Table 6 Air quality, donations, and...

Table 6

Air quality, donations, and volunteers: entropy-balanced tests

Panel A. Covariate balance pre- and post-entropy balancing
	Mean – Pre-entropy balancing			Mean – Post-entropy balancing
	No Hazardous Days	Any Hazardous Days	Stdzed diff. in means	No Hazardous Days	Any Hazardous Days	Stdzed diff. in means
Fundraising Expenses	0.064	0.042	−0.100	0.040	0.042	0.000
Age	2.963	2.919	−0.050	2.918	2.919	0.000
Program Expense Ratio	0.812	0.815	0.010	0.816	0.815	0.000
Size	1.139	0.953	−0.140	0.953	0.953	0.000
Government Grants	0.201	0.224	0.040	0.224	0.224	0.000
Program Service Revenue	0.500	0.377	−0.120	0.376	0.377	0.000
Employees	1.975	1.881	−0.050	1.882	1.881	0.000
Governance Index	4.375	4.043	−0.230	4.042	4.043	0.000
Rural	0.002	0.003	0.010	0.003	0.003	0.000
Poverty Rate	10.096	10.960	0.210	10.960	10.960	0.000

Panel A. Covariate balance pre- and post-entropy balancing
	Mean – Pre-entropy balancing			Mean – Post-entropy balancing
	No Hazardous Days	AnyHazardous Days	Stdzed diff. in means	No Hazardous Days	AnyHazardous Days	Stdzed diff. in means
Fundraising Expenses	0.064	0.042	−0.100	0.040	0.042	0.000
Age	2.963	2.919	−0.050	2.918	2.919	0.000
Program Expense Ratio	0.812	0.815	0.010	0.816	0.815	0.000
Size	1.139	0.953	−0.140	0.953	0.953	0.000
Government Grants	0.201	0.224	0.040	0.224	0.224	0.000
Program Service Revenue	0.500	0.377	−0.120	0.376	0.377	0.000
Employees	1.975	1.881	−0.050	1.882	1.881	0.000
Governance Index	4.375	4.043	−0.230	4.042	4.043	0.000
Rural	0.002	0.003	0.010	0.003	0.003	0.000
Poverty Rate	10.096	10.960	0.210	10.960	10.960	0.000

Panel B. Entropy-balanced sample regressions
	DV =
	Donations (1)		Volunteers (2)
Any Hazardous Days	−0.014**	(−19.514)	0.069**	(17.995)
Fundraising Expenses_t-1	1.636**	(696.277)	2.584**	(207.790)
Age_t	−0.035**	(−96.933)	0.184**	(95.601)
Program Expense Ratio_t-1	0.147**	(98.051)	0.102**	(12.785)
Size_t-1	0.183**	(453.191)	0.003	(1.468)
Government Grants_t-1	−0.076**	(−128.822)	−0.179**	(−57.235)
Program Service Revenue_t-1	−0.162**	(−302.789)	−0.256**	(−90.306)
Governance Index_t-1	0.029**	(121.913)	0.351**	(282.566)
Employees_t-1	0.043**	(188.731)	0.191**	(157.679)
Rural_t-1	−0.030**	(−5.717)	−0.545**	(−19.625)
Poverty Rate_t-1	−0.002**	(−22.590)	−0.036**	(−77.956)
Industry, State, and Year Fixed Effects	YES		YES
Observations	1,633,609		1,633,609
F	18,070**		4,971**
Adj. R²	0.485		0.205

Panel B. Entropy-balanced sample regressions
	DV =
	Donations (1)		Volunteers (2)
Any Hazardous Days	−0.014**	(−19.514)	0.069**	(17.995)
Fundraising Expenses_t-1	1.636**	(696.277)	2.584**	(207.790)
Age_t	−0.035**	(−96.933)	0.184**	(95.601)
Program Expense Ratio_t-1	0.147**	(98.051)	0.102**	(12.785)
Size_t-1	0.183**	(453.191)	0.003	(1.468)
Government Grants_t-1	−0.076**	(−128.822)	−0.179**	(−57.235)
Program Service Revenue_t-1	−0.162**	(−302.789)	−0.256**	(−90.306)
Governance Index_t-1	0.029**	(121.913)	0.351**	(282.566)
Employees_t-1	0.043**	(188.731)	0.191**	(157.679)
Rural_t-1	−0.030**	(−5.717)	−0.545**	(−19.625)
Poverty Rate_t-1	−0.002**	(−22.590)	−0.036**	(−77.956)
Industry, State, and Year Fixed Effects	YES		YES
Observations	1,633,609		1,633,609
F	18,070**		4,971**
Adj. R²	0.485		0.205

Note(s): In line with Hainmuller (2012), Panel A shows mean values and standardized difference in means for the control variables in the low and high air pollution groups before and after entropy balancing. Panel B shows coefficients from weighted regression (using weights derived from entropy balancing), with t-stats in parentheses; *p < 0.05, **p < 0.01 (two-tailed). Any Hazardous Days is a binary variable indicating organizations in a county with one or more Hazardous Days. See Appendix for variable definitions.

Source(s): Authors' own creation

[ViewLarge]

Sharing Unavailable