Microsoft word - reagan author approved edits 12-7-05 _h__p_.doc

REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM THE HUNT FOR SEALED SETTLEMENT AGREEMENTS There is substantial concern over courts sealing settlement agree- ments.1 This article describes how the Federal Judicial Center conducted some research on the matter as it pertains to federal civil cases at the re-quest of a United States senator and the Civil Rules Advisory Committee.2 We found that the sealing of settlement agreements in federal courts is rare, and that typically the only part of the court record kept secret by the sealing of a settlement agreement is the amount of settlement. The following is the story of the secret injury, according to conven- tional wisdom. Big Bad Corporation manufactured Product X. Helpless Citizen used Product X and was severely injured. So Helpless Citizen sued Big Bad Corporation, which settled the suit before trial for $1 zillion. Be-cause of a sealed settlement agreement, the public does not know that Help-less Citizen was hurt by Product X and that Big Bad Corporation was willing to compensate her for her injury. Neither does the public know that scads and scads of other Product X users were also injured and secretly paid off. The following is a typical example of how the news media have cov- ered the secret injury. CBS’s popular television program 60 Minutes II presented a story in 2003 reporting that the federal District of South Caro-lina addressed the secret injury problem by banning sealed settlement agreements.3 A teacher named Annie Davis bought a prescription drug from CVS.4 The drug made her so sick that she needed to be hospitalized, ∗ Senior research associate, Federal Judicial Center. A.B. Stanford University 1980 (Psychol- ogy, Human Biology); Ph.D. Harvard University 1986 (Psychology); J.D. University of California, Hastings College of the Law 1993. The views expressed herein are those of the author and not necessar-ily those of the Federal Judicial Center. 1. See Laurie Kratky Doré, Settlement, Secrecy, and Judicial Discretion: South Carolina’s New Rules Governing the Sealing of Settlements, 55 S.C. L. REV. 791, 792–94 (2004). 2. For more information on research projects by the Federal Judicial Center for the Civil Rules Advisory Committee, see Thomas E. Willging, Past and Potential Uses of Empirical Research in Civil Rulemaking, 77 NOTRE DAME L. REV. 1121 (2002). 3. 60 Minutes II: Hush Money (CBS television broadcast Jan. 15, 2003). REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM and although the initial illness subsided, lingering symptoms remained.5 Ms. Davis sued, and CVS agreed to settle the case.6 But the settlement is secret, so Ms. Davis is prohibited from disclosing its terms.7 Although this story bears some similarities to the Product X story, there are important differences. Ms. Davis’s case is more of a slip-and-fall case than a mass-tort case. CVS did not manufacture the drug she took; one of its pharmacists filled her prescription with the wrong drug.8 She was prescribed a hormone drug called Halotestin9 as part of her treatment fol-lowing removal of a cyst from her breast.10 Instead the pharmacist gave her Haldol, a powerful anti-psychotic drug used to treat schizophrenia,11 which induced seizures in Ms. Davis.12 These facts alone suggest nothing more than a one-time accident, al- though it is possible that this pharmacist routinely filled prescriptions in-correctly, or that procedures at this particular CVS pharmacy inadequately prevented errors, or even that CVS had more general corporate problems making improper filling of prescriptions at its pharmacies too likely. Of course, these possibilities assume that the error was really the pharmacist’s and not the prescription-writing physician’s. CBS reported that Ms. Davis’s attorney searched for other lawsuits against CVS for improperly filled prescriptions and found one.13 This pre-vious suit also settled with a secret settlement agreement.14 A striking fact in Ms. Davis’s story is that although her lawsuit was settled with a secret settlement agreement we know a lot about the case. We know who the plaintiff is, who the defendant is, what the defendant was alleged to have done, and how the plaintiff was injured. We do not 9. Halotestin is a brand name for fluoxymesterone, see WebMD Health—Patient Handout for Halotestin Oral, http://my.webmd.com/drugs/drug-3498-Halotestin+Oral.aspx?drugid=3498 &drugname=Halotestin+Oral#brands (last visited Nov. 10, 2005), an androgenic hormone, see RxList, Halotestin Online, http://www.rxlist.com/cgi/generic2/fluoxy.htm (last visited Nov. 10, 2005), used to treat delayed puberty in males and breast cancer in females, see RxList, Halotestin Indications, http://www.rxlist.com/cgi/generic2/fluoxy_ids.htm (last visited Nov. 10, 2005). 10. See 60 Minutes II, supra note 3. 11. Haldol is a brand name for haloperidol. See WebMD Health—Patient Handout for Haldol Oral, http://my.webmd.com/drugs/drug-5419-Haldol+Oral.aspx?drugid=5419&drugname=Haldol+Oral (last visited Nov. 10, 2005). 12. 60 Minutes II, supra note 3. 13. Id. CVS pharmacies fill many prescriptions and the existence of only one previous lawsuit is actually rather encouraging. I do not know, however, what methods the attorney used to search for cases, so I cannot evaluate how effective his search was likely to be. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM THE HUNT FOR SEALED SETTLEMENT AGREEMENTS 441 know only two things. We do not know how much CVS paid Ms. Davis and her attorney, and we do not know what Ms. Davis and her attorney learned in discovery.15 For its 60 Minutes II program, CBS interviewed Judge Joe Anderson, the chief judge of the federal District of South Carolina.16 Judge Ander-son’s district had recently amended its local rules to proscribe the sealing of settlement agreements.17 In 2001, the court codified standards and pro-cedures for “filing documents under seal” in a new Local Rule 5.03.18 In 2002 the court adopted an amendment to Rule 5.03: “No settlement agree-ment filed with the Court shall be sealed pursuant to the terms of this Rule.”19 United States Senator Herb Kohl, from Wisconsin, has been proposing legislation for the past dozen years to curtail secret settlement agreements. He has sponsored proposed Sunshine in Litigation Acts in 1993,20 1994,21 1995,22 1996,23 1997,24 1998, 25 1999, 26 2000,27 2003,28 and 2005.29 On 18. Joseph F. Anderson, Jr., Hidden from the Public by Order of the Court: The Case Against Government-Enforced Secrecy, 55 S.C. L. REV. 711, 719–20 (2004). See also D.S.C. LOCAL R. 5.03. 19. D.S.C. LOCAL R. 5.03. See also Anderson, supra note 18, at 720–26. Initially this proscription was Rule 5.03(C), see id. at 720, but after further amendments to Rule 5.03, it is now Rule 5.03(E). The proscription is tempered somewhat by the court’s Local Rule 1.02, which provides that “For good cause shown in a particular case, the Court may suspend or modify any Local Civil Rule.” D.S.C. LOCAL R. 1.02. 21. See 140 Cong. Rec. S7685–86 (daily ed. June 27, 1994) (statement of Sen. Kohl); id. at S7719; 140 Cong. Rec. S13041 (daily ed. Sept. 12, 1994) (statement of Sen. Grassley). 