The Police Department (PD) would like to know whether forgiving a first-time offender for a drug related crime might actually reduce the likelihood they commit crime in the future. Consider the crime of possession of a small quantity of marijuana committed by a minor (an individuals aged less than 18 years of age). Assume that the law states that if a minor is caught with 5 grams of marijuana for the first time, they only get a warning. If the quantity is higher than 5 grams, the minor gets charged and convicted of a serious crime. Adults (age 18 or more) get arrested and convicted regardless of the quantity of marijuana they are caught carrying.
You are hired as a consultant to help the PD to answer the question whether forgiving a first- time offender for a drug related crime might actually reduce the likelihood they commit crime in the future. The PD holds administrative records of the universe of warnings, offenses, as well as the individuals’ gender, age, and address.
In answering questions below, note that general statements about econometric methods will yield no marks. Adapt and assume the necessary assumptions to this particular setting and use equations and or graphs to support your argument. If using graphs, make sure that the axes are labelled correctly.
(i) Discuss the endogeneity problem in studying the above problem
(ii) What econometric method would be suitable for you to use in this setting to answer the question of interest to the PD?
(iii) Describe in detail the method in (ii) applied to this setting. What are the identifying assumptions of the method in (ii)?
(iv) How can the identifying assumptions be tested? Which graphs would be useful in supporting your arguments?
