Feb. 14-15: Online Days
Feb. 18-22: On-site Days in Washington DC

How To Reduce Educational Inequality Between Urban and Rural Regions in China: Evidence from Field Experiments

Tong Ru

University of Pittsburgh

Education: The Footstone of Economic Growth

Many scholars have discussed how education contributes to economic growth around the world. Theoretically, there are three main explanations of how education may improve economic growth: 1) By increasing the human capital of the labor force, or their productivity; 2) By advancing the innovative capacity and the development of new techniques for the economy, which leads to the increasing productivity; 3) By making more people understand knowledge and updating information, followed by more progress of technology (Hanushek & Woessmann, 2010). Referring to the first explanation -education improves the human capital of the labor force-, what is Human Capital? According to Schultz (1961), human capital has quantitative and qualitative dimensions, such as the number of people, the hours people worked, years of schooling, and so on. Generally, human capital can be categorized into the following five scopes: 1) health-related outcomes; 2) on-the-job training; 3) formal education; 4) study after adulthood; 5) migration for job seeking.

Understanding what is Human Capital, how does education matter for human capital? The Human Capital Theory (HCT) states that acquiring education can improve people’s productivity and earnings, which varies by education level. According to the World Bank Research Paper (Psacharopoulos & Patrinos, 2018), globally on average, the private returns for primary education are the highest (25.4 percent), and secondary education has the least private returns (15.1 percent). But for low-income countries, higher education has the highest private return rates (26.8 percent); and for high-income countries, higher education has the lowest private return rates (12.8 percent), highest for primary education (28.4 percent). Though varying by education level, we can see that both developed countries and developing countries benefit from improving educational attainment (Shafiq et al., 2019).


China’s Urban-Rural Educational Inequality

Throughout the recent 40 years of economic growth in China, the improving human capital and labor productivity have played essential roles and will continue to serve as the foundation of future growth for urban and rural China (Li et al., 2017; Yue et al., 2018). Understanding the positive effects of improving educational achievements on human capital, the Chinese government, researchers, and program practitioners have launched many policies and programs to advance educational outcomes. Looking at the big picture, China’s public expense for education increased to 4.11% of the GDP in 2018 from 2.1% in 1978 (Ministry of Education of the People’s Republic of China, 2019). From the individual’s perspective, according to China’s sixth population census in 2010, more than 96% of the males and 90% of the females have accepted certain education, which brings considerable returns to both individuals and society (Xie et al., 2019).

Despite China’s tremendous progress in human capital development in the past 40 years, the urban-rural human capital gap still exists, which should not be neglected since nearly two-thirds of Chinese children live in rural areas (MOE, 2020). According to the Sixth National Population Census, from 1990 to 2010, the average educational attainment of China’s rural people (Population Aged 6 and Over) increased from 5.75 years of schooling to 7.58 years, while compared to urban people (7.52 years and 9.98 years respectively), rural people just caught up the urban people after 20 years later (National Bureau of Statistics of China, 2001; NBSC, 2011). From China’s latest Seventh National Population Census, the overall national average educational attainment was 9.82 years of schooling (Population Aged 3 and Over), 11.19 years for urban people and 8.45 years for rural people respectively (NBSC, 2021). Admittedly, rural education improves a lot yearly, while still being left behind compared to urban areas.

When considering educational attainment, not only years of schooling but also other micro-level indicators can tell something. Researchers found that millions of children in rural China are suffering the unequal distribution of educational resources, lower academic outcomes, poorer health and mental health conditions (for example 25% ~ 34% of rural children had anemia in northwestern China), and less attention paid to their social-emotional and non-cognitive development, compared with their urban peers (Lai et al., 2014; Zhou et al., 2015; Yue et al., 2018; Loyalka et al., 2019; Bai et al., 2020). The unequal educational development between urban and rural China not only is harmful to the poor’s productivity in the short term but also heightens the urban-rural income inequality in the long run, which should be addressed to achieve China’s grand goal of social justice (Park et al., 2010; Li et al., 2015).