23. S. 1887, 104th Cong. § 211 (as reported in Senate, July 30, 1996). The bill was enacted with- out the Sunshine in Litigation provisions. Federal Courts Improvement Act of 1996. Pub. L. No. 104-317, 110 Stat. 3847 (codified as amended in scattered sections of 18 U.S.C., 28 U.S.C., and 42 U.S.C. (2000)). See 144 Cong. Rec. S6054 (daily ed. June 10, 1998). Senator Kohl’s efforts were preceded by efforts by Representative Larry Smith of Florida to enact H.R. 2017, the “Federal Sunshine in Litigation Act of 1990.” H.R. 2017, 102d Cong. (1991). REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM September 18, 2002, Senator Kohl asked the Judicial Conference30 to con-sider appropriate changes to the Federal Rules of Civil Procedure to pro-scribe secret settlements: “The American people are outraged that our judicial system is complicit in these secret settlements.”31 Ralph Mecham, director of the Administrative Office of the United States Courts and secretary to the Judicial Conference, responded to Sena-tor Kohl’s letter on October 3, 2002. Mr. Mecham promised that the Civil Rules Advisory Committee was “now planning a thorough investigation of the matter. It [would] begin its work with a review of the pertinent local district court rules and state court rules as well as the case law to determine whether a national rule governing sealing settlements is feasible and appro-priate.”32 In October 2002, Judge David Levi, a federal district judge in Sacra- mento, California, was chair of the Civil Rules Advisory Committee.33 On October 1, 2002, Judge Levi sent an e-mail message to Tom Willging, a senior research associate at the Federal Judicial Center and the Center’s usual representative at meetings of the Civil Rules Advisory Committee: Tom: Senator Kohl has asked the advisory committee to look at sealing orders entered to keep settlements confidential. You may know that the District of South Carolina has published a local rule for comment that would prohibit the sealing of settlements. We have agreed to put the mat-ter on the committee’s agenda. I’m not certain whether I will refer this to the Discovery sub-committee (on the theory that it is related to protective orders) or not, but I do hope we may have your empirical support. A number of questions come to mind: how often are sealing orders en-tered? why? why are settlements filed? (usually they are not in my ex-perience) in what kinds of cases are sealing orders entered? is the entire file sealed? why do lawyers seek such orders? (fact of settlement or amount?) are there local/state rules that apply? I’m sure you will think of 30. The Judicial Conference of the United States is the federal judiciary’s policy-making body. It is chaired by the Chief Justice of the United States and includes the chief judge of each circuit and one district judge from each circuit. See 28 U.S.C. § 331 (2000). 31. Letter from Herb Kohl, U.S. Senator, to L. Ralph Mecham, Sec’y, Judicial Conference of the U.S. (Sept. 18, 2002) (on file with author). This letter was included in agenda materials at the October 2003 meeting of the Advisory Committee on Civil Rules. 32. Letter from Leonidas Ralph Mecham, Sec’y, Judicial Conference of the U.S., to Herb Kohl, U.S. Senator (Oct. 3, 2002) (on file with author). This letter was included in agenda materials at the October 2003 meeting of the Advisory Committee on Civil Rules. 33. Judge Levi is now chief judge of the Eastern District of California and chair of the standing Committee on Rules of Practice and Procedure, a parent committee that coordinates the work of the Civil Rules and other advisory committees. The Civil Rules Advisory Committee is now chaired by Judge Lee Rosenthal (S.D. Tex.). REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM THE HUNT FOR SEALED SETTLEMENT AGREEMENTS 443 other questions. I do hope that the FJC will be able to support us on this. I didn’t want to spring this on you in Santa Fe. Thank you.34 The Federal Judicial Center asked me to direct the project responding to Judge Levi’s request. Within a few weeks of my receiving the assign-ment, after some pilot investigations, I developed a research approach cen-tering on the three following questions:35 1. How often are sealed settlement agreements filed? It looks like we can get a reasonably good idea of this by examining electronic docket sheets. An initial order-of-magnitude estimate is one in a thousand civil filings, by which I mean substantially more than one in ten thousand and sub-stantially fewer than one in a hundred. 2. What are sealed settlement agreements all about? To answer this ques-tion, I think we first have to find the cases, which we can do to some ex-tent by first examining electronic docket sheets and then perhaps examining unsealed documents in the cases, such as the complaints, mo-tions to seal, orders of judgment, etc. After we figure out what we can learn from these public sources we will be in a better position to think about whether we want to know more and how to get it. 3. What are local rules and practices? For example, the District of South Carolina has a new local rule forbidding the filing of sealed settlement agreements (Rule 5.03(C)). Both the Southern District of Florida and the Eastern District of Michigan have local rules limiting the time that filed settlement agreements can be sealed (S.D. Fla. Rule 5.4 para. B.2, E.D. Mich. Rule 5.4). I already have the strong sense that sealed documents are seen as a significant burden on the clerk’s office, because they take up valuable vault space even after the rest of the file has been shipped off-site.36 In this section, I describe our answers to these research questions, be- ginning with the third and ending with the second. 34. E-mail from David F. Levi, District Court Judge, E.D. Cal., to Tom Willging, Senior Research Assoc., Fed. Judicial Ctr. (Oct. 1, 2002) (on file with author). The Civil Rules Advisory Committee met October 3–4, 2002, in Santa Fe, New Mexico. 35. For more on this research approach, see TIM REAGAN ET AL., FED. JUDICIAL CTR., SEALED SETTLEMENT AGREEMENTS IN FEDERAL DISTRICT COURT—MAY 2003 PROGRESS REPORT 1–2 (2003) [hereinafter FJC http://www.fjc.gov/public/pdf.nsf/lookup/SealSet2.pdf/$file/SealSet2.pdf. 36. E-mail from author to John Rabiej, Chief, Rules Comm. Support Office, Admin. Office of U.S. Courts (Dec. 6, 2002) (on file with author) (preceding a conference telephone call to discuss the research). See also D.S.C. LOCAL R. 5.03(E); S.D. FLA. LOCAL R. 5.4 para. B.2; E.D. MICH. LOCAL R. 5.4. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM A. What Are the Local Rules and Practices? What are local rules and practices? Our research on this question had two parts. The first, major part was a compilation of state and federal rules concerning the sealing of court documents. My colleague Marie Leary undertook this endeavor, and we were able to complete the task in time to present the results to the Civil Rules Advisory Committee at its May 2003 meeting.37 The second part of this research effort involved interviews with district clerks and their staffs concerning local practices. We endeavored to find every state and federal statute and rule pertain- ing to the sealing of court records in civil cases. In the state courts, sealing is generally a statutory matter. In federal courts, it is generally a matter of local rule. We found statutes or rules concerning the sealing of court records in civil cases in twenty-nine states (58%).38 For seven states (14%), the rules did not limit the sealing of documents, but just covered such issues as ad-ministrative mechanics.39 Each of the states’ rules limiting the sealing of court records included one or more of five themes: (1) a good-cause requirement (eleven states40); (2) a least-restrictive-means requirement (seven states41); (3) a time limita-tion on how long documents may remain sealed (three states42); (4) special concern about not sealing settlement agreements with public parties (ten states43); and (5) special concern about cases involving public hazards (one state44). The themes in the federal rules were a little different. Among the fed- eral rules there was more concern with how long documents remain sealed and no singling out of settlement agreements with public parties. Of the ninety-four federal district courts, forty-seven districts (50%) had local rules concerning the sealing of court records in civil cases.45 For FJC 2003 REPORT, supra note 35, at 2–5. 40. California, Delaware, Georgia, Indiana, Michigan, New Hampshire, New York, North Caro- lina, Tennessee, Utah, and Vermont. See id. 41. California, Florida, Idaho, Michigan, New Hampshire, North Carolina, and Texas. See id. 42. Delaware, Indiana, and New Mexico (two of its thirteen judicial districts). See id. at 5–6. 43. Arkansas, Colorado, Florida, Iowa, Montana, Nevada, North Carolina, Oklahoma, Rhode Island, and Texas. See id. at 5. 44. