China's low consumption rate has been called the "Chinese mystery" by the international academic community. Some scholars have suggested that increasing social security levels can stimulate consumption. For the insurance system, it includes two stages, which are insurance paying stage and pension receiving stage. Some scholars have proved that increasing the pension replacement rate at the receiving stage has a significant promotion effect on consumption, but few scholars study the impact of pension contribution rate on consumption.

This article focuses on the impact of the Primary Pension Insurance For Urban Employees(PPIUE) contribution rate on household consumption. Using the CHIP2013 and CHIP2018 national sample survey data, applying the two-stage least squares regression method, choosing the average contribution rate within the industry and the province as the instrumental variable of household contribution rate, we analyzed and discussed the impact of the PPIUE on household consumption.

The impact of pension insurance contribution rate on consumption can be divided into "Current Income Effect" and "Expected Income Effect", the former will reduce consumption, while the latter will promote it, thus weakening the effect of contribution rate on consumption to a certain extent. And due to the existence of the Expected Income Effect, the impact of pension contribution rate on consumption may be non-linear. We found that: (1) For households participating in PPIUE, for every 1 percentage point increase in household consumption, household consumption in 2013 will decrease by 1.47%, while household consumption in 2018 will decrease by 1.25%. The increased income expectations in 2018 may be the reason of the difference. (2) Family age has a moderating effect on the relationship between the contribution rate and consumption: Compared with young families, mature families have stronger expected income effect , so the contribution rate may have a larger decreasing effect on the consumption of young families. (3) Pension contribution rate has a significant crowding out effect on the consumption of low-income families. Low-income families have a stronger current income effect, so consumption levels are more sensitive to changes in contribution rates. In addition, the contribution rate itself may have a nonlinear effect on consumption. (4) In terms of consumption structure, the increase in the contribution rate will squeeze out food, clothing, transportation & communication expenditures, but have no significant effect on inflexible expenditures like housing, daily necessities and services, education & entertainment, and healthcare.

Finally, we changed the measurement indicators of pension insurance contributions and the model estimation method to test the robustness of the model. The regression results are consistent, indicating that the empirical conclusions of this article are reliable.

Based on the above discussion, this article puts forward policy recommendations. When advocating to improve the pension insurance level to stimulate consumption, it is necessary to consider the crowding out effect of contribution rate on the consumption. In addition, improving the construction of the pension system and enhancing the expected income effect can alleviate the inhibitory effect of contribution rate on household consumption.