New equilibrium? Dynamics of intercity mobility in China during COVID-19 pandemic period

1. Introduction

Travel restrictions has been widely adopted by public authorities globally to reduce the spread of COVID-19 pandemic. However, the government and public acceptability of strict restrictions on intercity mobility had been depleted to varying levels when the more transmissible/deadly variants of the virus wreaked havoc. Paradoxically, while travel restrictions may reduce risks, they also degrade mobility and reduce the critical socio-economic functionality(Massaro et al., 2018). As a result, a dynamic ‘new equilibrium’ will emerge through travel behavioral adaptations impacted by the long-term interaction between COVID-19 prevention and mobility recovery(van Wee and Witlox, 2021).

With various control measures (e.g., social distancing, mask-wearing, vaccination, Internet plus big data technology), the COVID-19 has been well contained in China(Chen et al., 2021). However, service industries (e.g., transport, tourism) are still greatly influenced by COVID-19 and has not totally recovered yet (Wang et al., 2021). Specifically, the recovery of human mobility is not only related to pandemic, but also to the resulting pandemic prevention policy, transport supply and travel psychology. For instance, the decrease in intercity travel demand can be attributed to travel behavioral adaptions due to the perceived risk of infection. Especially, China have experienced multiple local waves over the past two years, with a discrepancy between policy intervention and severity of infection and mortality. As control policy is adjusted according to the dynamics of the COVID-19, the paradox between control measures and mobility recovery resulting in dynamic of intercity mobility do not invalidate but the China experience helps to contextualize them. Therefore, this paper aims to study on the dynamics of intercity mobility with COVID-19 in China for a two-year period (January 1, 2020 to December 31, 2021).

2. Data and methods

This study uses daily intercity mobility data collected from the Baidu Huiyan (https://qianxi.baidu.com/). This website publishes the intercity mobility index (IMI) data, namely the relative measure of the volume of daily inbound and outbound passengers with the LBS service for each city (Li et al., 2021). In 2020, there were 932 million smartphone users in China, accounting for 64.54% of population in China. We obtained the data from January 1 to May 22 of 2019, January 1 to May 5 of 2020, September 22 to December 31 of 2020, and the whole year of 2021. Based on the existing data, spatial variations in IMI changes were analyzed by confining the study scope in weekdays and national holiday, namely Labor Day holiday (between 1 May and 5 May). Besides, we obtained the monthly passenger volume (MPV) data from 2019 to 2021 collected by China's Ministry of Transport (MOT), which shows the intercity public transport use. The year of 2019 is considered as the baseline for comparison. Specifically, the loss rate of IMI was calculated as the ratio of IMI loss to the regular IMI in 2019.

3. Result

Fig. 1 shows the spatial variation of IMI. Cities with relatively higher IMI loss ratio or lower recovery ratio were mainly distributed in megacities, such as Beijing, Shanghai, Wuhan and provincial capitals, implying that the IMI of these hub and core cities has a long-term recovery trend. Additionally, the IMI also recovered slowly in cities of northeast region, where several local waves of COVID-19 were reported. Conversely, the recovery ratio of IMI is relatively higher and nearly similar beyond these core cities, especially in the labor-exporting cities of south-western China. The recovery ratio is also higher in northwestern of China as this area was relatively unaffected by COVID-19. The impact on holiday intercity travel has gradually recovered in 2021, with IMI nearly restored to the pre-pandemic level. To summarize, the spatial pattern of the largest scale of IMI loss matches the geographical characteristics of the spatial gradient difference on socio-economic development in China(Gibbs et al., 2020).

Fig. 1.

Distribution of intercity mobility at city level, China.

Note: Borders in the figure refer to the Ministry of Natural Resources of the People's Republic of China (http://www.gov.cn/guoqing/2005-09/13/content_5043917.htm).

The average loss ratio of IMI for weekdays at the city level in China is 30.4% for 2020 and 18.4% for 2021, respectively (Fig. 1 a-b). In other words, the average recovery ratio is 69.6% for 2020 and 81.6% for 2021, respectively. Regarding IMI of Labor Day holiday, the average IMI loss ratio of each city is 37.9% for 2020 and 7.4% for 2021, respectively (Fig. 1 c-d). Namely, the average recovery ratio of holiday IMI is 62.1% for 2020 and 92.6% for 2021, respectively. That is to say, the IMI gradually restored in China due to the combined effect of pandemic control, pandemic prevention policy, etc. Holiday intercity travel was more curbed than that in weekdays in the early days after the national wave of COVID-19, April 8, 2020.

Regarding MPV, up to December 31, 2021, the total MPV decreased almost 52.6% (9229 million persons) between 2019 and 2021. This suggests that transport passenger traffic is still greatly affected by the pandemic, and the recovery rate is less than half of the pre-pandemic level by the end of 2021. Among the four transport modes, the highway suffers the greatest loss, accounting for 60.9% (7925 million persons) of the total MPV loss, and has a lower level of recovery. Conversely, civil aviation suffers the least loss, accounting for 25.0% (146 million persons) of the total MPV loss, and showed a fluctuating high recovery level. Possible reason is that air transport is less regulated in China than railway, and airlines can re-schedule flights and adjust airfares according to market demand (Wang et al., 2021). The observed difference in recovery ratio between IMI and MPV supports previous studies on the modal shift influenced by COVID-19, implying the importance of car driving in intercity travel during the pandemic(van Wee and Witlox, 2021). In summary, although the IMI at the city level recovered to over 80% by the end of 2021 as the COVID-19 came under control, they have not fully recovered to pre-pandemic level over nearly two years, especially in core cities and weekdays. It is possible that the dynamic interaction between pandemic prevention and mobility recovery resulted in extended recovery periods or new equilibrium in intercity travel.

4. Conclusion

This study examined the dynamic of intercity mobility during the COVID-19 in China over the last two years. Although the IMI at the city level gradually restored as the COVID-19 came under control, they have not fully returned to pre-pandemic level over nearly two years, especially for core cities and weekdays. Highway has the greatest loss and low level of recovery, while civil aviation has the least loss and a fluctuating high recovery trend due to the less regulation of Chinese airline. We find clear spatial disparities in intercity mobility recovery across cities associated with their location, socioeconomic level, and pandemic level. Cities with relatively lower intercity recovery ratio were mainly distributed in megacities (Beijing, Shanghai) and provincial capitals, indicating that the long-term recovery trend of these core cities.

Although it is quite early to accurately answer the enduring changes of mobility impacted by COVID-19 since the pandemic is still unfolding, this paper at least provides new evidence on this question by incorporating the Chinese experience. An in-depth analysis of the dynamics on the enduring effect on mobility brought by pandemic could be further discussed.

CRediT authorship contribution statement

Tao Li: Conceptualization, Investigation, Methodology, Formal analysis, Writing – original draft, Writing – review & editing. Leibo Cui: Formal analysis, Writing – original draft, Visualization. Jiaoe Wang: Conceptualization, Investigation, Writing – original draft, Writing – review & editing.

Data availability

The authors do not have permission to share data.