Online education has been adopted nationwide due to the sudden break of Covid-19. Compared to the traditional education method, online education not only relies on good network condition, but also requires higher standard of home environment. Therefore, the learning outcomes of students with different family conditions may differ after the online education period. This paper takes the Covid-19 as an exogenous shock and discusses equity issues of online education under a quasi-natural experiment. Considering the internet penetration is relatively lower in middle and western areas, this paper uses a sample of students in the twelfth grade and ninth grade from certain schools in a city of middle areas, and explores the impact of family conditions including poverty, rural and only-child family on the quality of online education.

Based on the estimation using the DID method, scores of poor and rural students in the twelfth grade, as well as poor students in the ninth grade have declined after the Covid-19. In the perspective of discipline difference, the performance of science students in the twelfth grade is mainly influenced by poverty and the only-child condition, while liberal students are more sensitive to rural conditions. The gender gap in Math scores of the ninth grade widens after the Covid-19, while students in non-only child family experience improvement in Chinese scores. The paper further makes robust checks based on the parallel trend hypothesis and the bootstrap method. Although online education has been proved as an effective way to break the regional education monopolization and improve education equity, it also brings financial burdens to students and their families, and in that case, it may not be able to narrow the gap between students in the same region.

The marginal contribution and significance of this research are as follows: Firstly, this paper takes the break of Covid-19 as an exogenous shock and solves the endogeneity problem caused by sample selection under this quasi-experimental analysis. Secondly, this paper uses micro-level data of students and objective indicators as exam scores and family conditions to avoid subjectivity in personal evaluation. Finally, this paper not only supplements the research on the social socioeconomic influence of Covid-19, but also has practical significance on the further promotion of online education under educational informationization plan. Based on the preliminary study in this paper, there is still enough space for further study on the extension of samples and indicators, the long-term performance of education and employment after the Covid-19, as well as the comparison of impacts among different regions.