Exterro and Duke/EDRM Judges Survey 2019 Series: Part 2, Taking Affirmative Action to Address E-Discovery Problems

This is the second in a series of posts evaluating the results of Exterro and Duke/EDRM’s 2019 survey of Federal district court and magistrate judges. With information from over 250 judges, the survey data offers a rich trove to mine. Part 1 looked at responses to the first survey question, which was about failure to comply with Federal e-discovery rules. This article, Part 2, focuses on responses to the second substantive set of questions in the survey, which go to how often and why judges had to take affirmative action in cases to address e-discovery problems.

The question

Judges were asked, “In the past 12 months, how often have you had to take affirmative action (e.g., require additional conferences either in camera or unsupervised or issue a warning or sanction) in a case to address an e-discovery problem?” They were direct to choose one of five options: “one”, “two”, “three”, “four”, or “more than five”. The judges also were asked to comment, “on what basis” did they have to take affirmative action.

Three key takeaways

1. Most responding judges acted affirmatively at least once in the past 12 months. Only 15% said they had not needed to take any action. Of district judges, 78% said they acted affirmatively at least once. Magistrate judges were notably higher at 95%.
2. Although responding judges acted affirmatively, they did not do so often. Seventy-nine percent of responding judges took affirmative action just 4 times or less in past 12 months. For district judges, the number was 88%; for magistrate judges, 67%.
3. Judges also took affirmative action only in a tiny percentage of the cases they handled. (Here I need to add the important caveat that the following percentages are artificially high; more about that below.) For district court judges, the numbers range between 0.3% and 1.5%. For magistrate judges, the numbers range between 0.2% and 0.8%.

Placed in context, these numbers suggest that during the past 12 months responding judges (and Federal trial court judges generally) rarely felt the need to intervene to address e-discovery problems.

The responses

Adjustments

If judges had been given the option to answer “zero” or, in the case of district judges, “magistrates handle that”, the judges’ comments clearly indicate that a significant percentage of the judges would have selected one of those options. For the analyses in this article, I have adjusted the responses accordingly and am using “magistrate only”, “zero”, “one”, “two”, “three”, “four”, or “more than five”.

A closer look at responding judges

Just over 250 judges responded to this survey – 139 district court judges, 110 magistrate judges, and two judges who did not indicate their roles. The Administrative Office of the U.S. Courts reports that for calendar year 2018 there were 663 authorized district court judgeships. According to a 2016 paper prepared for the Federal Bar Association, at that time there were 531 full-time magistrate judges. Assuming the AO and FBA numbers to be accurate today, we got responses from approximately 21% of the bench.

Role	Number Responding	% Responding	Total Reported	% Reported	Responding as % of Reported
District Judge	139	56%	663	56%	21%
Magistrate Judge	110	44%	531	44%	21%
Total	249	100%	1,194	100%	21%

Most judges acted affirmatively

Ninety two percent of responding judges answered this set of questions. Of those judges, 15% said that in the past year they had not had to take any affirmative action to address e-discovery problems. Compare answers from magistrate and district court judges, and you see the expected difference between the two groups. Ninety-five percent of responding magistrate judges said they had taken affirmative action at least once in the past 12 months. For district court judges the number was lower, at 78%. Two district court judges specially noted that they passed these issues along to magistrate judges.

Took Affirmative Action	District Judge	Magistrate Judge	All
Never	29	5	34
At least once	100	97	198
Total	129	102	232

Overall, judges did not have to act affirmatively often. As the chart and tables below show, 79% of the judges took affirmative action less than five times in the past 12 months. For district court judges the number was 88%; for magistrate judges, it was 67%. Only 21% of responding judges took affirmative action five times or more. District judges were less at 12% and magistrate judges notably higher at 33%.

