Healey et al v. Jefferson County Kentucky Louisville Metro Government et al, No. 3:2017cv00071 - Document 72 (W.D. Ky. 2021)

Court Description: MEMORANDUM OPINION AND ORDER Signed by Judge Rebecca Grady Jennings on 1/15/2021. Plaintiffs' Motion to Certify Class (DE 50 ) is GRANTED IN PART and DENIED IN PART consistent with this OPINION. Counts I, II, III, IV, V, VII, and VIII of th e Third Amended Complaint shall be maintained as a class action under Fed. R. Civ. P. 23(b)(3) by Plaintiff Class Representatives on behalf of the subclasses. Attorneys James D. Ballinger and Gregory A. Belzley are appointed as co-class counsel for Plaintiffs as they meet all of the requirements of Fed. R. Civ. P. 23(g). Plaintiffs Larry Louis Hibbs, Jr. and Allen Goodfleisch are not part of the proposed class. Plaintiffs' counsel shall a file a status report as to Plaintiffs Hibbs and Goodfleisch within thirty (30) days of this Order. cc: Counsel (SMJ)

Download PDF

Healey et al v. Jefferson County Kentucky Louisville Metro Government et al01/15/21 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed Page 1 of 50 PageID #: 2087 Doc. 72 UNITED STATES DISTRICT COURT WESTERN DISTRICT OF KENTUCKY LOUISVILLE DIVISION CIVIL ACTION NO. 3:17-CV-00071-RGJ-RSE JACOB HEALEY, et al. PLAINTIFFS v. LOUISVILLE METRO GOVERNMENT, et al. DEFENDANTS MEMORANDUM OPINION AND ORDER Jacob Healey, James Michael Jarvis, Jr., Cynthia Dawn Yates, and Betty Melloan (“Plaintiffs or Named Plaintiffs”)1 move for class certification under Federal Rules of Civil Procedure 23(a) and 23(b)(1), (b)(2), and (b)(3). [DE 50]. Defendants responded and Plaintiffs replied. [DE 53; DE 56]. The Court conducted oral argument. [DE 67; DE 69; DE 71]. This matter is ripe. For the reasons below, Plaintiffs’ Motion [DE 50] is GRANTED IN PART and DENIED IN PART consistent with this Opinion. I. BACKGROUND Plaintiffs bring this putative class action against Louisville-Jefferson County Metro Government (“Metro Government”), Mark Bolton (“Bolton”), Director of the Louisville Metro Department of Corrections (“LMDC”), Arnetta Al-Amin (“Al-Amin”), Coordinator of LMDC’s records department, and Dwayne Clark (“Clark”), Chief of Staff of LMDC, alleging violations of Plaintiffs Larry Louis Hibbs, Jr. and Allen Goodfleisch are not part of the proposed class. [DE 50 at 1218 (“Movants do not include all of the named Plaintiffs in this Motion. The claims of Plaintiffs Larry Louis Hibbs, Jr. and Allen Goodfleisch relate solely to Defendants’ failure to effect their release for work or home Incarceration . . . Discovery has also shown that the determination of eligibility for work release and home incarceration is too complex, fact-intensive, and individualized for class certification and resolution”)]. During oral argument, Plaintiffs’ counsel stated that they would likely dismiss their claims if the Court granted class certification. [DE 69 at 1994]. 1 1 Dockets.Justia.com Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 2 of 50 PageID #: 2088 state law and seeking relief under 42 U.S.C. § 1983 for violating the Fourth, Eighth, and Fourteenth Amendments to the United States Constitution. [DE 37 at 346-349]. Plaintiffs allege that Defendants “regularly imprison, detain or incarcerate persons longer than ordered by Courts of the Commonwealth of Kentucky, and under conditions that violate the orders of such Courts. Such actions have been perpetrated by Defendants even though there exists no objectively reasonable legal grounds or justification for denying such persons timely and proper release from incarceration.” Id. at 347. Named Plaintiffs allege that LMDC held each of them for more than one day after they were to be released: 1. Plaintiff Jacob Healey was ordered by the court to serve only three days’ imprisonment at LMDC but ended up spending five – he was released only after his repeated inquiries as to why he was still being held [DN 37, Compl. at ¶ 14]; 2. Plaintiff James Michael Jarvis, Jr. remained incarcerated for three days after he was ordered released on his own recognizance (“ROR”). Id. at ¶ 16; 3. Plaintiff Cynthia Dawn Yates remained incarcerated for six days after her charges were dismissed. Id. at ¶ 18; and 4. Plaintiff Betty Melloan,2 despite being ROR’d at approximately 9:00 a.m. on January 8, 2018 (almost one year after this case was filed), was not released until after twelve noon the next day on January 9, 2018, almost 30 hours after the judge ordered her to be released. Id. at ¶ 19. [DE 50 at 1221-1222] (emphasis in original). Plaintiffs request certification of a class comprised of: All persons who, since February 3, 2016, were imprisoned, detained or incarcerated more than four hours after: (a) entry of an order directing their release; and/or (b) satisfaction of their term of incarceration set by prior court order. Id. at 1222. 