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM THE HUNT FOR SEALED SETTLEMENT AGREEMENTS 445 fifteen districts (16%), the rules covered administrative issues rather than limitations on sealing.46 Of the five themes addressed by the state rules, the federal rules addressed only three: (1) a good-cause requirement (eleven districts47); (2) a least-restrictive-means requirement (two districts48); and (3) a time limitation (twenty-nine districts49). The federal courts’ heavy emphasis on time limitations appears to be motivated by the clerks’ offices needing to clean out their vaults. Sealed documents generally are not sent to records centers because records centers are ill-equipped to enforce the seals. So, even after the rest of the case file has been shipped to records centers for long-term storage, clerks generally must retain sealed documents in their vaults. But upon the expiration of the sealing period sealed documents do not necessarily become public records. Courts’ local rules often provide that they can be destroyed or returned to the parties instead. We finished a compilation of state and federal rules on sealing court records in time for the May 2003 meeting of the Civil Rules Advisory Committee, but we did not finish our work on the other research questions until the April 2004 meeting. The other work involved mostly a review of federal case files. Because we were working for a federal rules committee, we decided to update our compilation of federal rules on sealing when we presented the other research results in 2004. We did not think we could also manage an updated compilation of state rules. We turned our April 2004 report into a Federal Judicial Center publication,50 and we received criti-cism from a special interest group for not including in this report a compi- 47. Northern District of California, Northern District of Illinois, District of Maryland, Western District of Michigan, Northern and Southern Districts of Mississippi, Eastern District of Missouri, Northern District of Oklahoma, Eastern District of Tennessee, District of Utah, and Western District of Washington. See generally id. The Northern and Southern Districts of Mississippi are listed as such instead of separately— “Northern District of Mississippi, the Southern District of Mississippi”—because one set of local rules governs both districts. 48. Northern District of California and District of Maryland. See generally, id. at 3. 49. Northern District of California, Southern District of California, District of Connecticut, South- ern District of Florida, District of Idaho, Northern District of Illinois, Northern and Southern Districts of Iowa, District of Kansas, District of Maryland, Eastern District of Michigan, Western District of Michigan, District of Minnesota, Northern and Southern Districts of Mississippi, Eastern District of Missouri, Eastern District of New York, Eastern District of North Carolina, Middle District of North Carolina, Western District of North Carolina, District of North Dakota, Northern District of Ohio, Middle District of Pennsylvania, Eastern District of Tennessee, Eastern District of Texas, Northern District of Texas, District of Utah, Western District of Virginia, Western District of Washington. See generally id. 50. See ROBERT TIMOTHY REAGAN ET AL., FED. JUDICIAL CTR., SEALED SETTLEMENT AGREEMENTS IN FEDERAL DISTRICT COURT (2004) [hereinafter FJC 2004 REPORT], available at http://www.fjc.gov/public/pdf.nsf/lookup/SealSet3.pdf/$file/SealSet3.pdf. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM lation of state rules. We did not do that because our compilation of state rules was already out of date. We had already met our obligation to the committee by providing it with a state and federal compilation in 2003, and in 2004 we presented them with an updated federal compilation to accom-pany our research on federal case files. To address the concerns of the spe-cial interest group, the Federal Judicial Center has posted on its website51 both the “published” 2004 report, containing a complete review of federal case files with a 2004 compilation of federal rules, and the “unpublished” 2003 report, containing a partial review of federal case files with a 2003 compilation of both state and federal rules. So, reader beware. The summary of state and federal rules given here is based on our 2003 report, and things may have changed a bit since then. The research I describe next involves a review of civil case files in a sample of federal districts. Before we contacted the districts to acquire the documents we needed to review, we interviewed their clerks or senior staff members concerning how sealed documents are handled in their districts. From these interviews we learned about the challenges of keeping sealed documents sealed. We also learned something that was not at all surprising: different districts, and sometimes different offices in the same district, handle sealed documents differently. One important way in which districts differ concerns how docket sheets refer to sealed documents. Sometimes the docket sheet specifies what kind of document is filed under seal—a settlement agreement, for example—and sometimes the docket sheet only notes that a “sealed document” is filed. Another way in which offices or districts can be different concerns how the sealing is achieved. One recently hired clerk told us that in one of her offices the prior practice had been for the clerk’s office to simply file under seal anything a person filing a document asked to be filed under seal.52 See Federal Judicial Center, http://www.fjc.gov (follow “Recent Publications” hyperlink) (last 52. This practice comports with somewhat frequent expectations. See Anderson, supra note 18, at 718–19. (“Many lawyers, and occasionally judges, assume all that is required for the sealing of discov-ery and other documents filed with the court is a simple directive from the attorney (or more commonly, the courier for the attorney) filing the documents to the intake deputy in the Clerk of Court’s office.”). REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM THE HUNT FOR SEALED SETTLEMENT AGREEMENTS 447 B. How Often Are Sealed Settlement Agreements Filed?53 In order to determine how often sealed settlement agreements are filed, we first had to locate them within the case files. Our research on case files essentially included combing through case records, searching for sealed settlement agreements, and then preparing short descriptions of cases with settlement agreements filed under seal. After some preliminary research we discovered that we could fairly reliably find sealed settlement agreements by searching for the word “seal” in docket sheets. We were fortunate in this respect. It would be very odd for part of the court record to be sealed without the word “seal” (including such variations as “sealed,” “sealing,” etc.) ap-pearing in the docket sheet (e.g., “document filed under seal,” “motion to seal,” “sealing order”). The presence of the word “seal” in a docket sheet, therefore, is extremely diagnostically sensitive to a sealed settlement agreement in the case file, which means there are likely to be very, very few false negatives, if any. Diagnostic specificity, on the other hand, is not as high with this method alone—there are likely to be many false positives. We observed a false positive rate of approximately 92%. Much more frequent than the sealing of settlement agreements is the sealing of exhibits submitted with discovery and summary judgment motions. And Crown Cork & Seal ap-pears as a litigant rather frequently. Still, looking for the word “seal” proved to be a very useful first step. A few years ago we could not have done this research. Much is possi- ble now that so much information is available electronically. We were able to download electronic copies of all unsealed docket sheets for federal civil cases terminated in 2001 and 2002, except for cases in the District of the Northern Marina Islands. I chose to look at two years’ worth of cases so that our results would not be distorted by a possibly idiosyncratic year, and I chose to look at all months so that there would be no seasonal distortions. I decided we should analyze the districts’ case data in a somewhat random order, so that if time did not permit a review of all districts’ data, the data we did review would likely be representative of all the data. I sup-posed that practices among districts in the same state might have some similarities, so I decided we should study all districts in the same state to-gether.54 Putting all the states and the federal districts outside states55 in an 53. For a more detailed discussion