It is true that there are CVI successes. But the literature is far less conclusive than proponents like to claim. In addition to the positive findings, there are many null or mixed results—studies where CVI has no effect on violence, or where it produced beneficial outcomes only in some places but not others, or on some metrics but not others. Some of the most supportive studies rely on weak control groups, comparing treated areas to the rest of a jurisdiction or matching treatment and control neighborhoods without using designs to address unmeasured variables that could drive the differences in outcomes. Reliance on weaker studies when stronger ones show smaller effects is usually a red flag.

eight years after release, men are 43% more likely to be taken back under arrest than women; African-Americans are 42% more likely than whites, and high-school dropouts are three times more likely to be rearrested than college graduates.