« Fair Sentencing Act about to alter crack sentencing ... with lots of transition issues to follow | Main | "Rep. Mike Rogers: Execute WikiLeaks leaker" »

August 3, 2010

An original response to "A Brief and Modest Proposal" from another US District Judge

Regular readers may recall this engaging post from last week concerning providing federal sentencing data for particular judges, titled "'A Brief and Modest Proposal' ... an original essay from US District Judge Richard Kopf." I am pleased now to be able to provide an original reponse to this original essay from another federal district judge:

Dear Professor Berman:

I write to share my views regarding a proposal to publish each year statistics which would disclose the percentage of times individual judges did, or did not, impose a sentence within the guideline range. Upon first blush, the idea may seem reasonable. Upon inspection, there are many reasons to oppose the proposal. At a minimum, such data should be understood as far more complex than what has been represented.

I note that it is hard to argue against “transparency.” I also acknowledge that most of the information involved is public record, albeit contained in individual cases. The concerns I offer are not that information on cases should be sealed or that federal judges should not be subject to criticism regarding sentencings. It is the aggregation of sentencing data that I fear will be misleading, subject to distortion, and of limited analytical value.

I give several reasons for this view. First, the proponent advocates making only a small portion of data available - information relating only to a judge’s adherence to or variance from the sentencing guidelines. In the age of electronic filings, much more information, by judge, is available. For example, with little effort, the Administrative Office of the Courts could compute a district judge’s reversal rate per year or the number of times an employer or an employee won or lost a case before a single judge. Singling out only guideline related data for disclosure is not a call for transparency. There can be much more disclosed, if that is the point.

Instead, the proposal seeks selective disclosure, by judge, with the knowledge that such information would be compiled each year. In civil cases, a statistical report by judge is issued twice a year, but only as to timeliness in the handling of cases and motions, with no reference to particular outcomes. If this proposal was adopted, for the first time, decisions made by a district judge would be statistically analyzed with regard to a potentially divisive issue. I can think of no other example in which decisions of specific district judges have been disclosed in such a manner.

Second, district judges do not select cases. When dealing with relatively small number of cases, anomalies can often occur. If I told you that a domestic relations judge gave custody or parenting rights to the mother in 95% of the cases last year, from that number alone, could you really draw any conclusions? If you looked at each case, maybe the fathers had played no parental role or other factors were at play making the decision to award parenting rights to the mother the absolute right thing to do. The statistics would give the impression of predisposition and bias. Sentencing should be focused on a unique individual and a unique crime. Defendants and crimes have infinite variations. Aggregating sentences into statistics de-emphasizes the individuality of the process.

Third, because we are dealing with relatively small numbers, the significance of the percentage of cases with departures or variances will be doubtful with most judges. Further, I have no doubt that the annual percentages for a particular judge will vary widely. Some simple math will illustrate this point. In our district, each judge sentenced roughly 60 defendants in 2009. 36.7% of these defendants received the benefit of a government motion made under U.S.S.C. § 5K1.1 or 18 U.S.C. § 3553(e) asking the Judge to impose a sentence below the guideline range. Another 15% of the defendants had a guideline range authorizing probation. The combination of these two groups is almost 50% of the cases. The remaining 28 defendants would be the only relevant basis for analyzing a judge’s adherence or deviation from the guidelines. The number is quite small. The type of offense, the backgrounds of defendants, the harm to the public can be highly variable from case to case and most certainly year to year.

To publish annual aggregate statistics relevant only as to whether a judge routinely imposes guideline sentences is an implied criticism of judicial decision making and a thinly veiled attempt to influence future sentencings. Better arguments, less intrusive upon judicial independence, are to simply leave matters as they stand (records are public but not aggregated) or open the electronic filing system to aggregation of any information requested. This way, we can use sunshine, and not a menacing searchlight, to aggregate all aspects of decision making.


United States District Judge

Related post:

August 3, 2010 at 11:27 AM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference An original response to "A Brief and Modest Proposal" from another US District Judge:


Left to their own discretion, judges caused a massive increase in the crime rate in the 1980's and 1990's, with massive increases in murder rates. Judges are not above scrutiny. Indeed most are lawyers, and know nothing about the subject matters before them, including criminality. They are all amateurs plunged into very difficult managerial positions. None is a professional judge, as they are in Europe. Despite their lifetime appointments, the dangerous or incompetent judge can be pressured to resign, and should be. Judges should not come to feel they have an inherited nor permanent title.