2 See infra pp. 41-42. 2 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 3 of 50 PageID #: 2089 A. Facts 1. LMDC’s process for releasing inmates a. A contempt order and two audits In January 2017, Jefferson County District Court Judge Stephanie Burke ordered Bolton, Clark, and Al-Amin to appear before her to show cause why LMDC should not be held in contempt for its “repeated and patterned failure to comply” with her orders. [DE 50-1 at 1249-1254]. After detailing more than fifteen incidents where LMDC allegedly violated her orders, Judge Burke noted: This Notice includes only a limited number of incidents that this Court has encountered in recent months. It is important to note that after discussion with a number of my colleagues, it is evident that the frequency of errors like those enumerated herein has greatly increased since the early part of 2016. Many more incidents of similar nature have occurred and therefore, this Court chose to take this action now so as to prevent the continued disruption of the dockets and to ensure that the potential negative or harmful results of such actions are avoided. ... The errors enumerated herein have caused Defendants to be wrongfully detained. Id. at 1254-1255. In response to Judge Burke’s contempt order and the filing of this lawsuit, Metro Government Councilman Raymond James requested that the Metro Office of Internal Audit review LMDC’s “inmate and release and in custody classification movement activity.” [DE 49-3 at 813814; DE 50 at 1228; DE 50-9 at 1597]. In their final report, the internal auditors made a series of recommendations to “improve the internal control structure and effectiveness of the process for inmate releases and in custody classification movements.” [DE 50-9 at 1599]. They recommended that: 3 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 4 of 50 PageID #: 2090 LMDC and the Courts should continue efforts to collaborate in developing a streamlined method to share all essential information at the time information is needed and in a format that is useful. Each of the considerations for corrective action, noted below, require a collaborative effort between the Courts and LMDC, to include leadership and staff of both organizations. ... LMDC and the Courts should continue efforts to collaborate in developing court orders that are notated in a clear, consistent manner. ... LMDC and the Courts should develop and implement a protocol for periodic communication between each organization. ... LMDC and the Courts should collaborate to determine the most efficient and effective method for identifying court orders that require special consideration. ... [LMDC should] [d]evelop a method to track and monitor court orders from the time of receipt through completed processing. This is imperative to the ability to ensure all court orders that are received are processed. Additionally, this will allow LMDC to quickly identify any delays based on the amount of time a court order remains in the queue. ... [LMDC should] continue to work with the vendor of X-jail to expand the functionality so that it can serve as the single repository for sentencing directives. Until the functionality of X-jail can be adjusted, the Time-out Book and X-jail should be reconciled during the secondary review of the inmate sentence calculation. This will help to ensure consistency between the two repositories . . . [T]he current process of maintaining two repositories increases the likelihood of errors and/or delays in processing release activity. Id. at 1601-1616. Metro Government also hired CGL Companies (“CGL”), an external auditor, to “conduct a comprehensive audit of the judicial directive (court order) system in the county. The system had 4 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 5 of 50 PageID #: 2091 experienced issues with the timely implementation of court orders, as well as their clarity and consistency.” [DE 50-10 at 1662]. In its report, CGL described the processing of court orders in Jefferson County: Court orders are issued by judges in the specific court. Typically, a deputy clerk from the Jefferson County Office of the Circuit Court Clerk (Clerk’s Office) is assigned to the court and records the order for the judge. The judge approves the order and the deputy clerk transmits the order via email to a specific email address specified by the LMDC. LMDC Records Office staff open the order and review it. If the order is clear