of this topic and the method used, see FJC 2004 REPORT, supra 54. It turned out that districts in the same state were not much more similar to each other than they REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM essentially random order, we worked our way through the list until we de-cided that it would be better to present the data than to continue the re-search. We ended up studying a little more than half of the districts (fifty-two out of ninety-four, or 55%).56 Staff member George Cort used his considerable computer skills to download the docket sheets and search for the word “seal.” For each dis-trict he was able to give me a text file that included only docket entries containing the word “seal.” Each docket entry included the case number, so I could see which entries were from the same case. He was also able to give me a text file that included not only these docket entries, but also docket entries containing the word “settle” if the word “seal” appeared somewhere in the docket sheet. The target words “seal” and “settle” were highlighted, which helped us to determine which cases needed further examination. Four research colleagues with law degrees—Natacha Blain, Steve Gensler, Marie Leary, and Shannon Wheatman—and I reviewed the docket sheets and case files. Looking through all of the docket entries containing the words “seal” or “settle,” we identified cases with docket sheets we should examine more thoroughly essentially by eliminating the cases that needed no further review. For example, if the target word “seal” appeared only in the party’s name (e.g., Henry Seale or Crown Cork & Seal), then we did not need to examine that case further. If the docket entry made clear that whatever sealing there was in the case pertained only to a discovery or summary judgment motion, then we did not need to examine that case fur-ther. If, on the other hand, the docket entry said merely “sealed document 55. Districts ouside states include the Districts of Columbia, Guam, Puerto Rico, and the Virgin Islands. Docket sheets for the District of the Northern Mariana Islands were not available electronically. Had we ended up studying all districts, I would have requested paper copies and reviewed them by hand. 56. Judge Levi assigned oversight of our work to a subcommittee chaired by the Honorable Brent McKnight, who was at first a magistrate judge for the Western District of North Carolina, then elevated to district judge during the course of the research, but who very unfortunately died shortly after the research was completed. We studied the districts of North Carolina first, so that Judge McKnight would have an opportunity to compare our results to his experience. We also studied the districts of Michigan and South Carolina early in the project, because the Eastern District of Michigan and the District of South Carolina had local rules specifically addressing the sealing of settlement agreements. We studied the districts of Florida early in the project because of the state’s groundbreaking “Sunshine in Litiga-tion” law. I sorted the other states and districts outside states randomly after North Carolina, Michigan, South Carolina, and Florida. We decided to stop work after studying half of the districts. Because I suspected someone would want to know whether the frequency of sealed settlement agreements was related to whether the district had a local rule explicitly requiring good cause to seal a filed document, in addition to studying the first forty-seven districts on our essentially random list, we studied the five districts on the second half of the list with good-cause rules. That way we could compare all eleven districts with good-cause rules to a sizeable sample of other districts. We did not observe an overall difference in sealed settlement rates between districts with good-cause rules and districts without good-cause rules. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM THE HUNT FOR SEALED SETTLEMENT AGREEMENTS 449 filed,” then we had to examine the complete docket sheet for that case to get a better idea of whether the sealed document might be a settlement agreement. The Administrative Office of the United States Courts maintains data on all federal district court cases. Each data record includes (among other data) district, case number, filing date, and termination date. We used this database to obtain case numbers for our search for docket sheets. Of the 288,846 unsealed docket sheets in civil cases terminated in our fifty-two study districts in 2001 and 2002, we found the word “seal” in 15,043.57 Examining docket entries with the words “seal” or “settle” in those cases, we identified 2,262 cases as having docket sheets we needed to examine more thoroughly.58 Reviewing those 2,262 docket sheets, we were able to eliminate 852 cases.59 Most of these cases had sealed documents only at the beginning of qui tam actions or attached only to discovery motions, motions for sum-mary judgment, or motions in limine. For the other 1,410 cases, we exam-ined substantial portions of the case files.60 If we determined from a review of a docket sheet that the case might or did include a sealed settlement agreement, then we reviewed parts of the case file in order to determine two things: (1) What was the case about? (2) Was what looked like it might be a sealed settlement agreement really a sealed settlement agreement, and, if so, what could we learn about the set-tlement? We could answer these questions largely by examining the com-plaint and any documents that appeared to relate to the sealed document or the settlement of the case. Again the times were in our favor. Many of the court documents we wanted to examine were available electronically. The courts’ staffs were very helpful in providing us with access to the documents we could not access electronically. Often they were willing to retrieve, photocopy, and mail the documents we needed. Sometimes we visited the court to examine documents there, or we sent a research assistant to photocopy them for us. Some case files had already been sent to records centers by the time we asked to review them. For some of these, either the court retrieved the files or we ordered copies of the documents we needed from the records centers. For others, we visited the records centers and examined documents there. 57. See FJC 2004 REPORT, supra note 50, at A-2 to A-3. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM After reviewing selected unsealed documents of each case that might or did have a sealed settlement agreement, we determined whether the case in fact had a sealed settlement agreement. Usually we could make this de-termination with confidence. In unclear cases, we simply made our best guess. The following is an example of a case about which we could only guess, but we are reasonably confident that our guess is accurate: Robson v. Dale County Board of Education (AL-M 1:00-cv-01037 filed 08/02/2000). Civil rights action by a substitute teacher for retaliation for exercising her First Amendment freedom of speech. The parties settled, and the court granted the parties’ joint stipulation to dismiss with preju-dice the individual defendants (the principal and the school superinten-dent). The settlement agreement with the remaining defendant school board apparently was filed under seal, because the docket sheet indicates that after the court granted the parties’ sealed joint motion to seal, the case was closed pursuant to a sealed order and a document was filed un-der seal the same day.61 We counted a case as having a sealed settlement agreement even if the settlement agreement subsequently became unsealed for some reason. Supnick v. Amazon.com Inc. (WA-W 2:00-cv-00221 filed 02/11/2000). Class action involving Web navigation software that gave the defendant access to users’ names, passwords, and other confidential information. A sealed settlement agreement was filed. One week after the settlement agreement was filed, it was unsealed. The defendant agreed to modify its software so that it does not collect confidential information. The defen-dant agreed to pay $1.9 million to named plaintiffs and a class of ap-proximately 47,500, and $100,000 to a fund that will provide grants to university-based programs with Internet public policy issues.62 Occasionally a sealed settlement agreement was filed with the court for the first time in order to enforce it. Erway v. Mayport Wholesale Seafood Inc. (FL-M 3:01-cv-00733 filed 06/27/2001). Employment action by a supervisor for sexual harassment and retaliation. The case was dismissed as settled. A sealed settlement agreement was filed as an attachment to the plaintiff’s motion to enforce the settlement agreement. The court denied the motion.63 Sometimes the party filing a motion to enforce did not take steps to Charles River Associates v. Hale Trans Inc. (MD 1:00-cv-02760 filed 09/14/2000). Contract case involving failure to pay for services rendered to assist the defendant in an antitrust lawsuit. The defendant filed a third- Id. at C-1. Cases are identified by district and case number. “AL-M,” for example, means the Middle District of Alabama. The beginning of the case number, “1:00,” means that the case was filed in office 1 of the district in the year 2000. The notation “cv” means that this is a civil case. The five digits following “cv” complete the case number. