Suppose we wish to know if long prison sentences reduce the probability of recidivism. The basic model is Repeat crime_i=B₀+B₁SentenceLength_i+e_i Where repeat crime is 1 for released prisoners who end up being sent back to jail after release. Which of the following is viable as an instrumental variable? A) A variable indicating the severity of the initial crime. B) A variable indicating that the person had been randomly assigned to a judge with a record of harsh sentencing. C) A variable indicating that the person had been randomly assigned to an educational program while in prison. D) A variable indicating whether the person was a high school dropout.

Question

Quizplus · Accepted Answer

Option B, a variable indicating that the person had been randomly assigned to a judge with a record of harsh sentencing, is a good instrumental variable because it is plausibly correlated with the probability of receiving a long prison sentence, but is not directly correlated with the probability of committing future crimes. The severity of the initial crime and high school graduation status might be correlated with the likelihood of recidivism as well as with sentence length, so they would not be good instrumental variables. The educational program might also be directly related to future criminal behavior and cannot be used as an instrumental variable.

Suppose We Wish to Know If Long Prison Sentences Reduce