and accurate and does not conflict with other orders, then the order is processed for implementation. However, if the court order is unclear or inaccurate, then further work must be completed before the order can be implemented. This involves LMDC Records Office staff contacting either a deputy clerk or staff from the Courts to get clarification or to request the order be corrected. We note that this can be a laborintensive process and can potentially delay order implementation. Id. at 1628-1629. This system has “been in place for decades” and “system leadership is well aware of the existing limitations.” Id. at 1639. CGL found that system leadership has “dedicated significant resources over the past several years to improve the process,” but that these improvements “only begin to address a portion of the issues with the current system, and fall short of the benefits of an entirely automated court order workflow process.” Id. CGL estimated that between 2.5% to 6.3% of court orders require the issuing court to clarify or correct them before LMDC can implement them. Id. at 1626. LMDC informed CGL “that the initial effort to review the order, and email or call the Clerk’s Office or court supervisors can typically occupy up to 30 minutes of staff time.” Id. at 1630. LMDC also provided CGL access to information about “release errors” from 2016, 2017, and until September 2018. Id. at 1629. According to LMDC, there were sixteen release errors in 2016, four in 2017, and seventeen in 2018. Id. CGL found that the release errors in 2017 and 2018 “could be attributed to obvious 5 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 6 of 50 PageID #: 2092 mistakes by LMDC Records Office staff, while others reflected the complicated nature of interpreting multiple court orders that can be conflicting. It was also clear from [its] review that high turnover rates in the LMDC Records Office, resulting in inexperienced staff, likely contributed to a number of release errors.” Id. at 1629-1630. “The nature of the release errors information” prevented CGL for determining the “exact length of delay-of-release errors, as in many cases the verified release dates were not established in the incident reports.” Id. at 1630. After reviewing a sample set of 700 court orders, CGL determined that “[t]he underlying issue with the court order process is that it effectively is a manual workflow system. A court order is generated and manually transmitted to LMDC, who [sic] manually opens the document, prints it out, and processes it.” Id. at 1636. Although LMDC’s use of “email in-boxes” is an “improvement“ over the “hand-delivery of court orders,” it still “represents a process that requires individuals [to] manually initiate, send, receive, and print the court orders.” Id. As a result, the “manual system . . . opens the potential for manipulation and security intrusions.” Id. at 1637. CGL further cautioned that “[w]ith the existing system, paper documents could easily go missing, be deleted from email inboxes, or even be taken by individuals with ill-intent who are involved or have access to the process.” Id. b. Metro Council’s resolution In October 2018, Metro Council approved a resolution “supporting Metro Technology’s inclusion in the development of a new case management system for use in Kentucky’s courts.” [DE 50 at 1232]. Councilman Bill Hollander described the purpose of the resolution: We had a discussion at the budget committee as a result of an audit last year . . . The audit was about the way our court system and the clerk’s office speaks to the jail in terms of who should be in the jail, who should be out of the jail, when they can get out of the jail. And what we found as a result of that audit is that, first of all, there are, as you would expect, a very large number of these cases in Jefferson 6 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 7 of 50 PageID #: 2093 County, far more than in any other county in the state. And, in a paper-based system, there are also errors. And the errors result in inconsistencies, incompleteness, lack of an audit trail, and sometimes potential liability, our liability, for having individuals detained longer and not being released in a timely fashion . . . So . . . one of the strong recommendations from the audit was that as the state is rolling out and developing a new computer system that Metro people have some seats at the table so that we can develop a system, so the state can develop a system, that will be easily used and interface with our jail operation and avoid some of the problems that have been identified. So the resolution simply asks that