Outcomes of their sentencing decisions should be included in their track records. If a judge awards custody to mothers 95% of the time, instead of the statistically more neutral 50%, a rebuttable presumption of bias is entirely appropriate. Let the judge defend the record or be assigned to a desk in the basement boiler room by his supervisor.

This is for the good of the judges as well as in the interest of the public. Liability is a substitute for violence. Absolute immunity, lack of accountability logically justify a recourse to violence. Judges are widely hated and held in public contempt. Making them explain their decision will teach the public, matters are more complicated and difficult than implied in the newspaper.

I seconded Judge Kopf's proposal and extended it. "I would add these correlates to be followed by the judges themselves, 1) the original charges as a more accurate reflection of the crime, rather than the plea bargained final charged crime; 2) follow up by telephone of the progress of the defendant on the path of maturation, perhaps via a confidential, immunized interview, a list of the crimes committed subsequent to the sentencing; 3) the overall criminal record of the defendant, as a better predictor of outcomes."

Posted by: Supremacy Claus | Aug 3, 2010 12:26:22 PM

Brilliant, thoughtful and true.

Posted by: lawyer | Aug 3, 2010 1:20:15 PM

Dear Professor Berman,

I have read Judge Sargus’ thoughtful response to my “Brief and Modest Proposal.” I do not wish to debate the judge about the value of transparency–we simply don’t see things the same way. However, because I think he misunderstands a key point, I wish to correct his apparent misunderstanding.

I am not suggesting anything that is radical or new. I am only suggesting that the same statistics that are annually produced and made public by the Sentencing Commission be expanded slightly. Right now, the Commission produces a statistical table for “Sentences Relative to the Guideline Range by Circuit and District” (available at http://ussc.gov (click on “Publications” and “Statistics”). That statistical table has the following categories: (1) “Sentenced within Guideline Range”; (2) “Above Range Departure”; (3) “Above Range Departure w/Booker”; (4) “Above Range w/Booker”; (5) “Remaining Above Range”; (6) “§ 5K1.1 Substantial Assistance”; (7) “§5K3.1 Early Disposition”; (8) “Other Gov’t Sponsored”; (9) “Below Range Departure”; (10) “Below Range Departure w/Booker”; (11) “Below Range w/Booker”; and (12) “Remaining Below Range.”

Realizing that the data is now available for each district, expanding the date set to include each judge who sentenced a substantial number of individuals would be easy to do and, more importantly, it would be helpful to scholars and researchers. Using Pacer, academics could use the judge-specific data produced by the Commission to isolate and study the sentencing decisions of particular judges and particular groups of judges. In other words, the Commission’s judge-specific data would give researchers a starting point; that is, an ability to sort out what judge or what group of judges warranted a more in-depth inquiry. I see that as very useful.

As your readers may remember, I was moved to make this suggestion after reading Professor Masur’s article. As you pointed out, Professor Masur argued that in many “cases the judges who diverge from the advisory Guidelines ranges will do so for the wrong reason.” I was very interested to know how one might empirically test that argument. Given the large number of federal sentencing judges, it seemed to me that one could not easily test that assertion without the data I now propose be made public.

Thank you very much for providing this forum. It is very helpful.

Richard G. Kopf
United States District Judge

Posted by: Richard Kopf | Aug 3, 2010 1:22:14 PM

The difference between J. Kopf and J. Sargus is that J. Kopf sees breaking down the stats by individual judge as just a "slight expansion" of existing practice, rather than as something new, and J. Sargus sees it as something indeed quite new. I think J. Sargus' view is the more convincing. J. Kopf's response would have been more persuasive had he been willing to engage J. Sargus' suggestion that, if we're going to publish stats for individual judges in one type of case (criminal sentencings), we should do do so for other types of cases (e.g., custody disputes, etc.).

