The paper utilizes the statistical methods of variable selection and model selection, together with the computation power of language R, and builds the ARIMA model and dynamic regression model respectively, which is different from the commonly used modelling methods in econometrics. Since cross validation is also performed on the ARIMA prediction model, satisfactory results from the model have shown that it is reasonable to conclude that the growth rate of the number of Chinese high school graduates who will study aboard is going to slow down. The dynamic regression model quantifies to what extent the key factors are affecting the number of people studying aboard on the other hand, and ranks the factors affecting import of china’s higher education service in the following manner: employment opportunity, number of high school graduates, and average dispensable income. Further analysis reveals that the lack of global competitiveness of China’s higher education industry is the root cause for the import to remain high.