Blaise OSS

Table 1. Descriptive statistics, 2020–2024 panel.

Variable	Treated (9)	Control (65)	Total
Community districts	9	65	74
Months in window	75	75	75
CD-month observations	675	4875	5,550
Pre-treatment mean complaints/CD/month	51.1	1.6	—
Total Rodent complaints in window	—	—	224,889

Table 2. Four DiD estimators of the containerization-pilot effect.

Estimator	ATT	SE	p
TWFE	-10.27	1.78	<.001
CS	-4.87	2.01	0.015
SA	-12.85	2.81	<.001
BJS	-11.90	0.70	<.001

Table 3. Robustness checks.

Check	ATT	p	Reading
Placebo t₀ = 2022-07-01 (BJS)	+9.97	<.001	Non-significant at raw α; sign reversal vs. headline.
Log(1 + complaints) TWFE	+0.227 (+25.4%)	0.092	Same sign; smaller magnitude.
Post-COVID (2022-01 →) (BJS)	-6.70	<.001	Strengthens headline.
Manhattan-only controls (BJS)	+3.62	0.059	Same sign, wide CI.

Table 4. RDD bandwidth sensitivity (cutoff = 35.1).

Bandwidth	h	ATT	SE	p
h/2	6.59	+1.07	4.72	0.821
h	13.18	+1.64	3.41	0.631
2h	26.36	-1.51	2.91	0.604

Table 5. Identification + statistical diagnostic checklist.

Diagnostic	Value	p	Reading
Parallel-trends F	F = 4.40	<.001	Reject flat pre-trends.
Breusch-Pagan	see §4.2	<.001	Heteroskedastic; cluster-robust SE.
TWFE R²	0.802	—	Within-panel variance absorbed.
MDE (α=.05, power=.80)	~17.6 complaints (	d	~ 1.00)
BH survivors	9/13	BH 0.05	Main BJS + SA survive.