Table 6.

Elasticities of mobility to policy stringency: what drives the relationship between democracy and compliance?

	(1)	(2)	(3)	(4)	(5)	(6)
	Mobility	Mobility	Mobility	Mobility	Mobility	Mobility
Policy stringency	-67.05**	-100.4***	-99.34***	-63.22**	-63.30*	-27.98
	[31.11]	[24.36]	[25.64]	[25.90]	[33.90]	[22.38]
Policy stringency $\times$ welfare system (QoG)	1.007
	[2.298]
Policy stringency $\times$ human capital index (PWT)		4.632
		[7.206]
Policy stringency $\times$ tax revenue (% GDP) (WB)			-0.0438
			[0.530]
Policy stringency $\times$ bureaucratic quality (VDem)				3.010
				[3.645]
Policy stringency $\times$ access to Internet (QoG)					0.361
					[0.239]
Policy stringency $\times$ social capital (WVS)						-12.46***
						[4.054]
Observations	15,815	22,114	20,564	25,234	23,089	11,605
R-squared	0.740	0.662	0.652	0.665	0.669	0.708
Number of country-mobility category pairs	340	435	405	490	450	205
Controls	$✓$	$✓$	$✓$	$✓$	$✓$	$✓$
Country FE	$✓$	$✓$	$✓$	$✓$	$✓$	$✓$
Time FE	$✓$	$✓$	$✓$	$✓$	$✓$	$✓$

Notes: This table presents OLS estimates from regressing mobility on the policy stringency index and an interaction between policy stringency, the Polity democracy index and several variables aimed at explaining the relationship between democracy and compliance. All specifications include: i) the number of confirmed cases and its interaction with policy stringency.; ii) the interaction between cases and $D_c$, as well as a three-way interaction between cases, our measure of democracy and policy stringency; iii) the time varying index of testing policy; iv) the interaction between cases and a dummy for the accountability and transparency of government; v) the interaction between policy stringency and country-level log of GDP per capita; vi) the interaction between policy stringency and the number of mobile phone subscriptions per hundred people; and vii) the interaction between policy stringency and the dummy for experience with previous epidemics. Errors are clustered at the country-level. The coefficients with ${}^{★}{}^{★}{}^{★}$ are significant at the 1% level, with ${}^{★}{}^{★}$ are significant at the 5% level, and with ${}^{★}$ are significant at the 10% level.