Abstract
This pre-analysis plan specifies the main empirical analyses in a study of the effect of winning the lottery on subsequent criminal behavior of winners and their children. We first discuss previous theoretical literature, the Swedish criminal justice system, our data on criminal convictions, and provide descriptive statistics for crime in Sweden. We then present our sample of lottery players and how they compare to the population in terms of criminal behavior. Next, we use Monte Carlo simulations to evaluate analytical standard errors and the statistical power of different sample restrictions and outcome variable definitions. Finally, we pre-specify the analyses that will be reported in the paper.