Table 5.

Alternative specifications

	Must ID assault	Must ID assault, no crime	cluster	Combined cities	Include late May
	(1)	(2)	(3)	(4)	(5)
Fraction of people staying at home	0.8591*** (0.2749)	1.3480*** (0.3569)	0.9922** (0.4639)	0.9957*** (0.2522)	1.0178*** (0.2453)
Observations	3770	2610	5220	4640	5400
R2	0.9058	0.9157	0.8977	0.9052	0.8990

All regressions are weighted by population and include local maximum temperature, the lagged domestic violence rate, insured unemployment rate, and month, day of week, city, and holiday dummies. Robust standard errors are in parentheses. Robust standard errors clustered at state level are in parentheses in column 3. ^*p < 0.10, **p < 0.05, ***p < 0.01