Global Experiences in Improving Educational Achievements

Looking at the Education Production Function (EPF), input and output are the two main elements (Hanushek, 1989; Hanushek, 2020). To clarify the output in EPF, the most commonly used one is student’s educational achievement. Regarding the input in EPF, various indicators can be generally distributed to the school-level inputs (labors, facilities, etc.), teacher-level (working hours, teaching qualifications, etc.), family-level (books, parental care, etc.), student-level (studying hours, studying interests, etc.), and community-level (children center, libraries, etc.). Empirically, from the input perspective, many researchers around the world have examined which factors may influence students’ educational achievements and discussed how to improve the input on education.

School-level & Teacher-level

According to the review by Pelayo and Brewer (2010), the most important teachers characteristics is teacher quality. Generally, there are two kinds of variables that can be used to indicate teacher quality from the empirical evidence: 1) Easy to observe: years of experience (while not always the more the better), degree level & major (only affect student’s educational outcomes in certain subject teaching), pedagogical training, certification (conflicting evidence); 2) Hard to measure: philosophy of teaching, interpersonal or social skills, etc. For teachers’ demographic characteristics, several studies also stated that teachers’ ethnicity and gender also matter to students’ educational outcomes (Dee, 2005; Paredes, 2014).

For improving teachers’ performance, according to Podgursky et al (2010), there are mainly two different levels to improve teachers’ teaching performance: 1) Teacher training or teaching license, which represents that the teacher meets the minimum standard of proficiency; 2) More complicated teacher certification or teacher professional development program, for example, the Teach For America Program improved teachers’ teaching in math compared to new teachers who didn’t attend this program. Among the so many policies and projects targeting teacher performance, including teacher training (both pre-service and in-service), teacher incentive plans, and teaching curriculum revolution, teacher incentive plans attract much attention both in the United States and China.


It has been proved that family-level factors do affect students’ educational outcomes in different ways (Rothstein, 2010). Generally speaking, the following characteristics at the family-level matter for children’s production of education: 1) Educational attainments of parents, parental migration, number of siblings, family assets, etc; 2) Occupation of parents, income level of parents, the social network of family members, etc; 3) Demographics, including gender, race, age, etc. For example, the higher educational attainment of parents is found to positively correlate to students’ reading ability and critical thinking; higher socioeconomic families put more effort into students’ after-school art-related activities; also, children from poorer families usually suffer worse health, which has negative effects on the human capital development of children since they were very young.


Regarding the role of social context on education, the effects of social context on children’s educational outcomes are different in different countries (Carnoy & Marshall, 2005). Taking Cuba as an example, researchers tried to figure out why Cuban students performed better in math and language than in other countries, finding that the coefficients of indicators of social capital -including whether attending preschool, whether students work after school, violence in school, and average socioeconomic background in school– for Cuban samples explained more students’ outcomes than others.

What Have Been Done to Narrow the Urban-Rural Educational Inequality in China?

Understanding the Education Production Function and aiming to narrow China’s urban-rural educational gap, scholars have explored various empirical interventions to improve rural children’s educational outcomes. Inspired by the Abdul Latif Jameel Poverty Action Lab (J-PAL) and the Innovations for Poverty Action (IPA), which have examined the effects of RCTs on improving educational outcomes and reducing poverty in developing countries in South Asia, Africa, and Latin America since 2003, researchers in China started to explore the effect of conducting an RCT in rural education from 2005 and continued to apply this method into more than 50 field experiments in rural China till today.

Briefly speaking, an RCT in educational research is implemented mainly by the following three steps: 1) Baseline survey, to collect the necessary information of all samples, including students’ standardized test scores, demographics, family background characteristics, teachers’ demographics, teaching time and workload, and school characteristics; 2) Randomized allocation and intervention, randomly selecting a part of samples to become the “treatment group” and launched certain intervention activities to the “treatment group” while doing nothing to the remaining samples who are randomly assigned to be the “comparison group”; 3) Endline survey, after a certain period of intervention implementation -usually an academic semester, researchers collect all information same as the baseline survey once again and examine whether there are any differences of all outcome variables between the “treatment group” and the “comparison group”.

