Table 1.

Relationship between Racial Animus, Hate Crimes, and Chinese Restaurant Visits.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
VARIABLES	Incidents	Incidents	Incidents	Incidents	Visits	Visits	Visits	Visits
Panel A: Google search index
Google ch-word(t)	0.00057*	0.00057*	0.00057*	0.00057*	$-$21.069*	$-$22.017*	$-$23.489**	$-$24.381**
	(0.00031)	(0.00031)	(0.00031)	(0.00031)	(12.629)	(12.704)	(11.632)	(11.697)
Google ch-word($t-1$)			$-$0.00019	$-$0.00018			$-$13.934	$-$14.900
			(0.00035)	(0.00035)			(11.467)	(11.594)
Total visits					0.052***	0.052***	0.048***	0.048***
					(0.005)	(0.005)	(0.006)	(0.006)
Population(m)	0.51926***	0.51907***	0.52042***	0.52287***
	(0.19939)	(0.19942)	(0.19936)	(0.19991)
Google Asian(s)(t)		$-$0.00005		$-$0.00018		$-$138.636**		$-$120.800**
		(0.00109)		(0.00111)		(60.169)		(54.317)
Google Asian(s)($t-1$)				0.00075				10.816
				(0.00163)				(59.669)
Observations	3600	3600	3600	3600	1440	1440	1380	1380
R-squared	0.309	0.309	0.309	0.309	0.996	0.996	0.997	0.997
Outcome mean	0.147	0.147	0.147	0.147	104962.736	104962.736	104962.736	104962.736
Panel B: Twitter post index
Twitter ch-word	$-$0.00041	$-$0.00037	$-$0.00048	$-$0.00047	$-$60.249	$-$61.190	$-$30.883	$-$30.178
	(0.00099)	(0.00096)	(0.00099)	(0.00100)	(49.798)	(50.150)	(42.641)	(42.737)
Twitter ch-word ($t-1$)			$-$0.00063	$-$0.00064			$-$10.216	$-$10.089
			(0.00065)	(0.00065)			(36.060)	(36.286)
Total visits					0.047***	0.047***	0.044***	0.044***
					(0.005)	(0.005)	(0.006)	(0.006)
Population(m)	0.56488***	0.56517***	0.58097***	0.58136***
	(0.17387)	(0.17391)	(0.17623)	(0.17630)
Twitter Asian(s)(t)		$-$0.00003		$-$0.00002		0.294		$-$0.279
		(0.00003)		(0.00002)		(0.682)		(0.646)
Twitter Asian(s)($t-1$)				$-$0.00001				$-$0.266
				(0.00002)				(0.693)
Observations	11,116	11,116	10,921	10,921	4493	4493	4300	4300
R-squared	0.220	0.220	0.220	0.230	0.996	0.996	0.997	0.997
Outcome mean	0.057	0.057	0.057	0.057	41065.784	41065.784	41065.784	41065.784

Notes: The table presents the relationship between the racially charged Google search index and the Twitter post index, anti-Asian hate crimes, and visits to Chinese restaurants. Hate crime data are from the FBI UCR, visit data are from Safegraph, and all data are at the media market$\times$year-month level. Outcome variables are the monthly number of anti-Asian hate crimes between January 2014 and December 2018 (columns (1)-(4)) and the monthly number of visits to Chinese restaurants between January 2018 and December 2019 (columns (5)-(8)). Google Asian(s) is the Google search index for the word “Asian(s).” Twitter Asian(s) is the number of tweets including “Asian(s)” per 100,000 “the” tweets. All regressions control for local unemployment rate, year-month fixed effects, and media market fixed effects. *** $p<$0.01, ** $p<$0.05, * $p<$0.1.