Investigating the association between perceived risk levels and commuting mode shifts after the lifting of the COVID-19 pandemic prevention and control policies

•Investigating commuters' perceived risks of contracting COVID-19 and their commuting mode transfers after anti-pandemic policies were lifted.•More than 60% of respondents had a high-perceived risk level of contacting COVID-19.•More than half of public transport commuters turn to private vehicles.•There was significant association between commuters' perceived risks and mode transfers.•Commuters with higher perceived risk levels, longer commuting distances, or car ownership were more likely to switch from other modes to private cars. There is a research gap in understanding people's perceived risks and their commute mode shifts after the major shift in anti-pandemic policies. Our study aims to reveal the relationship between commuters' perceived risks and their commuting mode transfers in the specific context of canceling anti-pandemic policies. We conducted an online sample survey of residents in 6 neighborhoods after one month the lifting of anti-pandemic policies in Kunming, China. Measured perceived risk data suggested that a perceived risk score of 23 ∼ 30 accounted for 62 % of the respondents, who were defined as the high-perceived risk group; while the perceived risk score of 14 ∼ 22 accounted for 36 % of the respondents, who were defined as the middle-perceived risk group; only 2 % of respondents with a perceived risk score of 6 ∼ 14. Commuting mode transfer statistics showed that 22.2 % of the respondents switched from other commuting modes to private cars, of which 56.1 % came from public transportation. Conversely, out of 81 car commuters, only 3 respondents moved to other commuting modes. We used nonparametric tests to find that there were group differences in commuting mode shifts. Specifically, the proportion of commuters with high-perceived risk levels shifted from other travel modes to private cars was 11% larger than that of commuters with middle-perceived risk levels. Public commuters were more likely to switch to car commuting than active commuters. The nonparametric test results also showed that single variables such as car ownership, commute distance, age, and marital status was significantly correlated with the distribution of the shifting in commuting mode. Furthermore, we employed a binary logistic regression model to reveal that commuters with higher perceived risk levels, longer commuting distances, or car ownership were more likely to switch from other travel modes to private cars than other commuters. The conclusion of this study