Posted by: lawyer | Aug 3, 2010 1:43:12 PM

This is an interesting debate, but I can tell you, speaking as someone who has worked for the US Sentencing Comm'n, that the Comm'n has a hard enough time collecting sentencing data (and collecting it in a timely fashion) from judges as it is. At least when I worked there some years ago, judges submitted the data voluntarily, without identifiers, pursuant to a Memorandum of Understanding between the Sentencing Comm'n and the Judicial Conference. I'm not sure even Congress could order the judges to provide more detailed data, as there are separation of powers issues lurking here. And the data the Comm'n collects is incomplete. They only see presentence reports and judgments of conviction, which often do not accurately reflect what happened at sentencing. To really know what happened at sentencing and why it happened, the Comm'n would have to see sentencing transcripts, which are very expensive and are typically only ordered (and paid for, typically, by defense counsel) in cases that are appealed. For these practical reasons and for the thoughtful reasons submitted by Judge Sargus, I think Judge Kopf's proposal is a non-starter.

Posted by: attorney | Aug 3, 2010 2:04:05 PM

The problem with Judge Kopf's idea, and Judge Sargus alluded to it, is that it treats sentencing as if it involves nothing more than a Guideline calculation and a court's decision to follow it or not. In reality, there are (or at least should be) so many other factors involved.

Kopf would like to reduce a sentencing judge's decisions to a statistic. Unfortunately, sentencing cannot be -- and should not be -- so simplistic.

Posted by: DEJ | Aug 3, 2010 2:18:14 PM

You could get cooperation if the judges' paychecks were exchanged for the data requested, every pay period.

Posted by: Supremacy Claus | Aug 3, 2010 2:26:41 PM

Dear Professor Berman,

With reference to the person who submitted a comment under the name “attorney” and who represented that he worked for the Sentencing Commission “some years ago,” I respectfully suggest that several of assertions in that post are inaccurate. In particular,

1. Judge-specific data is and always has been collected by the Sentencing Commission as evidenced by the fact that the Commission uses that data to compute the statistics for each circuit and district that are published annually. Indeed, in the past I have requested my own statistics and the Commission provided them to me.

2. The Sentencing Commission receives a “Statement of Reasons” for each sentence and that statement describes not only the basis for departures but also the basis for Booker-type variances. See, e.g., Form AO245B

3. The submission rates for SORs and other documents is very high. For example, during the most recent time frame (2nd quarter of 2010), the submission rate for SORs was 98 percent. See Rate of Submission of Statement of Reasons Form in Each Circuit and District, 2nd Quarter 2010 Preliminary Cumulative Data (Table 28) at page 51, available at http://www.ussc.gov (click on Second Quarter FY10 Quarterly Sentencing Update).

As for the other comments by “attorney,” I am not suggesting that Congress or the Executive branch order the release of this information. I am suggesting that the Sentencing Commission, an agency of the judicial branch, independently decide to release the data. In fact, I hope that individual judges would be supportive. Hence, there are no “separation of powers issues lurking”in my proposal.

Furthermore, and not to beat the hide off a dead horse, I am not suggesting that judge-specific statistics tell the whole story. I am suggesting that judge-specific statistics are a starting point. That is why my proposal was, shall I say, a "modest" one.

Richard G. Kopf
United States District Judge

Posted by: Richard Kopf | Aug 3, 2010 5:46:12 PM

Every one of Judge Sargus's concerns could be explained in a prominent preface to the table setting forth the data.

Federal judges have lifetime tenure and salary. That secures their independence. They do not have -- no federal official wielding that amount of power has or should have -- immunity from scrutiny or criticism.

A judge should welcome questions, including hostile questions, about his decisions, including sentencing decisions. If he can explain them, the education thus provided will be a service to the public. If he can't, maybe he should have thought more carefully to begin with.

If a judge is consistently above or below the guidelines to a degree substantially different from his colleagues, the public has a right to know this and ask why. There may be perfectly valid reasons (for example, having to do with the particular district in which that judge sits). But there may also be questionalbe reasons (for example, that the judge is sentencing based on a pre-existing bias flowing from his former work as a prosecutor or defender). Either way, we need a judge-by-judge breakdown, as Judge Kopf suggests.

The Founders valued, and created, an independent judiciary. They did not create an imperial one. We pay the freight and we live with the results. We have a right to know.