According to the review of the empirical studies which use the Randomized Controlled Trials (RCTs) method -the golden standard to identify causal effects- in rural China’s education from 2007 to 2019, researchers found that most of the field trials focus on the following perspectives to improve rural children’s educational outcomes: Teacher and Teaching, Computer-Assisted Learning, Nutrition and Health, and Early Childhood Development (Shi et al., 2020).

School-level and Teacher-level

Teacher and Teaching

At the teacher-level, after several trials using the randomized controlled experiment method, scholars mainly found the following two effective approaches to improving human capital outcomes of children in rural China: life teacher training in boarding schools (Yue et al., 2014) and a “pay-for-percentile” teacher incentive plan (Loyalka et al., 2019; Chang et al., 2020). For the effects of the life teacher training, though the research team didn’t observe students’ improving academic outcomes in the experiments, they did find that training on life teachers (who are responsible for students’ accommodation administration, out-of-class activities, and health management) decreased students’ bad behaviors at school. Specifically, they noticed that after being treated, students in the treatment group (whose life teachers were trained by the research team) decreased the probability of being late or leaving early for class by 48% and less fighting with classmates by 78%.

Regarding students’ academic outcomes, scholars examined different ways of ranking teacher’s performance according to students’ academic outcomes: “levels incentive” (teacher performance was measured by the average year-end exam scores of all students in the class, and teachers were ranked based on average-class achievement), “gains incentive” (teachers were ranked by the average gains of all students between year-start exam scores and year-end exam scores), and “pay-for-percentile” (in which teacher performance was measured by the average percentile rank of all students at the endline exam in the class). After one year of intervention, scholars found that the “pay-for-percentile group” significantly improved students’ math test scores by 0.148 standard deviations while the other two didn’t improve students’ scores significantly. Furthermore, they found that the “pay-for-percentile” teacher incentive plan promoted teachers to focus on all students in their classes, especially low-achieving students (Chang et al., 2020).

Computer-Assisted Learning

Computer Assisted Learning (CAL) has been proven to improve students’ academic outcomes by tons of evidence from different countries, which is also considered to be a feasible option for children in rural China to substitute for their lack of after-class commercial private tutors. Conducting 7 randomized controlled trials from 2011 to 2016 in both migrant communities in cities and rural villages, scholars examined the effectiveness of CAL on improving rural students’ academic outcomes and non-academic outcomes and also tried to explore its mechanism (Lai et al., 2013; Mo et al., 2013; Lai et al., 2015; Bai et al., 2016). In China’s migrant schools in big cities (special schools designed for children who live in big cities with parents’ migration, who have no access to public schools due to the registration restriction), researchers found that CAL intervention (twice a week, each for 40 minutes) improved students’ computer skills by 0.33 standard deviations and test score in Math between 0.15 ~ 0.17 standard deviations. In rural schools, CAL increased students’ test scores in Math by 0.12 ~ 0.22 standard deviations, especially for minority students whose test scores in Chinese also improved by 0.14 ~ 0.20 standard deviations. Additionally, by attending the computer class twice a week, students also gained more self-efficacy and more educational expectations.

Nutrition and Health

Regarding the other critical ingredient of human capital: health outcomes, looking at the so many trials of advancing rural children’s health outcomes, we found that these interventions not only improved students’ health conditions but also their academic outcomes. To address the universal iron-deficiency anemia in rural Northwestern China (about 25% ~ 30%), researchers mainly explored the following three interventions targeting students’ malnutrition problem: providing free multivitamin supplements (with iron) at school (Luo et al., 2012; Wong et al., 2014), information intervention targeting school principals (Miller et al., 2012), and incentive intervention toward school managers (Luo et al., 2019a). Though providing free multivitamin supplements at school to students did improve their hemoglobin level by 2.14 ~ 3.69 g/L (0.2 ~ 0.33 standard deviations) and test scores of Math by 0.1 ~ 0.2 standard deviations, scholars found that information intervention on school principals (nutritional knowledge training) didn’t affect students’ academic outcomes. Furthermore, researchers showed that for schools that assessed school quality using students’ academic outcomes, providing incentives or subsidies to school managers by examining how many students in their schools had anemia improved students’ hemoglobin levels by 2.14 ~ 8.6 g/L, decreased the ratio of anemia by 13.8% ~ 14.5%.