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM THE HUNT FOR SEALED SETTLEMENT AGREEMENTS 451 party complaint for legal malpractice against its former attorneys. After a settlement conference the court dismissed the case. The plaintiff filed a motion to revoke the dismissal and reopen the case, because the primary defendant did not pay the settlement. This motion included a letter from the magistrate judge that revealed the settlement amount of $162,000. The third-party defendant filed a sealed settlement agreement as an ex-hibit to a motion to enforce it.64 There were some cases that we counted as including sealed settlement agreements that others might not have counted. If the court sealed the tran-script of a successful settlement conference, we counted the case as one with a sealed settlement agreement. Giobbi v. Lahaina Divers Inc. (HI 1:00-cv-00005 filed 01/04/2000). Per-sonal injury action by a woman who was injured by a boat propeller while swimming. The settlement was placed on the record under seal during a settlement conference.65 I mentioned that these were the methods we used for cases with un- sealed docket sheets. We found that among the fifty-two study districts 138 cases were terminated in 2001 or 2002 with sealed docket sheets.66 These were cases for which we had a case number but could not retrieve the docket sheet electronically. For a few of these cases, the courts were will-ing to let us examine the docket sheets, and we could usually reliably de-termine whether the cases contained sealed settlement agreements. For the majority of cases with sealed docket sheets, however, we regarded cases terminated by consent judgment or settlement—based on the Administra-tive Office’s data—as cases with sealed settlement agreements, and cases terminated otherwise as cases without sealed settlement agreements. We can say very little else about these cases. Sometimes the Administrative Office data include case names, and sometimes it does not. Compass Marine v. Lambert Fenchurch (MS-S 1:99-cv-00252 filed 04/05/1999). Fraud action. The docket sheet is sealed. The case was dis-missed as settled.67 Sealed Plaintiff v. Sealed Defendant (MN 0:02-cv-04270 filed 11/07/2002). Contract action. The docket sheet is sealed. The case was dismissed as settled.68 66. I considered the possibility that a district might keep a civil case so secret that it does not report any data about the case to the Administrative Office. I do not know whether this ever occurs, but if it does occur it is likely to occur only very, very rarely. Caseload is an important factor in the alloca-tion of resources to the districts, so it is in their interest to report all cases, even if they withhold some information about some of them. 67. FJC 2004 REPORT, supra note 50, at C-57. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM Using these methods we determined that of the 288,846 civil cases terminated in 2001 or 2002 in the fifty-two study districts, 1,270 had sealed settlement agreements, which is 0.44%—less than one-half of one per-cent.69 C. What Are Sealed Settlement Agreements All About?70 From our analysis of the court records, we were able to determine what types of cases have sealed settlement agreements, specify what we know about such cases, and infer why settlement agreements are sealed. We then considered these findings in light of the relevant case law. We found that most of the cases with sealed settlement agreements are not cases the public would be especially interested in, and that typically the only part of the court record kept secret by a sealed settlement agreement is the amount of settlement. 1. Types of Cases with Sealed Settlement Agreements Every civil case filed in federal district courts has a civil cover sheet that specifies, among other things, the case’s “nature of suit.”71 The Ad-ministrative Office’s data on all civil cases include nature-of-suit codes. We can use these data to get a sense of what kinds of cases are settled by sealed settlement agreements, and whether the rate of sealed settlement agreements is higher or lower than average for particular natures of suit. For example, 378 of our cases with sealed settlement agreements had “personal injury” nature-of-suit codes.72 This group represented 30% of our 70. For a more detailed discussion of this topic, see id. at 1–8. 71. The civil cover sheet, also known as form JS 44, specifies that it was “approved by the Judi- cial Conference of the United States in September 1974 [and] is required for the use of the Clerk of Court for the purpose of initiating the civil docket sheet. Consequently, a civil cover sheet is submitted to the Clerk of Court for each civil complaint filed.” JUDICIAL CONFERENCE OF THE UNITED STATES, JS 44 CIVIL COVER SHEET (rev. 2004) [hereinafter CIVIL COVER SHEET JS 44]. The September 1974 Report of the Proceedings of the Judicial Conference of the United States states that the civil cover sheet was adopted as part of a civil docket package adopted pursuant to Federal Rule of Civil Procedure 79(a), DIR. OF THE ADMIN. OFFICE OF THE U.S. COURTS, REPORTS OF THE PROCEEDINGS OF THE JUDICIAL CONFERENCE OF THE UNITED STATES 50 (1974), which states that “the Director of the Ad-ministrative Office of the United States Courts with the approval of the Judicial Conference of the Unites States” may prescribe the district courts’ civil docket form and style. FED. R. CIV. P. 79(a). 72. See FJC 2004 REPORT, supra note 50, at 5 tbl.1. The more specific natures of suit include airplane (code 310, 168 cases); airplane product liability (code 315, 5 cases); assault, libel, and slander (code 320, 3 cases); federal employers’ liability (code 330, 4 cases); marine (code 340, 6 cases); marine product liability (code 345, 0 cases); motor vehicle (code 350, 26 cases); motor vehicle product liability (code 355, 14 cases); medical malpractice (code 362, 28 cases); product liability (code 365, 57 cases); asbestos personal injury product liability (code 368, 12 cases); and other personal injury (code 360, 55 cases). See CIVIL COVER SHEET JS 44, supra note 71. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM THE HUNT FOR SEALED SETTLEMENT AGREEMENTS 453 sealed settlement cases; more importantly, this group represented 0.82% of the personal injury cases terminated in our study districts during the study period, which is a rate nearly twice the average rate of 0.44%.73 Not sur-prisingly, personal injury actions are more likely to be resolved by sealed settlement agreement than other natures of suit; but even so, only a rela-tively small fraction of personal injury actions are resolved by sealed set-tlement agreement in federal district courts. Patent actions74 accounted for only 5% (sixty-two cases) of our cases with sealed settlement agreements, but that 5% included 2% of the patent cases terminated during the study years in the study courts.75 It is not sur-prising that patent actions might be resolved with sealed settlement agree-ments at nearly five times the average rate because patent actions often involve trade secrets, and it appears that settlement agreements often have to refer to the trade secrets. On the other end of the scale were real property cases.76 They accounted for 1% of our cases with sealed settlement agree-ments, and 0.07% of the real property cases in our study cohort had sealed settlement agreements.77 Frequencies for other natures of suit are in our 2004 report.78 Reviewing the 1,270 cases in our study with sealed settlement agree- ments, it was clear that the public is likely to be considerably more inter-ested in some of the cases than in others. For example, the public is likely to be very interested in a case involving sexual molestation of a child by a school bus driver. Doe v. Holcomb (VA-E 2:00-cv-00597 filed 08/15/2000). Personal in-jury action for sexual molestation of a Headstart student by a school bus driver. An agreed protective order held confidential (1) medical and psy-chological information about the plaintiff, (2) information concerning the criminal investigation of the bus driver, and (3) the identity of the plain-tiff. The court approved a sealed settlement agreement.79 The public is likely to be considerably less interested in a copyright Forestry Suppliers Inc. v. General Supply Corp. (MS-S 3:01-cv-00014 filed 01/09/2001). Copyright infringement action concerning forestry 73. FJC 2004 REPORT, supra note 50, at 5 tbl.1. See CIVIL COVER SHEET JS 44, supra note 71 (nature-of-suit code 830). FJC 2004 REPORT, supra note 50, at 5 tbl.1. 76. Real property cases include six more specific natures of suit: land condemnation (code 210, 0 cases); foreclosure (code 220, 3 cases); rent lease and ejectment (code 230, 2 cases); torts to land (code 240, 0 cases); tort product liability (code 245, 0 cases); and all other real property (code 290, 2 cases). See CIVIL COVER SHEET JS 44, supra note 71. FJC 2004 REPORT, supra note 50, at 5 tbl.1. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM catalogues. The case was dismissed pursuant to a sealed settlement agreement.80 What makes a case of special public interest? I discovered that re- searcher colleagues of mine did not agree. Some would have classified product liability actions, even if they did not involve personal injury, as of special public interest because the marketing of a defective product to the public puts the public at risk of injury, whether it is serious physical injury or less serious economic injury. So a case concerning defective windows might be of special public interest. Island Developers Ltd. v. Martin Lumber and Cedar Co. (FL-S 1:99-cv-02969 filed 11/03/1999). Contract action involving breach of implied warranty when defective wood windows were installed. In the order of dismissal the court retained jurisdiction to enforce the terms of the set-tlement agreement. Two months after the case was dismissed, a sealed document was filed the same day the plaintiff filed a motion to expedite enforcement of the settlement agreement. A sealed settlement agreement apparently was filed. The court denied the motion for oral argument, and the plaintiff withdrew the motion to expedite enforcement, because the parties resolved the issue.81 Others among us were inclined to regard cases concerning very seri- ous physical injuries as cases of special public interest even if the injuries resulted from a traffic accident. Estate of Carr v. Louisiana-Pacific Corp. (NC-W 5:99-cv-00023 filed 02/24/1999), consolidated with Estate of Carr v. Louisiana-Pacific Corp. (NC-W 5:99-cv-00024 filed 02/24/1999), Estate of Carr v. Louisiana-Pacific Corp. (NC-W 5:99-cv-00025 filed 02/24/1999), Estate of Phillips v. Louisiana-Pacific Corp. (NC-W 5:99-cv-00026 filed 02/24/1999), and Estate of Carr v. Louisiana-Pacific Corp. (NC-W 5:99-cv-00027 filed 02/24/1999). Consolidated motor vehicle tort action in which five dece-dents’ estates sued the alleged employers of a logging truck driver. Ac-cording to the complaints, the driver became distracted while changing a tape in his cab. He veered into oncoming traffic and ran a church van off the road. He then swerved back into the correct lane, and the truck’s logs spilled, crushing the van’s five occupants. The district court granted summary judgment to the defendants on the grounds that the driver was not their agent, and the plaintiffs appealed. The case settled on appeal. The court had to approve the settlement agreement, because one of the plaintiffs was a minor representing her father’s estate. Terms of the set-tlement agreement are under seal.82 So what we did was identify a few features that might make a case of special public interest, and we reported how many sealed settlement cases have each feature. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM THE HUNT FOR SEALED SETTLEMENT AGREEMENTS 455 Environmental cases are likely to be of special public interest. There were ten environmental cases with sealed settlement agreements in our study (1%).83 The following is an example: Castle & Cooke Properties Inc. v. BHP Hawaii Inc. (HI 1:98-cv-00923 filed 11/17/1998). Environmental case in which hazardous chemicals and petroleum products allegedly migrated from the defendant’s property to the plaintiff’s property, causing contamination of groundwater and soil. An unexecuted sealed settlement agreement was filed.84 Product liability cases are also likely to be of special public interest. We classified cases as product liability cases for these purposes based on all parts of the case file we reviewed; we did not rely on nature-of-suit codes alone. Among our sealed settlement cases there were 258 product liability cases (20%).85 The following is one example: Figueroa Rodríguez v. Volkswagen of America Inc. (PR 3:99-cv-01422 filed 04/16/1999). Motor vehicle product liability action for complica-tions from a fractured arm, which resulted from a Volkswagen Golf’s airbag exploding in a fender-bender. The plaintiffs filed a sealed infor-mative motion on September 18, which the court granted “until October 19.” The parties filed a stipulation for dismissal on October 3, and the case was dismissed on October 18. The sealed motion probably con-tained terms of settlement.86 Most of the product liability cases were in two large consolidations of cases arising from airplane crashes. A 1998 airline crash in Nova Scotia resulted in 144 consolidated cases in the Eastern District of Pennsylvania with sealed settlement agreements. In re Air Crash Disaster Near Peggy’s Cove, Nova Scotia on September 2, 1998 (MDL 1269 transferred 04/07/1999). Airline cases claiming that an aircraft crashed because of a malfunction in the in-flight entertain-ment center, killing 229 passengers and crew members. A stipulation and order was filed under seal in each member case on the same day that the case was closed.87 Another thirty-three sealed settlement cases in our termination cohort arose from the ill-fated TWA flight 800 from New York to Paris on July 17, 1996. In re Air Crash TWA (lead case filed 10/24/1996). Consolidated airplane actions for wrongful death (with one case designated an airplane product liability action) arising from the July 17, 1996, crash of TWA flight 800 from Kennedy Airport to Paris, France, which was allegedly caused by a fuel cell explosion. In each of these cases, the plaintiffs settled with the 87. Id. at C-91 (footnote omitted). REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM defendants and dismissed their actions with prejudice. The docket sheet for each of these cases indicates “sealed document placed in vault” either the same day as settlement or several days before or after settlement.88 It is important not to think that just because two large consolidations account for so many cases that they somehow do not count. There is a temptation to reason, for example, that if the plane had not crashed in Nova Scotia, then the sealed settlement rate would have been 0.39% (1,126 di-vided by 288,702, subtracting 144 from both the numerator and denomina-tor). It is possible that the plane might not have crashed in Nova Scotia, but it is also possible that 250 sealed settlement agreements might have been filed in consolidated actions terminated January 2, 2003, and were not in-cluded in our study merely because the cases did not terminate the previous week. It is important to recognize that results could change substantially if the facts were