The Manhattan Bail Project randomly assigned some defendants
with strong community ties to be recommended for release without bail and some
not to be recommended. There was a very high (99%) rate of appearance at trial
for those released without bail, but of course, the researchers did not know
that this would be the case before they conducted the study. Would you consider
this study design ethical? What if one of the persons in the 1% who did not
come to trial murdered someone when he should have been at court? Are there any
conditions when randomization should not be permitted when evaluating criminal
justice programs?
Leave a Reply