Direct Consequences of Social Distancing Policies on Mobility During COVID-19 in South Korea

OBJECTIVES: This study aimed to investigate the impact of social distancing policies on mobility in South Korea during the COVID-19 pandemic in 2020.

METHODS: The research period encompassed two nationwide pandemic phases in late 2020 in South Korea, following the implementation of the social distancing policy. Mobility data from Google community mobility report, which show the percent change in visit or duration of stay compared to the baseline, were utilized. To assess the impact of social distancing on mobility, Pooled Ordinary Least Squares (OLS) was employed to examine overall policy effects, while a fixed-effects approach addressed regional variations and day-of-week effects. The number of new COVID-19 cases with a one-day lag for each region was included as control variable. For metropolitan areas, an interrupted time series (ITS) analysis was conducted to assess impact over time, considering a broader range of confounding factors.

RESULTS: In the first study period, Pooled OLS analysis revealed the following changes: When social distancing measures were first tightened, there was a decrease of 17.7% in mobility; for subsequent tightening, there was an additional decrease of 10.6% (R-squared: 75.9%). In the second period, we observed the changes of -12.0%, +2.8%, and -8.7% respectively in response to tightening social distancing levels (R-squared: 73.7%). When fixed effects were added to the OLS model, the policy’s effect slightly diminished compared to the OLS model. However, a significant impact on mobility reduction was still observed. When analyzing changes over time using ITS, a gradual increase in mobility reflecting fatigue was observed one month after the implementation of the policy.

CONCLUSIONS: The research findings indicate that social distancing policies had a significant impact on reducing mobility in South Korea during the COVID-19 pandemic. This effect was observed even when controlling for other variables that could influence mobility.