The Importance of School Systems: Evidence from International Differences in Student Achievement The Importance of School Systems: Evidence from International Differences in Student Achievement *
Students in some countries do far better on international achievement tests than students in other countries. Is this all due to differences in what students bring with them to school-socioeconomic background, cultural factors, and the like? Or do school systems make a difference? This essay argues that differences in features of countries' school systems, and in particular their institutional structures, account for a substantial part of the crosscountry variation in student achievement. It
... st documents the size and cross-test consistency of international differences in student achievement. Next, it uses the framework of an education production function to provide descriptive analysis of the extent to which different factors of the school system, as well as factors beyond the school system, account for crosscountry achievement differences. Finally, it covers research that goes beyond descriptive associations by addressing leading concerns of bias in crosscountry analysis. The available evidence suggests that differences in expenditures and class size play a limited role in explaining crosscountry achievement differences, but that differences in teacher quality and instruction time do matter. This suggests that what matters is not so much the amount of inputs that school systems are endowed with, but rather how they use them. Correspondingly, international differences in institutional structures of school systems such as external exams, school autonomy, private competition, and tracking have been found to be important sources of international differences in student achievement. Average achievement levels of students differ markedly across countries. On the most recent international achievement tests in math and science, the average 15-year-old student in Singapore, Hong Kong, Korea, Japan, and Taiwan is more than half a standard deviation ahead of the average student of the same age in the United States (Hanushek and Woessmann (2015b)). Following the rule of thumb that average student learning in a year is equal to about one-quarter to one-third of a standard deviation, these differences are roughly equivalent to what students learn during 1.5-2 years of schooling. Similarly, the average student in Finland and Estonia is 40 percent of a standard deviation ahead of the United States, and the average Canadian student is about one-third of a standard deviation ahead. On the other hand, the average student in Peru and Indonesia is more than 1.1 standard deviations behind the United States. When put on the same metric, achievement in Ghana, South Africa, and Honduras lags more than 1.5 standard deviations behind the United States. Overall, average achievement levels among 15-year-olds between the top-and bottom-performing countries easily differ by more than two standard deviations, or the equivalent of 6-8 years of learning. Why do students in different countries achieve at such vastly different levels? Apart from differences in socioeconomic and cultural backgrounds, do differences in the organization and governance of school systems play a role?