Table 5.

Restricted sample: Sensitivity overview

	Dependent variable
	Cases	log Cases	R0	log R0
	(1)	(2)	(3)	(4)
Original model	− 1.007	− 0.350	− 0.332∗	− 0.385∗
	(0.904)	(0.293)	(0.173)	(0.189)
w/o Košice district	− 1.347	− 0.492	− 0.320	− 0.361
	(0.996)	(0.324)	(0.198)	(0.215)
Unweighted	− 1.427	− 0.519	− 0.340	− 0.411
	(0.968)	(0.336)	(0.222)	(0.257)
PCR only	− 0.741	− 0.301	− 0.387∗∗	− 0.420∗∗
	(0.706)	(0.277)	(0.179)	(0.191)
Considering mobility	− 0.926	− 0.319	− 0.319	− 0.360∗
	(1.021)	(0.320)	(0.193)	(0.201)
R0 based on 7 day avg			− 0.191	− 0.171
			(0.160)	(0.148)
7 day avg of R0			− 0.152	− 0.138
			(0.175)	(0.158)
Placebo regression — Threshold 1.2%	− 1.325	− 0.229	− 0.052	− 0.005
	(0.989)	(0.246)	(0.160)	(0.186)
Placebo regression — Threshold 0.5%	− 0.298	− 0.242	0.057	0.121
	(0.795)	(0.372)	(0.370)	(0.436)
Date of second mass testing (Nov 1)	0.470	0.145	− 0.109	− 0.082
	(1.141)	(0.312)	(0.258)	(0.201)

This table presents the results for the key parameter β₁ from Eq. (1) estimated on restricted samples (visualized on Fig. 4) with population sizes used as weights. Row 2 and 3 show the results when excluding large urban districts from our analysis and using unweighted regressions respectively (Appendix A2.1). Row 4 presents the results considering only test results from PCR data only (Appendix A2.2) and row 5 shows the results when including workplace mobility in the estimation (Appendix A2.3). Rows 6 and 7 present results based on R0 measure calculated on a 7-day average of infection cases and on a 7-day average of R0 respectively (Appendix A2.4). Finally, the last three rows show estimates from placebo regressions using arbitrary thresholds (using samples depicted in Fig. 14) or dates (Appendix A2.5)

Districts weighted by their population size

^∗∗∗p < 0.01; ^∗∗p < 0.05; ^∗p < 0.1