This paper uses housing policies from January 2019 to August 2024 as keywords to conduct content searches on the housing preference choices of college student entrepreneurs at the national, provincial, and municipal levels. After text mining using ROSTCM software, a quantitative evaluation was conducted using the Policy Index Model (PMC). Subsequently, propensity score matching and a multi-period double difference model were employed to calculate and match the impact of housing policy on the housing preferences of college student entrepreneurs, analyze the heterogeneity among different types of housing, and investigate the effects of housing policy implementation on the housing preferences of college student entrepreneurs. The results indicate that migration duration, migration scope, and the type of destination city all significantly influence the housing preference choices of college student entrepreneurs. The government primarily relies on environmental policy tools to exert indirect influence, while supply-side and demand-side tools have limited direct supply and incentive effects. The evaluation grades of the nine policies fall within the range of good to acceptable, with PMC scores ranging from 4.000 to 7.294. Under different housing policies, the impact on the housing preference choices of college student entrepreneurs is related to the number of properties owned by their families and their urban-rural distribution. For example, college students from families with two or three properties exhibit higher or lower preference choices under different housing policies compared to those from families with one property, and this effect exhibits heterogeneity across regions and urban-rural areas.