Family-level and Community-level

Early Childhood Development

Except at the school level and teacher level, scholars also explored how to improve rural children’s human capital at the family level and community level, mainly focusing on early childhood. At the family level, Zhou et al (2016) explored that providing free Micronutrient supplements to infants (6 to 36 months old) not only improved their hemoglobin levels by 1.77 g/L but also advanced their cognitive development index by 2.23 points of the score (using Bayley Scale of Infant Development, BSID). Using the specially designed early childhood development curriculum, the research team trained local village officials as parenting trainers to visit these infants’ families and teach family members how to use the curriculum to interact with their children, finding that the weekly home-visit parenting instruction improved infants’ cognitive development score by 0.27 standard deviations compared to the control group (who didn’t receive the treatment) (Yue et al., 2019).

To decrease the cost of the intervention, researchers also explored whether bi-weekly home-visit parenting instruction has similar effects and found positive effects of 0.24 standard deviations (Luo et al., 2019b). At the community level, to cover more infants in rural villages and decrease the cost of parenting instruction treatment, scholars have tried to build village-level parenting centers that not only provide one-on-one parenting instruction but also group reading activities and communication platforms for caregivers. While the effects of parenting centers are under evaluation, they preliminarily found that the community-level parenting intervention may not only benefit infants’ cognitive and language development but also their social-emotional development (Shi et al., 2020).

Next Steps to Close China’s Urban-Rural Educational Gap

Looking at the above studies which examined the effects of various interventions targeting the human capital development of children in rural China, one of my most critical learning is that empirical research never ends. Though scholars have done many trials which benefited millions of children in rural China, more explorative studies to narrow China’s urban-rural educational gap are still needed to make the systematic change from the bottom (local rural schools and rural families) to the top (the Ministry of Education, policymakers). Taking the advantage of RCTs to identify effective ways to improve children’s educational and health outcomes and promoting evidence-based policy changes will definitely benefit more rural children. While transferring experiment results to policies that will be implemented nationally is challenging, we can still make some efforts to make ourselves closer to our final grand goal.

Considering China’s policy-making process, the more detailed procedure for conducting interventions, the easier it is to operate in local schools or rural families, and the bigger effects we would expect from the well-designed treatments targeting rural children’s education and health. Under the centralized educational system in China, educational policies are usually announced by the Ministry of Education and delivered to local educational departments level by level, from province to city and county. After education policies are delivered to the most basic government departments -the County Education Bureau-, principals of schools in the townships and villages are informed of the latest policies and implement policies at their schools. During the process of transferring educational policies from the central government to local educational departments, local educational departments have a certain freedom to supplement the detailed regulations according to their regional characteristics, which provides the confidence for us to expect higher effectiveness if we make sure to cooperate with local government at the beginning of a project.

Another key to promoting policy change using empirical evidence from field experiments in China’s rural education is the cost-effectiveness analysis. Thinking back to the above trials which researchers explored to improve rural children’s human capital outcomes, there are many ways that can be tried: providing free nutritional supplements to students, training and incentive for teachers, parenting instructions on family members, and so on. Among the so many verified effective interventions, which one will be chosen by policymakers? Accurate cost-effectiveness analysis should be one of the answers. By comparing the costs and benefits of different treatments or interventions, we could have an estimation of which intervention can acquire the greatest benefits with the lowest costs, which would help policymakers a lot to choose what they really would like to pay for.


Bai, Y., Mo, D., Zhang, L., Boswell, M., & Rozelle, S. (2016). The impact of integrating ICT with teaching: Evidence from a randomized controlled trial in rural schools in China. Computers & Education, 96, 1-14.

Bai, Y., Neubauer, M., Ru, T., Shi, Y., Kenny, K., & Rozelle, S. (2020). Impact of second-parent migration on student academic performance in Northwest China and its implications. The Journal of Development Studies, 56, 1523-1540.

Carnoy, M., & Marshall, J. H. (2005). Comparing Cuban academic performance with the rest of Latin America. Comparative Education Review, 49(2), 230-61.

