Why does the new institution often go back to its original state after its implementation for a while? How to prevent favorable institutions from having regressive effect is a frequently recurring difficulty in the institutional change and reform. It is necessary and vitally important to study these questions to promote the theoretical development of institutional changes and optimize the current institutional reform and change. This paper has, based on the convenient sampling to control effectiveness, chosen 21 cases from different places, fields and levels. It has adopted the method of cross-case cluster analysis and discussed the regressive effect of institutional change and anti-regression mechanism. Research findings have shown that there are six main influencing factors causing the regressive effect of institutional change. They are: a. the will pullback of the main body of institutional supply; b. the intervention of groups with reverse interest; c. the lack of combining institutions; d. the imbalance of the goals of institutional changes; e. the inertia conflict of the old institution; f. the low embedding strength of informal restraint. Meanwhile, the paper has found that an effective anti-regression mechanism should meet the following six basic design principles: a. the constant progress of willpower entity of institutional supply; b. the participation and balance of multiple entities; c. the construction of combined institutional system; d. the coordination and integration of influencing factors of supply goals; e. the adoption of slow-moving institutional change; f. strengthening the guiding function of informal restraint. These findings not only have guiding significance and instructive value on the development of theories of institutional changes, but also on the design, reform and change of actual institutions.