Fig. 2: Adherence change estimates from mobile-phone mobility data.

a,b, Coefficients representing changes in mobility as measured by time spent in residential locations (a) and retail and recreation visits (b) relative to the baseline period (0–30 days after first required measure), controlling for containment policies. These were estimated in multi-level models with two levels: time (level 1, days) and countries (level 2) with a first-order auto-regressive residual covariance structure. Dashed lines represent baseline behaviour, and error bars indicate 95% confidence intervals. The apparent non-linear pattern is confirmed in Tables 2 and 3, which present results of models with additional controls and residual models with significant quadratic trends. See Supplementary Fig. 2a–e and Supplementary Table 1a,c,d,f–i,l,m for models with alternative specifications (including fixed-effects models with clustered standard errors at the country level) and sets of control variables.