism that reproduces inequality (Yaschine, 2012: 262; Mancera et al, 2012).

The program itself may have perverse effects on the quality of education, In his qualitative research in rural areas, Agudo (2008) found evidence of an extended practice where teachers do not report absences and do not fail students in order to avoid conflict with parents, which may explain the reduced levels of repetition in elementary and secondary levels that have been attributed to the program. The author argues thatthis practice can result “in diluting the impact of Oportunidades and converting the Program into solely a mechanism to transfer income” (2008: 130). Behrman et al. (2005) have suggested that Progresa-Oportunidades may have actually lowered school quality, given the increased number of students derived from the program support, and the resulting

“congestion effect”. González de la Rocha (2008: 26) among others, acknowledge that services have expanded in response to demand generated by the program, but this expansion is led by services of the poorest quality.

Early evaluations showed marginal impacts of the program on school enrollment in primary school, as enrollment rates at this level were already high before the program was implemented, and suggested shifting these transfers to secondary and high school (Skoufias, 2000). According to Levy, these suggestions led program officials to add in 2001 three additional years of grants to students attending high school (2006: 60); but transfers to students in primary school were kept the same. More recently, a number of scholars have suggested eliminating transfers at this level in order to improve the program’s efficiency by reducing the costs associated with targeting and monitoring of compliance (Schultz, 2004; Skoufias, 2005; Escobar and González de la Rocha, 2005;

De Janvry and Sadoulet, 2006; Attanasio, Meghir and Santiago, 2007; Agudo, 2008;

Azevedo and Robles, 2010; SEDESOL, 2012). They suggest redirecting the resources from primary to secondary school, where most students dropout from school. Despite these suggestions, education transfers to elementary school have continued, and in fact were expanded in 2010 to cover students in the first three years of primary school in rural areas.

It is also important to disentangle how much of the increased enrollment rates in higher levels of education can be attributed to the scholarships offered by Progresa-Oportunidades rather than to other factors. Levy argued that “it is not correct to conclude, for example, that the program’s impact on school enrollment derives from the education transfers” recognizing that school enrollment might also been affected by other factors (2006: 43). González de la Rocha (2008) argues that the magnitude of the impacts of Progresa-Oportunidades on increasing years of schooling depends on contextual, household and individual factors, including school proximity, quality of education, parents' incomes, and family members support (remittances), among others (2008: 30). Proximity to school plays a significant role in determining the impacts of Progresa-Oportunidades, as evidence suggests that children in recipient households that are further away from secondary schools are less likely to enroll (Schultz, 2004; De Janvry and Sadoulet, 2006; Winters and Chiodi, 2011). González de la Rocha found that “one of the factors explaining the longer school careers of the Oportunidades beneficiaries is the short distance between schools and homes. While primary school coverage is widespread, it starts to decrease as the schooling level increases, so there is a shortage of junior high schools and above all, senior high schools” (2008: 37).

Agudo (2008) came to a similar conclusion, stating that part of the explanation for increased enrollment rates is due to the expansion of educational services, and that “the prolongation of educational trajectories and the increase in average schooling to which Oportunidades contributes take place especially in situations where there are sufficient educational offerings (for example, nearby secondary schools or in the locality itself) (2008: 131). According to Agudo ““school dropout is due, in part, to problems with educational coverage more than coverage by the Oportunidades Program itself, as the students with scholarships who drop out of school are usually those who must travel long distances from small localities where there are no schools to reach the school they attend. When such traveling becomes problematic (for example, due to transportation costs or the need for children to help with farm work), Program scholarships are not a sufficient incentive to stay in school” (2008: 127). Agudo recommends adapting the program to the diverse conditions of recipient households, by establishing a differentiated system of grants for those children or youth who have to commute outside their communities to study (2008: 129). Recognizing that implementing a differentiated

system could be complex, “it would introduce greater equity in the recipient population”

(2008: 128). He also recommends expanding boarding schools at all levels, particularly in those regions where school coverage is limited (2008: 129).

A differentiated system of scholarships paying more to those students living far away from schools has been recommended in several studies (González-Konig et al, 2007; Sariego, 2008; Rodríguez-Castelán, 2010; Ulrichs and Roelen, 2012). González-Konig et al. (2007) suggest that when the marginal utility of children is higher than the transfers, the general amount offered by the program is insufficient to offset the opportunity costs and as a consequence, the small amount conditioned to minimal attendance could exclude the poorest from participating in the program. Ulrichs and Roelen (2012) criticize the program for providing the same amount of transfers to all households, without taking into account “distance from schools or health centers, or depth of poverty” (2012: 14), and recommend integrating “a differential amount of cash for beneficiaries living in particularly remote areas, to cover the higher participation cost to comply with conditionality requirement” (2012: 15). They maintain that the program does not respond to “the particular needs and structural causes of vulnerability in the programme design, to compensate for higher opportunity costs in complying with the conditions, incurred due to remoteness and limited access to services” (2012: 6). The authors criticize the program’s “one-size-fits-all” application as “it fails to appropriately and adequately address the structural causes that perpetuate poverty, particularly among indigenous people” (2012: 17).