Extended Data Table 5 Crime outcome treatment-on-the-treated estimates

From: Knowledge about others reduces one’s own sense of anonymity

  1. Specifications show the percent change in crimes per resident attributed to the intervention for four TOT analyses. For the “Received Mailer” analysis, treated corresponds to receiving NCO mailers (four randomly assigned treatment developments did not receive mailers). For the “Received Outreach” analysis, treated corresponds to receiving NCO outreach (11 randomly assigned treatment developments did not receive NCO outreach, while two randomly assigned control developments did receive NCO outreach). For the “Received Mailer AND Outreach” analysis, treated corresponds to receiving NCO mailers and NCO outreach. For the “Received Mailer OR Outreach” analysis, treated corresponds to receiving either NCO mailers, NCO outreach, or both. TOT analyses use two-stage least squares. P-values are from two-tailed tests.
Back to article page