the Metro Technology Department be involved in having a seat at the table. [DE 50 at 1232 (Louisville/Jefferson County Inclusion in Development of New Case Management System, Discussion of Resolution During October 11, 2018 Meeting of Louisville Metro Council (statements by council members, 36:33-38:17))]. After Hollander’s remarks, Councilwoman Denton asked him: “I’m just curious to know if anyone from the Mayor’s office or from Corrections or anywhere in Metro has contacted the state to request to be included.” Id. at 38:28-38:39. Hollander responded: And I don’t know the answer to that . . . I think, and I don’t want overstate this, but I think it is a fair, a fair reading of the audit results that there has not been a lot of cooperation between various players here, for a number of years, and I’m trying to be diplomatic about this, and I’m not ascribing any blame to anybody for that, but I think . . . the point of the resolution is to say, it would be a good idea if there was that kind of discussion going forward. Id. at 39:03-39:35. During the meeting, Councilman Brent Ackerson discussed Metro Government’s lack of a response to this “problem that has been ripening for years”: I’ve heard a lot of political correctness here today and I’m going to play the opposite role, as I usually do. I mean, this resolution solves a little problem that is: please invite us to the table. But, it doesn’t solve the bigger problem. The bigger problem is, why haven’t we already asked and it shouldn’t be this body’s responsibility, it should be the administration’s responsibility. I sat through that hearing and they talked about you know, 30-plus mistakes a day, they talked about the cost, there is a huge cost involved here, that cost isn’t really boiled down to dollars, it boils down to manpower. We have a 100% turnover rate . . . within seven months on average of the clerks handling these matters because the system is in such chaos . . . This is not something that came out of nowhere. 7 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 8 of 50 PageID #: 2094 This has been a problem that has been ripening for years. And the fact of the matter is, that this city hasn’t asked for a seat at the table. We are the number one stakeholder here. I think we handle, the closest county is Fayette County and they handle half the volume that we handle through our jail system. And so there are no surprises here. So, while this resolution may solve the problem of the lack of communication coming from the city at this very moment, it does not solve the greater problem of why hasn’t that already been initiated without this body requesting it to initiate it . . . So the heck with political correctness, let’s shame some people into doing the right thing and get on with business. Id. at 44:40-46:30. 2. The timeliness of LMDC’s release of inmates During discovery, Plaintiffs asked Defendants to “identify by name and last known address and telephone number all persons at [LMDC] from February 3, 2016 to the present who were: a. Imprisoned, detained, or incarcerated longer than ordered by Courts of the Commonwealth of Kentucky.” [DE 50 at 1223]. In response, Defendants provided Plaintiffs with two spreadsheets. Id. The first spreadsheet “listed . . . 29 late releases from LMDC’s mai[n] jail for the time period February 3, 2016 through the date of production, August 17, 2018.” [DE 50 at 1223]. In the first spreadsheet, LMDC listed the “causes of delay” as: “Records had lack of staff, was behind in paperwork” (1 day); “Records misplaced the paperwork” (2 days); “Records failed to release” (2 days, 3 days, 4 days); “Records failed to process the release from custody” (3 days, 5 days, 5 days, 26 days); “File folder was misplaced in records” (2 days, 2 days, 2 days, 5 days, 5 days); “Bail credit had not been applied” (1 day, 1 day, 9 days); “Good time credit was not applied to sentence” (1 day, 5 days); “Records miscalculated sentence credit” (10 days, 10 days); “Records incorrectly calculated sentence” (12 days); “Unknown records error” (15 days); and, “Charges were amended down, but were not changed in records” (142 days). Id. at 1224. Defendants’ second spreadsheet “list[ed] late releases from LMDC’s Home Incarceration Program (‘HIP’) for the same relevant time period. The spreadsheet listed only 14 late releases 8 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 9 of 50 PageID #: 2095 from HIP for the same time period.” Id. (emphasis in original). These fourteen late releases were caused by, among other things: “[G]ood time credit not applied to sentence” (32 days, 4 days, 4 days, 4 days), “Records did not place