17 Figure: Grade 6 language achievement by province in South Africa, where the home language is the same as the language of learning and teaching (LOLT) and the home language is different from the LOLT. Home language same as LOLT Home language different to LOLT 100% 90% 80% National average 69 70% 60% 50% 40% 30% National average 32 20% 10% 0 te pe n r ste Ea Ca e e Fr Sta g ten au aZ ulu al at N G po ga lan po Lim a um p es W e pe inc ov r tP n r the Ca pe n r ste Ca e or W N N Source: Grade 6 Systemic Evaluation National Report, Pretoria: Department of Education, 2005 Kw M or th Note: The blue line indicates the much better academic performance of children taught in their own language in the first years of education, as opposed to those taught in a language that is not their own (black line). The issue of the cost-effectiveness of education in a minority language is considered far less often but just as important. Education in a minority language is more cost-effective, even if it has some slightly higher initial costs in terms of teaching materials or training, because it produces more secondary school graduates than a cheaper official language-only education. The cost of public education per successful secondary school graduate has been shown in the few studies that examined this issue directly to be lower than in other public schools, because of the higher success rate in minority schools. Schools that also use minority languages to communicate with parents have been shown to increase their involvement and improve their understanding of their children’s education.27 For example, in Guatemala, the long-term cost saving as a result of first language-based education for all children who do not speak the official language was estimated to equal the cost of primary education for 100,000 students, or a potential saving of over 31 million quetzals (US$ 5 million).28 A similar study in Mali found that French-only programs cost about 8% less per year than mother-tongue schooling, but the total cost of educating a student through the six-year primary cycle is about 27% more, largely because of the difference in repetition and dropout rates.29 27 28 29 Child Trends Data Dank, School communication in parents’ native language, October 2015, <http://www.childtrends. org/?indicators=school-communication-in-parents-native-language>. Patrinos, H. and Velez, E. ‘Costs and benefits of bilingual education in Guatemala: a partial analysis’, International Journal of Educational Development, vol. 29, no. 6 (2009), pp. 594–598, at p. 597. World Bank (note 2).