Although universities have expanded in size, it is unclear if the poor have benefited. If there are high returns to college education, then increasing access of the poor to college has important welfare implications. The objective of this paper is to document the rates of enrollment into college of the poor and to identify the hurdles to doing so. Relying on several sets of data, including a survey of college students from universities in three poor provinces in China, we have found that the college matriculation rate of the poor is substantially lower than students from non-poor families; the same is true for rural women and minorities. Clearly, there are barriers that are keeping the rural poor out. The paper also demonstrates that the real hurdles are not during the years of secondary schooling or at the time of admissions to college. The real impediments keeping the rural poor from pursuing a college education arise long before high school—as early as preschool and elementary school years—and are present throughout the entire schooling system.

Impact evaluation has become an increasingly integral part of development project design and execution in recent years. Many questions remain, however, about what methods yield the most compelling evaluations, and how best to implement them. The Rural Education Action Project (REAP) is among the most successful impact evaluation groups currently operating in China. The goal of this paper is to share five practical strategies that REAP has employed to maximize the effectiveness of our impact evaluations. These strategies include the use of randomization and other experimental and quasi experimental research designs; pursuit of local and international collaboration; strict attention to policy relevance; a modular, incremental research approach; and robust outreach. 

The overall goal of this paper is to try to understand in the context China’s middle school education system what is the tradeoff between community service and educational performance. In addition, we seek to understand some of the other effects of participation in community service activities. When students participate in community service, does their self-esteem rise and/or self-efficacy rise? Finally, we also want to know if students participate in community service activities will their sense of social responsibility rise.

Because the nature of this question is so broad we must necessarily reduce the scope of the paper. To do so, we study a community service program funded and run by a private US foundation in 592 junior high and high school classes in two counties in Shaanxi province. While we understand that there are many institutional and pedagogical issues that will affect the relationship among community service, grades and a sense of social responsibility (e.g., citation), we will focus our efforts on empirically evaluating the linkages. We also recognize that by focusing on the activities of an NGO in a single part of China that many of the findings will be affected by the context of the program. However, we also believe that we may be able to observe and measure more fundamental behavioral relationships. At the very least we seek to provide rigorous, statistically-based evidence from a case study which hopefully can be combined with other work to help us more fully understand community service, social responsibility and educational performance.

The education of the disadvantaged population has been a long-standing challenge to education systems in both developed and developing countries. Although computer-assisted learning (CAL) has been considered one alternative to improve learning outcomes in a cost-effective way, the empirical evidence of its impacts on improving learning outcomes is mixed. This paper uses a randomized field experiment to explore the effects of CAL on student academic and non-academic outcomes for students in migrant schools in Beijing. Our results show that a remedial CAL program held out of regular school hours improved the student standardized math scores by 0.15 standard deviations and most of the program effect took place within 2 months after the start of the program. Students with less-educated parents benefited more from the program. Moreover, CAL also significantly increased the students’ interest in learning