only a little different, but this does not give us license to change the facts. Professional malpractice is a sort of product liability for people, so professional malpractice cases are likely to be of special public interest. Our sealed settlement cases included forty professional malpractice cases (3%).89 Most of them are medical malpractice cases, such as the following example: Estate of Mayo v. Kindred Nursing Centers East LLC (NC-M 1:02-cv-00260 filed 04/05/2002). Medical malpractice action against a nursing home for wrongful death resulting from the insertion of a feeding tube into a patient’s trachea instead of her esophagus, resulting in her lungs receiving feeding solution. The case was dismissed pursuant to a sealed consent order.90 Ritchie v. Yanchunis (AZ 2:00-cv-01533 filed 08/09/2000). Personal in-jury action for legal malpractice in allowing the statute of limitation on an action for wrongful termination to lapse. The parties agreed to a con-fidential settlement agreement, and the court ordered the transcript of the settlement agreement filed under seal.91 The public is likely to have a special interest in actions against public parties. As observed above, many of the state limitations on sealed settle-ment agreements apply with special force to cases with public defendants. Our sealed settlement cases included 152 against public parties (12%),92 including the following case: 88. Id. at C-77 to C-78 (footnote omitted). REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM THE HUNT FOR SEALED SETTLEMENT AGREEMENTS 457 Jeswald v. Frontier Central School District (NY-W 1:00-cv-00824 filed 09/22/2000). Employment discrimination action by an obese Native-American woman, alleging wrongful termination on the basis of her race and disability by a school district. The defendant informed the court that the parties had settled. Five months later a document was filed under seal, and the case was closed the same day.93 We found that approximately one-quarter of the cases in which we found sealed settlement agreements involved very serious physical inju-ries—death or serious permanent disabilities.94 Most of these 334 cases (26% of the sealed settlement cases) are cases of special public interest because they are environmental, product liability, professional malpractice, or public-party defendant cases. Thirty-nine cases have sealed settlement agreements and involve very serious injuries but are not environmental, product liability, professional malpractice, or public-party defendant cases. The following is an example: Parkhill v. Starwood Hotels & Resorts Worldwide Inc. (MI-E 2:00-cv-71877 filed 04/24/2000). Personal injury action for quadriplegic spinal cord injuries sustained by the plaintiff while swimming in the ocean at the defendant’s hotel. The case settled, and approximately three months after the filing of the stipulated order of dismissal—on the statistical date of termination—a sealed document was filed; this may be a settlement agreement.95 The final case feature we identified as possibly making a case one of special public interest is sexual abuse. Thirty-one of the sealed settlement cases (2%) were actions for sexual abuse.96 Some of them involved sexual abuse of children. McCollins v. Woods Services Inc. (PA-E 2:01-cv-00110 filed 01/08/2001). Personal injury action by a woman and her 12-year-old son, who was sexually abused by a 15-year-old student at a private residential school for the mentally retarded. The plaintiffs’ petition to approve the settlement and the order approving settlement were sealed.97 Some of them involved sexual abuse of adults. Cerpas v. University of California, San Francisco (CA-N 3:00-cv-04505 filed 12/01/2000). Employment action by a customer service representa-tive claiming that her supervisor coerced her into accompanying him to a motel room for sex. The case settled at a settlement conference, and the confidential settlement agreement was placed on the record. The re-porter’s transcript was filed under seal.98 REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM Of all of the cases with sealed settlement agreements in our study, 40% had one or more of the features described here that might make them cases of special public interest.99 2. What We Know About Cases with Sealed Settlement Agreements The selection of case descriptions from our 2004 report reproduced for this article should make clear that the sealing of a settlement agreement does not deprive the public of all information about a case. In 97% of the cases in which we found sealed settlement agreements the complaint was not sealed.100 Thirty-nine cases had sealed complaints.101 In twenty-three of these cases, the whole case file, including the docket sheet, was sealed.102 We counted them as cases with sealed settlement agreements because they were coded as settled in the Administrative Office’s data. In fifteen of the cases with sealed complaints, the docket sheet was not sealed but all of the documents filed in the case were sealed.103 We can often learn a lot about a case from the docket sheet alone, however. Baker v. Bollinger (MI-E 4:00-cv-40239 filed 06/26/2000). Employment case against the University of Michigan and some of its employees. The case file includes a protective order concerning confidential health in-formation. The court granted the parties’ joint motion for a stipulated permanent injunction and sealing of the record.104 The other case with a sealed settlement agreement and a sealed com- plaint is one in which an agreed judgment specifying the terms of settle-ment was the only unsealed document in the file. Doe v. City of Tulsa (OK-N 4:00-cv-00896 filed 10/18/2000). Civil rights action alleging failure to take adequate precautions to avoid public disclosure of the plaintiff’s medical condition. The case was dismissed as settled. The court ordered all matters related to the case sealed except for the agreed judgment, which discloses the amount of settlement to be $9,000. The sealed documents include the complaint, settlement confer-ence report, and administrative closing order.105 We counted this as a case with a sealed settlement agreement because other documents containing terms of settlement were sealed. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM THE HUNT FOR SEALED SETTLEMENT AGREEMENTS 459 As we saw with CBS’s 60 Minutes II presentation of Annie Davis’s action against CVS, the public may have access to a lot of information about a case even if a settlement agreement is sealed. 3. Why Are Settlement Agreements Sealed? It is well-known that a large proportion of civil cases settle. Prelimi- nary research for this project showed that 22% of a selection of federal civil cases were coded in the Administrative Office’s database of cases as set-tled, and another 2% were coded as terminated by consent judgments.106 Many cases coded as voluntary dismissals or “other” dismissals probably were settled as well. It is very common for settlement agreements to include confidentiality provisions. Uncertainty about case outcomes works to the advantage of repeat defendants. If the settlement is large, confidentiality can also protect plaintiffs, especially minors, from predators. Typically, however, it is not necessary for settlement agreements to be filed. The case can be dismissed voluntarily by the plaintiff, and the settlement agreement can be enforced as a private contract.107 Why would a settlement agreement be filed? First, settlement agree- ments are filed in cases requiring court approval of the agreements.108 Such cases include cases with minors as parties, class actions, and cases under the Fair Labor Standards Act.109 Second, a settlement agreement might be filed in federal court to preserve federal jurisdiction over its enforce-ment.110 Parties typically seek to file settlement agreements under seal to give effect to confidentiality terms. What is kept secret by secret settlements? In the course of our research we were able to examine many settlement agreements. Sometimes a sealed settlement agreement became unsealed or was, perhaps inadvertently, filed unsealed attached to a motion to enforce it. Sometimes we were given ac-cess to sealed documents. The agreements we saw confirmed what we might have supposed in advance. [S]ettlement agreements, sealed or otherwise, generally contain four es-sential elements: (1) a denial of liability, (2) a release of liability, (3) the amount of settlement, and (4) a requirement of confidentiality. In unfair 107. See Doré, supra note 1, at 798–800. 