in computer nor the time out book” (22 days) and “HIP never received release from records” (13 days, 13 days). Id. Of the 43 inmates LMDC identified, 39 (or 92%) of them were not timely released seemingly because of errors originating in LMDC’s records department. [See DE 50-2 and DE 50-3]. The remaining four inmates were not released for these reasons: “Xjail error,” “Citation had wrong inmate information,” “Unknown records error,” and “Div 6 clerk had not sent information to modify conditions of probation.” [DE 50-2 at 1258-1261 (emphasis added)]. Thus, of the 43 inmates LMDC identified, only one was not timely released because of an issue with a court order. Id. Plaintiffs next asked Defendants to provide the following information for all “persons who were released from Louisville Metro Department of Corrections (‘LMDC’) for the time period between February 3, 2016 to the present”: name; address; date and time a court ordered inmate to be released from LMDC; and date and time LMDC released inmate.3 [DE 50 at 1224-1225]. LMDC objected to this discovery request: RESPONSE: Objection. This response is overly broad and unduly burdensome. The Defendants are not required to place this information in a spreadsheet as part of the discovery request and reject that request. The Defendants[’] attempt to comply with discovery discovered that a report for just one month generated a 304 page report with 2,949 releases. Plaintiffs are requesting information compiled from each release for a period of about 30 months. 30 months times 3000 releases each month = 90,000 files. LMDC personnel spent 20 minutes pulling the information for 2 pages of the 304 pages. It would take approximately 55 hours to complete just 304 pages. It is estimated that the time it would take to review 30 LMDC is required to maintain a “written record” of when an inmate is released. 501 Ky. Admin. Regs. 3:120(3) (“A written record shall be kept of the time, purpose, date, and authority for release or removal from confinement, and into whose custody the prisoner is released or removed”). 3 9 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 10 of 50 PageID #: 2096 months’ worth of releases; 30 months x 304 pages/month x 55 hours = over 500 thousand hours. Id. at 1225. Because Defendants claimed “that it would take them over 500,000 hours (or 42.8 years) to identify all persons’ release times that they had ‘released’ from February 3, 2016 to the present, Plaintiffs . . . requested to view for themselves LMDC’s inmate release records for two different time periods.” Id. Defendants agreed and Plaintiffs reviewed two sample sets of inmate records: one from February 3, 2016 through March 31, 2016 [DE 50-5], and the other from June 1, 2017 through July 31, 2017 [DE 50-6]. Id. Plaintiffs documented their findings in two spreadsheets.4 Based on the review, Plaintiffs identified thousands of inmates held for more than four hours after LMDC’s receipt of their release orders and hundreds of inmates held for more than twelve hours after they had completed service of their sentence. [Compare DE 50-5; 50-6 with DE 50-10 at 1629 (“In total, LMDC provided information on 16 release errors during 2016, 4 during 2017, and 17 during 2018”)]. Yet, of the thousands of inmates not timely released, none of their inmate files contained an incident report explaining why. [DE 69 at 2044]. 3. Defendants’ knowledge about alleged issues with its process Plaintiffs assert that the deposition testimony of Bolton, Al-Amin, Clark, and Sharonda Simmons (“Simmons”) reveals what Defendants knew about defects in LMDC’s process and how they attempted (or did not attempt) to fix them. [DE 50 at 1232]. a. Bolton’s Deposition Testimony Bolton was the Director of LMDC from 2008 until summer of 2019. Plaintiffs deposed him in April 2019. [DE 49-3 at 699]. During his deposition, Bolton testified that he has known 4 Jennifer Moss, a paralegal for Plaintiffs’ counsel, reviewed the files. [See DE 50-4]. 