Factors driving the implementation of the 'Local New Year' policy to prevent COVID-19 in China

The Spring Festival Transportation (SFT) in China is the largest periodic human migration globally, involving over 2 billion people. Since the 1980s, increased urbanization has led to a significant migrant worker population. The SFT, a crucial cultural event, represents the annual homecoming for millions. The COVID-19 pandemic drastically impacted SFT, causing a significant decrease in passenger volume in 2020 and 2021 due to government-imposed travel restrictions aimed at controlling the virus spread. The 'local new year' policy, encouraging people to stay in their work cities, was a significant measure in controlling the initial outbreak and was again suggested in 2022. This policy, however, created a conflict between pandemic prevention and people's emotional need to return home for the festival. This study aims to understand the factors driving the willingness of people, particularly migrant workers, to support the 'local new year' policy, aiming to inform the development of more effective and humane pandemic prevention strategies.

Previous research highlights the link between population mobility and pandemic spread. Studies show that increased population mobility correlates with a higher number of confirmed cases. Reducing mobility is crucial in controlling infection rates. Existing literature examines the behavioral changes during the pandemic, such as increased cooking at home, reduced shopping, and less frequent travel. While these studies focus on the changes in behavior, they lack in-depth analysis of the underlying reasons behind these shifts. This study aims to fill this gap by exploring the specific factors influencing people's decision-making during the pandemic, especially around the culturally significant Spring Festival.

This study employed a two-stage questionnaire survey targeting migrant workers in large Chinese cities. The first survey identified potential influencing factors related to the 'local new year' policy. The second survey collected data on these factors, alongside demographic information from 584 valid responses. The Cp method, based on Mallows' Cp, was used to screen and select the primary influencing factors from the data collected in the second survey. This method helps to prevent overfitting in the model by selecting the best combination of independent variables that minimize the mean square error. A Logit model, a discrete choice model, was then constructed using maximum-likelihood estimation to analyze the relationship between the primary influencing factors and the probability of choosing to celebrate a 'local new year'. The Logit model allowed for the estimation of the relative importance of each influencing factor in predicting the likelihood of supporting the policy.

The Cp method identified four primary factors influencing the decision to support 'local new year': (1) presence of relatives (elderly/children) in the hometown; (2) ability to buy travel tickets; (3) risk of contracting COVID-19 during travel; and (4) availability of activities in the city of work. Further analysis using the Logit model and maximum likelihood estimation revealed that 'availability of activities in the city of work' was the most significant predictor of supporting the policy, having a much greater impact than the other factors. 'Presence of relatives (elderly/children) at home' and 'risk of contracting COVID-19 during travel' were identified as secondary factors. The analysis showed that those with no activities arranged in their work city were 100 times more likely to go home than those with arrangements, highlighting the importance of social and recreational provisions in the policy's effectiveness. Those who did not fear contracting COVID-19 during travel were 8.4 times more likely to return home compared to those who were concerned, emphasizing the importance of perceived risk. The study also found the presence of relatives was a factor with a significant influence (7.5 times more likely to go home for those with relatives compared to those without). The analysis using the Logit model and maximum-likelihood estimation provided further insights into the relative importance of these factors. Table 6 and Table 7 present the results of the maximum-likelihood estimation analysis, demonstrating the statistical significance of these factors and their odds ratios.

The findings highlight the crucial role of social and recreational activities in influencing the success of the 'local new year' policy. The strong impact of this factor suggests that focusing solely on material compensation, as seen in some existing policies, may be insufficient. The study's results suggest a shift towards prioritizing the provision of culturally relevant and engaging activities for migrant workers during the Spring Festival to encourage compliance with the policy. The significant influence of having relatives in the hometown reinforces the importance of addressing the emotional and familial aspects of the policy. The fear of contracting COVID-19 remains a considerable factor, but the availability of activities significantly outweighs the risk aversion, suggesting a need for better alternatives in the work city. These findings offer a valuable contribution to the understanding of the factors shaping behavior during public health emergencies, demonstrating that a holistic approach to policy-making that addresses both practical and emotional needs is essential.

This study provides valuable insights into the factors driving the adoption of the 'local new year' policy in China. The findings demonstrate the primacy of providing local activities and cultural programs to encourage compliance. Material incentives alone are not sufficient. Future research should focus on more nuanced cultural and contextual aspects, regionally tailored strategies, and the effectiveness of different types of activity programs in fostering a sense of community and well-being. The methodology and results of this study can be applied to inform pandemic response strategies in other contexts.

The study's reliance on self-reported data through online questionnaires may introduce biases. The sample, while relatively large, may not be fully representative of all migrant workers in China. The cross-sectional nature of the study limits the ability to make strong causal inferences. Finally, the study focused primarily on migrant workers and may not fully capture the experiences of other demographic groups.