Posted by: Bill Otis | Aug 4, 2010 9:17:40 AM

I wonder if Judge Kopf recalls what happened to Judge Rosenbaum back in 2002? He was the victim of a Congressional witch hunt as a result of his individual sentencing statistics. Yes, it would be nice if we lived in a rational country where judges could simply explain their decisions and everyone would understand and feel good about it. But the truth of the matter is we have an absolutely insane political culture. All it takes is one Congressman with an agenda to scare judges away from guidelines variances less they become the next Judge Rosenbaum. See Feeney Amendment.

With respect to Judge Kopf's criticism of my previous post, as I said I worked for the Comm'n years ago. The information gathering changed after Booker. The Statement of Reasons policy was implemented after Booker. So I concede that his knowledge of the Comm'n's information gathering practices is more accurate than mine at this time. However, I do practice law in federal court everyday (not in Judge Kopf's district), and I can tell you that the Statement of Reasons forms are prepared by courtroom deputies here, not by district judges. They do not always accurately reflect what happened in court.

Posted by: attorney | Aug 4, 2010 10:21:01 AM

Judge Sargus concedes the folly of arguing against transparency but proceeds to do just that.

The public has a right to decide whether the data is of "limited analytical value."

Posted by: mjs | Aug 4, 2010 10:28:20 AM

In related news, Threats Against Judge on Arizona Immigration Case Fit Larger Pattern.

The threats against Bolton fit into a wider nationwide pattern of increased threats against judges. Improper communications and threats to federal prosecutors and federal judges more than doubled during the mid-2000’s, according to an Office of the Inspector General report released earlier this year.

Posted by: George | Aug 4, 2010 11:55:20 AM

maybe if they stopped acting like the judges in nazi germany in the 1930-40's and whitewashing anything crooked the govt and law enforcment is doing...people would stop wanting to remove them any way they can.

Posted by: rodsmith | Aug 4, 2010 4:22:03 PM

If liability is a substitute for violence, then immunity is a full moral and intellectual justification for violence. I would prefer it if a strong executive sent federal marshals, arrested the judge, tried her for treason and insurrection against the Constitution, for an hour, with her wearing a giant dunce cap, with the sole evidence being her decision, and then hanged her in the court basement. Judicial review violates Article I, Section 1, and is not permitted by any other clause.

Posted by: Supremacy Claus | Aug 4, 2010 7:48:11 PM

By nature, the judicial branch is anti-majoritarian. Judges stand in the way of the State and as the last line of defense for the individual against the over-arching depredations of the State. Such actions are often not popular with the mob. I am not surprised therefore, that our torture cheerleader would enthusiastically support supplying more fodder for the mob, in a transparent attempt to pander to the basest human emotions, rather then recognize that justice is supposed to be much more than mere vengeance and retribution. Is there no shame in tirelessly hyperventilating over the prospect of putting yet another human in a cage for yet a longer period of time?

Posted by: Mark # 1 | Aug 5, 2010 1:59:40 AM

As a criminal defense practitioner, I respectfully disagree with Judge Kopf's proposal. I am very much persuaded by Judge Sargus's point that aggregating this type of statistic may be used as an attempt to influence future decisions and undermine judicial independence.

I can also add, as an example, that aggegating this statistic does not take into account, for example, the nature of the Guideline at issue in a particular case. For example, child pornography guideline has recently come int for just criticism because it aggravates the sentence based on factors inherent just in every instance of the crime and, thus, makes a vast majority of the offenders eligible for the statutory 10-year maximum. For that reason, many district court judges have recently exercised their Booker discretion to impose below the guidelines sentences. It would be manifestly unfair to read anything negative in the judge's decision to impose a below-the-Guideline sentence in those situations, particularly since even appellate courts recognize the problems with that particular guideline.

Posted by: Gene | Aug 5, 2010 4:15:29 PM

Omgeee i just noticed i dont have my "Mom" ring on i hope i didnt lose it the other night when i was drunk.

Posted by: charms thomas sabo | Dec 16, 2010 3:02:09 AM

Post a comment

In the body of your email, please indicate if you are a professor, student, prosecutor, defense attorney, etc. so I can gain a sense of who is reading my blog. Thank you, DAB