Took Action – All	Percentage	Cumulative Percentage
Magistrates handle	1%	1%
0 times	14%	15%
1 times	21%	35%
2 times	24%	59%
3 times	13%	73%
4 times	6%	79%
5+ times	21%	100%
Total	100%

Took Action – District Court	Percentage	Cumulative Percentage
Magistrates handle	2%	2%
0 times	21%	22%
1 times	22%	44%
2 times	22%	67%
3 times	16%	82%
4 times	6%	88%
5+ times	12%	100%
Total	100%

Took Action – Magistrate	Percentage	Cumulative Percentage
Magistrates handle	N/A	N/A
0 times	5%	5%
1 times	20%	25%
2 times	25%	50%
3 times	11%	61%
4 times	6%	67%
5+ times	33%	100%
Total	100%

But most judges acted affirmatively only rarely

Judges took affirmative action only in a tiny percentage of the cases they handled, at most just over 1.5% and most likely almost always far less than that. The first step toward arriving at this conclusion is to calculate the average number of cases handled each year by district court and magistrate court judges. As explained below, it appears those numbers are somewhere in the range of 336 matters per year for district court judges and 656 for magistrate judges. According to Federal Judicial Caseload Statistics 2018 from the Administrative Office, in the 12-month period ending March 31, 2018 there were 358,563 civil and criminal matters filed in U.S. district courts. Of those, 277,010 were civil matters. Subtracting prisoner petitions, there remain 233,045 matters. The Administrative Office also reports that for calendar year 2018 there are 663 authorized district court judgeships. Combined those numbers, and it appears that on average district court judges handle approximately 336 matters a year. The Administrative Office tells us that in 2018 magistrate judges handled 348,421 civil matters. According to a 2016 paper prepared for the Federal Bar Association, at the time there were 531 full-time magistrate judges. Assuming the number of magistrate judges has held steady, then on average magistrate judges handled approximately 656 matters last year.

	District Judges	Magistrate Judges
Number of matters handled	223,045	348,421
Number of judges	663	531
Average number of matters per judge	336	656

The second step is to calculate frequency of affirmative actions, where that frequency is defined as the number of times in the year a judge took affirmative action divided by the number of matters a judge handled. Here is where the above-mentioned caveat comes into play. Because of the specific wording of our question (“how often have you had to take affirmative action … in a case”), we do not know from the response whether judges taking affirmative action more than once did so in the same case or in two separate cases. For this exercise, however, I need to assume that each time a judge took an affirmative action, the judge did so in a different matter. I know this assumption is wrong, as the judges’ indicate otherwise, but I have no better data to work with. This means that the numbers below overstate, most likely substantially, the percentage of cases on judges’ dockets where the judges had to take affirmative action. Nonetheless, here are the calculations.

Number of Matters with Affirmative Action Taken as % of Docket
No. of Matters with Affirm Action	District Judge	Magistrate Judge
1 matter	0.3%	0.2%
2 matters	0.6%	0.3%
3 matters	0.9%	0.5%
4 matters	1.2%	0.6%
5+ matters	1.5%	0.8%

The implication is clear. Matters in which judges find themselves having to take affirmative action only account for a tiny part of the judges’ overall dockets. For district court judges, the numbers range between 0.3% and 1.5% according the above calculations and should be noticeably lower. For magistrate judges, the numbers range between 0.2% and 0.8% with the same cautionary comment.

Comments

In addition to being asked how often they took affirmative action in cases to address e-discovery problems, judges were asked to comment as to the bases for which they took that action. DUKE/EDRM organized the comments into seven groups:

• On the basis of misconduct, neglect, or rule violation
• On the basis of a dispute
• On the basis of communication issues
• To provide guidance
• Regular practice
• On the basis of privilege issues
• Basis unspecified (a catch-all group we won’t go into here)

As in Part 1, I won’t fill out this post with the full set of comments; those are available on the DUKE/EDRM site. Following, instead, is my distillation of the comments.