108. One might think that for the court to approve a settlement agreement, the agreement would have to be filed, but my impression in reviewing case files for this project was that sometimes settle-ment agreements were approved without being made part of the record. FJC 2004 REPORT, supra note 50, at 5. 110. See Doré, supra note 1, at 801. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM competition cases, especially cases involving patents, the terms of set-tlement typically bind the parties to certain actions in addition to or in-stead of the payment of a settlement amount. In general, however, the only thing kept secret by the sealing of a settlement agreement is the amount of settlement.111 Court documents are presumptively public. The prevailing view ap- pears to be that if part of the court record is sealed, it should only be sealed on specific findings of a need to seal, and the public record should include as much information as is reasonable about what is sealed and why. As the Supreme Court stated, “It is clear that the courts of this country recognize a general right to inspect and copy public records and docu-ments, including judicial records and documents.”113 The Court went on to state, “It is uncontested, however, that the right to inspect and copy judicial records is not absolute. Every court has supervisory power over its own records and files, and access has been denied where court files might have become a vehicle for improper purposes.”114 Accountability is a principal reason for public access. “The political branches of government claim legitimacy by election, judges by reason. Any step that withdraws an element of the judicial process from public view makes the ensuing decision look more like fiat, which requires com-pelling justification.”115 “As with other branches of government, the bright light cast upon the judicial process by public observation diminishes possi-bilities for injustice, incompetence, perjury, and fraud.”116 Although the common law presumption of access applies to docu- ments filed with the court, this presumption of access does not apply to documents exchanged in discovery,117 or to settlement agreements not filed.118 But if a settlement agreement is filed with the court for the court’s FJC 2004 REPORT, supra note 50, at 7. 112. For a more detailed discussion of this topic, see id. at 1–2. 113. Nixon v. Warner Commc’ns, Inc., 435 U.S. 589, 597 (1978) (footnotes omitted). 115. Union Oil Co. of Cal. v. Leavell, 220 F.3d 562, 568 (7th Cir. 2000). 116. Littlejohn v. BIC Corp., 851 F.2d 673, 678 (3rd Cir. 1988). See also Joy v. North, 692 F.2d 880, 893 (2d Cir. 1982) (“An adjudication is a formal act of government, the basis of which should, absent exceptional circumstances, be subject to public scrutiny.”); Jessup v. Luther, 277 F.3d 926, 928 (7th Cir. 2002) (“[T]he public cannot monitor judicial performance adequately if the records of judicial proceedings are secret.”); id. at 929 (“The public has an interest in knowing what terms of settlement a federal judge would approve and perhaps therefore nudge the parties to agree to.”). 117. Fed. Trade Comm’n v. Standard Fin. Mgmt. Corp., 830 F.2d 404, 408 (1st Cir. 1987); United States v. Amodeo, 71 F.3d 1044, 1050 (2d Cir. 1995). 118. Gambale v. Deutsche Bank AG, 377 F.3d 133, 143 (2d Cir. 2004); Pansy v. Borough of Stroudsburg, 23 F.3d 772, 781–83 (3d Cir. 1994). REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM THE HUNT FOR SEALED SETTLEMENT AGREEMENTS 461 approval or interpretation, then denying the public access to the agreement requires special circumstances.119 And the court should make a record of specific reasons for sealing a filed document.120 If federal district courts were diligent in sealing only what really needs to be sealed, and making a public record of what is sealed and why, then our research for this project would have been easier than it was. In the large majority of cases, we would have been able to determine whether a case file contained a sealed settlement agreement by reviewing the docket sheet alone. My impression after reviewing many hundreds of case files for this project is that, although the sealing of settlement agreements in federal courts is considerably less frequent than many people fear, there often is an insufficient public record of what is sealed and why. It is clear that our research was helpful to the Civil Rules Advisory Committee. Relying on our research, “the Advisory Committee will con-tinue to monitor court practices but does not intend to propose any new rules at this time.”121 This does not mean that the sealing of court records is not a matter of concern for courts or the public. Nor does it mean that confidential, as op- The presumption of public access is stronger for documents filed in conjunction with substan- tive action by the court than for documents filed as part of discovery disputes. Anderson v. Cryovac, Inc., 805 F.2d 1, 11 (1st Cir. 1986); Leucadia Inc. v. Applied Extrusion Techs., Inc., 998 F.2d 157, 165 (3d Cir. 1993); Foltz v. State Farm Mut. Auto. Ins. Co., 331 F.3d 1122, 1135–36 (9th Cir. 2003); Chi. Tribune Co. v. Bridgestone/Firestone, Inc., 263 F.3d 1304, 1312 (11th Cir. 2001). 119. See Bank of Am. Nat’l Trust & Sav. Ass’n v. Hotel Rittenhouse Assocs., 800 F.2d 339, 345 (3d Cir. 1986) (“Once a settlement is filed in the district court, it becomes a judicial record, and subject to the access accorded such records.”); Herrnreiter v. Chi. Hous. Auth., 281 F.3d 634, 637 (7th Cir. 2002) (“[Defendant’s] desire to keep the amount of its payment quiet (perhaps to avoid looking like an easy mark, and thus drawing more suits) is not nearly on a par with national security and trade secret information. Now that the agreement itself has become a subject of litigation, it must be opened to the public just like other information (such as wages paid to an employee, or the price for an architect’s services) that becomes the subject of litigation.”); Brown v. Advantage Eng’g, Inc., 960 F.2d 1013, 1016 (11th Cir. 1992) (“It is immaterial whether the sealing of the record is an integral part of a negoti-ated settlement between the parties, even if the settlement comes with the court’s active encouragement. Once a matter is brought before a court for resolution, it is no longer solely the parties’ case, but also the public’s case. Absent a showing of extraordinary circumstances . . . , the court file must remain accessible to the public.”). 120. In re Cendant Corp., 260 F.3d 183, 194 (3d Cir. 2001) (“Broad allegations of harm, bereft of specific examples or articulated reasoning, are insufficient.”); Stone v. Univ. of Md. Med. Sys. Corp., 855 F.2d 178, 182 (4th Cir. 1988) (“[T]he district court must provide a clear statement, supported by specific findings, of its reasons for sealing any records or documents, as well as its reasons for rejecting measures less drastic than sealing them.”); Hagestad v. Tragesser, 49 F.3d 1430, 1435 (9th Cir. 1995) (“[B]ecause the district court failed to articulate any reason in support of its sealing order, meaningful appellate review is impossible.”). 121. CIVIL RULES ADVISORY COMM., MINUTES, OCT. 28–29, 2004, at 25, available at http://www.uscourts.gov/rules/Minutes/CRAC1004.pdf. REAGAN AUTHOR APPROVED EDITS 12-7-05 (H)(P) 9/12/2006 9:56:03 AM posed to sealed, settlement agreements are not a matter of concern. Nor does our research include any information on state cases.122 What our re-search shows is that sealed settlement agreements per se are not common in federal courts, and the seals typically keep secret only the amounts of set-tlement. 122. Note that state courts hear family law cases, in which there often are special reasons for seal- ing settlement agreements. See Jean Hoefer Toal & Bratton Riley, The New Role of Secret Settlements in the South Carolina Justice System, 55 S.C. L. REV. 761, 771 (2004).

Source: http://studentorgs.kentlaw.iit.edu/cklawreview/wp-content/uploads/sites/3/vol81no2/Reagan.pdf