After more than 1 year into the COVID-19 pandemic, governments worldwide still face the challenge of adopting non-pharmaceutical interventions to mitigate the risks posed by the emergence of new SARS-CoV-2 variants and the lack of a worldwide equitable vaccine allocation. Thus, it becomes crucial to identify the drivers of mobility responses to mitigation efforts during different restriction regimes, for planning interventions that are both economically and socially sustainable while effective in controlling an outbreak. Here, using anonymous and privacy-enhanced cell phone data from Italy, we investigate the determinants of spatial variations of reductions in mobility and co-location in response to the adoption and the lift of restrictions, considering both provinces and city neighbourhoods. In large urban areas, our analysis uncovers the desertification of historic city centres, which persisted after the end of the lockdown. Such centre-periphery gradient was mainly associated with differences in educational attainment. At the province level, the local structure of the labour market mainly explained the variations in mobility responses, together with other demographic factors, such as the population’s age and sex composition. In the future, targeted interventions should take into account how the ability to comply with restrictions varies across geographical areas and socio-demographic groups.

COVID-19 大流行一年多后，世界各国政府仍然面临着采取非药物干预措施来减轻新 SARS-CoV-2 变体出现和全球范围内缺乏公平疫苗分配所带来的风险的挑战。因此，确定不同限制制度下缓解措施的流动性反应的驱动因素变得至关重要，以便规划经济和社会可持续的干预措施，同时有效控制疫情。在这里，我们使用来自意大利的匿名和隐私增强型手机数据，研究了流动性和同地减少的空间变化的决定因素，以响应限制的采用和取消，同时考虑到各省和城市社区。在大城市地区，我们的分析揭示了历史悠久的城市中心的荒漠化，这种荒漠化在封锁结束后仍然持续存在。这种中心-外围梯度主要与教育程度的差异有关。在省一级，当地劳动力市场结构主要解释了流动反应的变化，以及其他人口因素，如人口年龄和性别构成。未来，有针对性的干预措施应考虑到不同地理区域和社会人口群体遵守限制的能力有何不同。