Beginning on March 16, 1974, the federal district judges of the Second Circuit, which takes in New York, Connecticut and Vermont, found an unusual assignment in their mail: presentence reports, five a week for six weeks. Forty-three active judges and seven senior judges took part. Over those weeks they handed down roughly as many sentences as they normally did in half a year, none of them real.
The first twenty reports were real files from probation offices in the circuit, with names, places and dates changed. Every judge got the same pages. A judge who had sentenced the actual defendant, or helped decide his case, didn’t get that file.
Case 1 was a union official over 40, convicted at trial on nine counts of extortionate credit transactions and related income tax violations. He had three prior convictions and one serious charge pending.
The median sentence came back at 10 years in prison and a $50,000 fine. The most severe was 20 years and a $65,000 fine. The most lenient was 3 years. At least six judges gave 15 years or more, and at least six others gave 5 years or less.
It wasn’t one odd file. In 16 of the 20 cases, the judges didn’t even agree on whether the defendant should go to prison at all. “It would appear that absence of consensus is the norm,” wrote Anthony Partridge and William Eldridge of the Federal Judicial Center, who ran the analysis.
The obvious suspects were a few harsh judges at one end and a few soft ones at the other. The report looked for them and mostly didn’t find them. Ranked across thirteen of the cases, only two judges came out consistently severe and one consistently lenient. Of 29 judges whose average rank sat near the middle, 26 had handed down one of the ten harshest sentences in some case and one of the ten mildest in another. “Consistency of relative position is thus very much the exception,” the authors wrote, and the disparity “cannot, on the whole, be explained by labeling the judges.” They found no evidence, either, that years on the federal bench moderated it.
The authors listed the study’s limits. These were paper defendants: no judge saw a face, and no sentence touched a real family. Parole might narrow the gap in time actually served; measured by parole eligibility, Case 1’s range would have been one year to 6 2/3 years. And a single file, they warned, said nothing about its whole class of crime. Still, they concluded the disparity was “a reasonably good approximation of what really happens in the courtrooms of the circuit.”
The judges had asked for the study themselves. Judge Marvin Frankel, who chaired the subcommittee that organized it, wrote in the foreword that it was meant as “this concrete picture of ourselves.”
Parole could shorten a prison term. It could do nothing for the threshold question. As the report put it, the offender sentenced to prison “does go to prison; the offender sentenced to probation or a fine does not.”