10 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 11 of 50 PageID #: 2097 since at least 2014 about issues with LMDC’s non-automated system for processing releases: Q. Were you aware of the issues with lack of automation and – A. Oh, of course. Yes. Q. —the problems that could be encountered in forming queries? Were you—had you been aware of that beforehand? A. Yeah. Sorry to interrupt you. Absolutely. Q. Now, how long have you been aware of that? A. For quite a while. Q. But would you say two years, three years, five years? A. Five years plus. Id. at 721. Bolton testified about his “efforts” “to automate the records department’s operations to facilitate and make that process more efficient”: Q. Okay. Once your records department receives an order to release an inmate, whose responsibility is it to implement that order? A. The records department personnel. ... Q. Okay. So my question is, from the time that the order comes into your records department and the time that the inmate is released, the process that occurs between those two periods, has that been implemented—has that been automated by the LMDC to any degree during your tenure? A. I—I—I don’t know. I don’t have the answer to that. Q. Have you ever told anybody, or suggested to anybody, or assigned anyone to look at how that process might be automated in order to increase its efficiency and its accuracy? A. Yeah. I think I’ve got a little bit clearer understanding of what you’re 11 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 12 of 50 PageID #: 2098 asking . . . So we would meet regularly with the Clerk’s Office and knowing that there was no budget to create automation in the Clerk’s Office . . . so there . . . was much discussion over the years as to how we might more effectively and efficiently move paper and create a—an electronic footprint for tracking purposes if we needed to go back and track. And that recommendation came out of Corrections . . . And that was to utilize email to move court orders, and that’s what we did. And that . . . helped a great deal, but it was certainly not a . . . permanent, best practice solution. It was a solution . . . or it was . . . a process that was better than what we had. ... Q. Okay. This wasn’t a particular automation feature that was put into place. People got e-mail, and they started communicating via e-mail as opposed to a clerk walking over to the records department or mailing something or faxing something. A. Correct. Q. Correct? A. Yeah. Id. at 724-728. Bolton also testified about CGL’s findings: Q. And—but what I—what’s interesting to me is the bottom line. A. Uh-huh. Q. “Percent requiring clarification.” A. Right. Q. March 26 through April 9, 2018, the number—or the percent of the orders received by your records department that required clarification was 6.3 percent. A. Right. Q. Okay? I take it that, conversely, that means there was 92.7 percent that did not need clarification? 12 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 13 of 50 PageID #: 2099 A. I don’t know. Q. September 4 through 14, 2018, the number of orders needing clarification was 2.5 percent. A. Uh-huh. Q. Did you—do you understand that to mean that the percentage of orders that did not need clarification was 97.5 percent? A. I—I—I—I would think so. Yeah. Q. Okay. Now, what correlation . . . have you made between overdetentions and court orders requiring clarification? Are all of the overdetentions a result of unclear court orders, or are some of these overdetentions occurring despite clear court orders? A. I don’t know. Q. Has anybody made an examination of that question? A. I don’t know the answer to that. Q. Well, we’ve spent a long time here talking about problems with the court orders— A. Right. Q. —in the context of this over-detentions case. But your testimony is you don’t even know whether that’s a problem when it comes to overdetentions or not? A. I don’t know specifically. I certainly would suspect that it can be. Id. at 743-745. b. Al-Amin’s Deposition Testimony Al-Amin is Coordinator of LMDC’s records department. [DE 49-1 at 386]. She reports to Clark. Id. at 387. Plaintiffs deposed Al-Amin twice. [DE 50 at 1232]. They first deposed her in September 2018. Id. Al-Amin’s job is to “make sure that that the daily operations . . . are carried out properly.” [DE 49-1 at 386]. LMDC’s “goal” is to process release orders within two 13 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 14 of 50 PageID #: 2100 to four hours of receipt. Id. at 427. That said, LMDC does not have a “system” for prioritizing the oldest orders first. Id. at 426. In fact, Al-Amin testified that orders are not prioritized based on when LMDC receives them: Q. Now, do they then go through the releases and prioritize the releases themselves, such that the oldest order is at the top and the—the most recent one is at the bottom of the release stack? A. No. Q. And why not? A. They’re just getting them processed. I mean, they’re—they’re all going to be processed in—during that time because they’re not going to be placed in a basket for somebody else to process because they was processed on that shift—I mean, they were printed on that shift during that time so they’re processing them. Id. at 417. Al-Amin described the process for releasing an inmate who is being held