Hi everyone,

I want to investigate the relationship between employment and crime on an individual and monthly level. I have data from around 1.5 million individuals for 96 time periods (8 years, 12 months per year), where I know whether they were employed or not, whether they committed an offence or not, monthly income and some other control variables.

The problem with this analysis is that there may be reciprocal effect (employment affects crime, but crime might also influence employment) and the fact that there are other variables/characteristics that affect crime as well as employment. Unfortunately I won't be able to do use an IV with the available data.

Does anyone have any suggestions on what would be a good method to use? On the internet I saw a random effects probit model to control for individual characteristics, and I also saw references to a tobit model or a bivariate probit selectivity model. Does anyone have experience with the kind of problems I am experiencing, and know what model would be best to use and how to implement this model in Stata (e.g. I'm not sure how I would implement a bivaraite probit selectivity model in Stata)?

Thanks a lot!