.@CrimAdvStats Using propensity score matching as one of their statistical procedures, Blomberg et al. (2011) found that higher educational achievement is an important predictor of recidivism.
sciencedirect.com/science/ar…
.@CrimAdvStats Worrall and Kovandzic (2010) use a panel dataset and employ the instrumental variables approach to find that police presence was inversely related to four crime index rates. sciencedirect.com/science/ar…
.@CrimAdvStats Hao and Cowan (2020) use the difference-in-differences design to examine the spillover effects of marijuana legalization in counties bordering non-legal states.
onlinelibrary.wiley.com/doi/…
.@CrimAdvStats Mitchell et al. (2016) used a regression discontinuity design in their study and found that imprisonment had little to no effect on recidivism. Alternatives to imprisonment might be more useful for non-violent/drug offenders.
tandfonline.com/doi/full/10.…