Chang, F., Wang, H., Qu, Y., Zheng, Q., Loyalka, P., Sylvia, S., Shi, Y., Dill, S., Rozelle, S. (2020).The impact of pay-for-percentile incentive on low-achieving students in rural China. Economics of Education Review, 75, 101954.

Dee, T. S. (2005). A teacher like me: Does race, ethnicity, or gender matter?. American Economic Review, 95(2), 158-165.

Hanushek, E. A. (1989). The impact of differential expenditures on school performance. Educational Researcher, 18(4), 45-62.

Hanushek, E. A., & Woessmann, L. (2010). Education and economic growth. Economics of Education, 60, 67.

Hanushek, E. A. (2020). Education production functions. In The economics of education (pp. 161-170). Academic Press.

Lai, F., Zhang, L., Hu, X., Qu, Q., Shi, Y., Qiao, Y., Boswell, M., & Rozelle, S. (2013). Computer assisted learning as extracurricular tutor? Evidence from a randomised experiment in rural boarding schools in Shaanxi. Journal of Development Effectiveness, 5(2), 208-231.

Lai, F., Liu, C., Luo, R., Zhang, L., Ma, X., Bai, Y., Sharbono, B., & Rozelle, S. (2014). The education of China’s migrant children: The missing link in China’s education system. International Journal of Educational Development, 37, 68-77.

Lai, F., Luo, R., Zhang, L., Huang, X., & Rozelle, S. (2015). Does computer-assisted learning improve learning outcomes? Evidence from a randomized experiment in migrant schools in Beijing. Economics of Education Review, 47, 34-48.

Li, H., Loyalka, P., Rozelle, S., Wu, B., & Xie, J.  (2015). Unequal access to college in China: How far have poor, rural students been left behind?. The China Quarterly, 221, 185-207.

Li, H., Loyalka, P., Rozelle, S., & Wu, B. (2017). Human capital and China’s future growth. Journal of Economic Perspectives, 31(1), 25-48.

Loyalka, P., Sylvia, S., Liu, C., Chu, J., & Shi, Y. (2019). Pay by design: Teacher performance pay design and the distribution of student achievement. Journal of Labor Economics, 37(3), 621-662.

Luo, R., Shi, Y., Zhang, L., Liu, C., Rozelle, S., Sharbono, B., Yue, A., Zhao, Q., & Martorell, R. (2012). Nutrition and educational performance in rural China’s elementary schools: Results of a randomized control trial in Shaanxi Province. Economic Development and Cultural Change, 60(4), 735-772.

Luo, R., Miller, G., Rozelle, S., Sylvia, S., & Vera-Hernández, M. (2019a). Can bureaucrats really be paid like ceos? Substitution between incentives and resources among school administrators in China. Journal of the European Economic Association.1-37.

Luo, R., Emmers, D., Warrinnier, N., Rozelle, S., & Sylvia, S. (2019b). Using community health workers to deliver a scalable integrated parenting program in rural China: A cluster-randomized controlled trial. Social Science & Medicine, 239, 112545.

Ministry of Education of the People’s Republic of China. (2019). Statistical Announcement on the Implementation of National Education Funds. http://www.moe.gov.cn/srcsite/A05/s3040/201910/t20191016_403859.html

Ministry of Education of the People’s Republic of China. (2020). Overview of educational achievements in China in 2019. http://en.moe.gov.cn/documents/reports/202102/t20210209_513095.html#:~:text=Tremendous%20achievements%20were%20made%20in,sectors%20at%20all%20educational%20stages.&text=In%202019%2C%20China%20had%20530%2C000,from%202019%2C%20up%202.4%25.

Miller, G., Luo, R., Zhang, L., Sylvia, S., Shi, Y., Foo, P., Zhao, Q., Martorell, R., Medina, A., & Rozelle, S. (2012). Effectiveness of provider incentives for anaemia reduction in rural China: A cluster randomised trial. BMJ, 345, e4809.

Mo, D., Swinnen, J., Zhang, L., Yi, H., Qu, Q., Boswell, M., & Rozelle, S. (2013). Can one-to-one computing narrow the digital divide and the educational gap in China? The case of Beijing migrant schools. World Development, 46, 14-29.