On the basis of misconduct, neglect, or rule violation

Most of this group of comments focused on attorneys’ failure to do something. These included failures to: confer with opponents meaningfully or even at all; be specific in responses and objections; understand their client’s data and e-discovery practices; search for data appropriately or diligently; search the data itself appropriately or diligently; deliver complete productions; act timely or act at all; comply with court orders; and understand e-discovery itself or turn to someone who does for assistance. Many of the comments centered on attorneys’ abuse of the e-discovery process, such as: making overbroad requests; using boilerplate or improper objections or responses; engineering inappropriate delay; or refusing to engage in genuine efforts to confer or to resolve disputes before turning to the court.

On the basis of a dispute

Perhaps the most fundamental dispute leading to judges taking affirmative action to address e-discovery problems arose when each side thought it was reasonable and the other side was wrong and neither side seemed willing to re-assess those positions. Many disputes leading to judges taking affirmative action stemmed from counsels’ inability to agree. Areas where they could not (or preferred not to) come to agreement included: scope and limits of discovery to be engaged in; initial e-discovery actions to be taken; search terms; deadlines; and scope of ESI to be produced. Other disputes came from lawyers taking unilateral action. One-sided actions include requesting parties asking for too much before trying to determine what actually might be reasonable, as well as responding parties producing only what they wanted to without conveying how they had limited their processes. Some disputes appeared to be born of ignorance, such as where attorneys over-promised because they did not understand what they are committing themselves or their clients to do. Finally, some disputes seemed to be the result of an apparent desire to have issues remain unresolved: parties refuse to engage in meaningful communications; stonewall on productions without offering reasonable alternatives; and demand data for data’s sake rather than to advance the theory of the case.

On the basis of communication issues

“What we’ve got here is a failure to communicate.” Too often attorneys didn’t meet, didn’t talk, didn’t write, didn’t plan. And when they did, too regularly they seemed to be the proverbial ships passing in the night.

To provide guidance

Some judges felt they had to act affirmatively to so as to give guidance, especially to inexperienced counsel, for example directing parties to refine and identify their e-discovery.

Regular practice

Some judges see themselves as pursuing a regular practice of acting affirmatively, where in their local rules they call for conferences, in more complex cases require counsel to meet and confer with IT people participating, and sometimes just bring counsel into the courthouse so the attorneys can hammer things out on a face-to-face basis.

On the basis of privilege issues

Privilege issues can lead to judges needing to take affirmative action. Part of this appears to be inherent in the difficulties and complexities posed by privilege concerns. Part arises from overly aggressive assertion of privilege.

About the survey

In late 2018, Duke/EDRM and Exterro conducted a survey of Federal judges, the fifth year for Exterro and the second for Duke/EDRM. The survey contained over 20 groups of questions. More than 250 Federal district court and magistrate judges responded to the survey. Exterro and Duke/EDRM published survey results in early 2019. Exterro’s report on the results is available here. Duke/EDRM’s compilation of results, including the text of comments from judges responding to specific questions, is available here. The astute observer will notice that tallies used in this and subsequent posts differ slightly from those appearing in the Duke/EDRM and Exterro materials. Generally those differences come from efforts on my part to reconcile answers to choose-an-option questions with narrative comments provided by the judges.

Survey demographics

Roles

Respondents are pretty evenly divided between district court and magistrate judges. Of the 251 Federal judges responding to the survey, 55% identified themselves as district court judges and 44% said they are magistrate judges. Of the district judges, 4% reported that they are chief judges, 6% are retired, and less than 1% are on senior status. Just over 1% of the magistrate judges noted that they are chief magistrate judges and less than 1% said they are retired.

Years on the bench

The responding judges are a seasoned group. Overall, 63% of the judges have been on the bench for at least 11 years; 83% for six years or more. District court judges have more experience – their largest cohort, at 46%, has been on the job for 21 years or more – while responding magistrate judges have not served as long – their biggest group, at 33%, clocks in at between 6 and 10 years. These differences are not surprising as district court judges have lifetime tenures while magistrate judges are appointed for renewable eight-year terms.