Table 3. Robustness check for the OLS estimates of beta.

Check	Width of River	Distance of River from city centre	Number of obs.	F	R2
First part: Sensitivity check across temporal domain
A. Using only plague outbreak during AD1347–1449
	0.1096634*** (0.3507051)	−1.053141* (−0.742794)	777	6.00	0.50
B. Using only plague outbreak during AD1450–1549
	0.0929646*** (0.315163)	−1.525797*** (−0.1034473)	1660	8.15	0.38
C. Using only plague outbreak during AD1550–1649
	0.1194134*** (0.404504)	−1.267226*** (−0.1084678)	2667	9.84	0.31
D. Using only plague outbreak during AD1650–1760
	0.0198132** (0.1449958)	−1.186286*** (−0.1735572)	455	2.28	0.27
Second part: Sensitivity check across spatial domain
E. Aggregating cities with the same number of plague outbreak
	0.1272769*** (0.3794992)	−10.69387** (−0.3156618)	56	11.10	0.30
F. Dropping all cases in Russia and Turkey before aggregating cities with the same number of plague outbreak
	0.1266675*** (0.3800006)	−10.6744** (−0.3155398)	56	11.18	0.30
G. Aggregating cities with the same number of plague outbreak only in UK, Germany, France and Italy
	0.1417548*** (0.424572)	−8.895052* (−0.2789527)	54	11.05	0.30

Notes: The dependent variable of check A, B, C and D is the total number of plague reoccurrence within our study period. Time fixed effect and regional fixed effect were added to the model. For check E, F and G, the dependent variable is number of count of plague outbreak. Width of river and distance of river away from city centre are aggregated and averaged for check E, F and G before statistical analysis. No fixed effect was added to the model.

***p < 0.005; **p < 0.01; *p < 0.05.