National Bureau of Statistics of China. (2001). http://www.stats.gov.cn/tjsj/tjgb/ndtjgb/qgndtjgb/200203/t20020331_30003.html

National Bureau of Statistics of China. (2011). http://www.stats.gov.cn/tjsj/pcsj/rkpc/6rp/indexch.htm

National Bureau of Statistics of China. (2021). http://www.stats.gov.cn/tjsj/pcsj/rkpc/7rp/zk/indexch.htm

Park, A., Cai, F., & Du, Y. (2010). Can China meet its employment challenges?. Growing pains: Tensions and opportunity in China’s transformation. The Walter H. Shorenstein Asia-Pacific Research Center, Stanford, 27-56.

Paredes, V. (2014). A teacher like me or a student like me? Role model versus teacher bias effect. Economics of Education Review, 39, 38-49.

Pelayo, I., & Brewer, D. J. (2010). Teacher quality in education production. In International encyclopedia of education (pp. 438-442). Elsevier Ltd.

Podgursky, M., Springer, M. G., & Hutton, R. (2010). “Teacher Training and Preparation in the United States,” in Dominic Brewer and Patrick McEwan, eds. Economics of Education, pp. 290-295.

Psacharopoulos, G., & Patrinos, H. A. (2018). Returns to investment in education: a decennial review of the global literature. Education Economics, 26(5), 445-458.

Rothstein, R. (2010). “Family Environment in the Production of Schooling,” In Dominic Brewer and Patrick McEwan, eds. Economics of Education, pp. 148-155.

Schultz, T. W. (1961). Investment in human capital. The American Economic Review, 51(1), 1-17.

Shafiq, M. N., Toutkoushian, R. K., & Valerio, A. (2019). Who benefits from higher education in Low-and Middle-Income countries?. The Journal of Development Studies, 55(11), 2403-2423.

Shi, Y., Zhang L., Chang F., Liu H., Tang B., Gao Q., Guan H., Nie J., Yang J., Bai Y., Li Y., Tang L., Yue A., & Ru T. (2020). Practices and Experiences of Randomized Controlled Trials in China’s Targeted Poverty Alleviations in Education (in Chinese). Journal of East China Normal University (Educational Sciences), 38(8): 1-67.

Wong, H. L., Shi, Y., Luo, R., Zhang, L., & Rozelle, S. (2014). Improving the health and education of elementary schoolchildren in rural China: Iron supplementation versus nutritional training for parents. Journal of Development Studies, 50(4), 502-519.

Xie Y., Li, W., & Jin, Y. (2019). The Popularization of High-school Education: Background, Significance and Challenges (in Chinese). Journal of Peking University (Philosophy and Social Sciences), 2019, 56(4): 122-135.

Yue, A., Shi, Y., Chang, F., Yang, C., Wang, H., Yi, H., Luo, R., Liu, C., Zhang, L., Chu, J., & Rozelle, S. (2014). Dormitory management and boarding students in China’s rural primary schools. China Agricultural Economic Review, 6(3): 523-550.

Yue, A., Tang, B., Shi, Y., Tang, J., Shang, G., Medina, A., & Rozelle, S. (2018). Rural education across China’s 40 years of reform: past successes and future challenges. China Agricultural Economic Review, 10(1), 93-118.

Yue A., Cai J., Bai Y., Tang L., Shi Y., Luo R., & Rozelle, Scott. (2019). Challenges and Possible Solutions for Children 0-3 Years Old in Poor Rural China (in Chinese). Journal of East China Normal University (Educational Sciences), 37(3): 1-16.

Zhou C., Sylvia, S., Zhang, L., Luo R., Yi, H., Liu C., Shi, Y., Loyalka, P., Chu, J., Medina, A., & Rozelle, S. (2015). China’s left-behind children: Impact of parental migration on health, nutrition, and educational outcomes. Health Affairs, 34(11), 1964-1971.

Zhou, H., Sun, S., Luo, R., Sylvia, S., Yue, A., Shi, Y., Zhang, L., Medina, Alexis., & Rozelle, S. (2016). Impact of text message reminders on caregivers’ adherence to a home fortification program against child anemia in rural western China: A cluster-randomized controlled trial. American Journal of Public Health, 106(7), 1256-1262.