Conversation

When controlling for poverty, black people still have higher crime rates - but this is not really the case if you control for *single motherhood*; as in, poor black communities with lots of fathers don't have more crime. Cool - if this is true, this seems useful! But- 2/
Replying to
It’s really an improper analysis to ‘control for’ either of those types of variables and then interpret it as if those factors were made to go away. All of these variables potentially affect each other over time and you can’t ‘control for’ some of them with data like this