in LMDC, but not serving a sentence (e.g., facing pending charges): A. Okay. Records, when they receive a release on an inmate, the release, of course, is going to come from the court or it’s going to come from pretrial but it has to come from a judge. Records will review the—the release paperwork and make sure that all of the charges and case numbers are covered. If they are, then they’ll process the release, okay. And then before they—we don’t actually release the person out the door, okay. So we’ll have to send—once we do our part in records, then we’ll send the paperwork over to intake and release. Intake and release conducts a warrant check. So they’ll check eWarrants to see if any outstanding warrants exist and they also check with NCIC to see if there's anything there. Q. Okay. And, if there’s not, then what happens? A. Then intake and release will go ahead and—and they’ll send the release —the booking copy. Id. at 392-393. 14 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 15 of 50 PageID #: 2101 Al-Amin also described the process for releasing an inmate who is serving a sentence: Q. When you get instead a sentence— A. Uh-huh. Q. —are you then subsequently told by the court when that person should —is supposed to be released or is the records division essentially expected to calculate the release date on its own from the sentence? A. It can go both ways because the judge can indicate when a person is to be released, okay, on an order of commitment. Q. Okay. A. And, basically, all we have to do then is—is just put the release date in the computer. I mean, there’s really no sentence to calculate because the judge has already given us that release date. With the sentences, however, we do have to calculate those sentences. Q. And who does that? A. The—the senior corrections technician. Q. Okay. So, when he gets the sentence order, he—he calculates the release date when he gets the sentence order? A. Right. Q. And then what happens with that release date, is that—is there a flag somewhere in the system that—that comes back and reminds you all that, okay, he’s completed his—his sentence and today’s the day he needs to be released? A. Yeah. Actually, once we’ve calculated the sentence, okay, the SCT will indicate the release date on the order and they’ll place their initials on there. Then it—that information’s transferred to what we call a time out book. The time out book has all the dates for that year, okay, and we’ll go to that date and we’ll put a sticker that includes the inmate’s name, CIN number, and number of days that the person is sentenced to on the date that he's due to be released, he or she's due to be released. Okay. So you can have that method. And then, we also calculate it in XJail, that’s our jail system, jail management system. 15 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 16 of 50 PageID #: 2102 ... A. On that date that the inmate is . . . due to be released from custody, third shift SCTs will print out—well, actually second shift prints out the time out list and gets it ready for third shift. Q. Okay. A. So any inmates that’s [sic] scheduled to be released for that day, that night, the SCT prints the list out for the following day if that makes, did that make sense? Q. Okay. A. And then when third shift comes on, then they’ll take that time out list and they’ll compare that time out list with the names that’s [sic] in the time out book to make sure that all of those names they mirror one another. Q. All right. A. And they’ll have to go through all of those names to see if the inmate is still eligible to be released from custody or how they—if they have other charges pending, then they cannot get out. And we’ll make a note in the time out book why the inmate cannot be released from custody. Q. Okay. A. Okay. Id. at 408-410. Beginning in 2012 or 2013, Bolton at times expressed concern to Al-Amin about “late releases” and told her that “we need to do what we need to do to prevent this from happening.” Id. at 532. But Bolton neither asked Al-Amin for nor provided her with suggestions—other than an “incentive” program that did not work—for how to prevent late releases: Q. Had [Bolton] come to you before then and told you that you needed to do something to correct the late releases? A. Yeah. We had discussions about the late releases but what—there wasn’t a remedy or . . . anything put in place at that time. ... 16 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 17 of 50 PageID #: 2103 Q. And, on each of these occasions he would . . . express to you, we need . . . to do what we need to do to prevent this from happening, essentially? A. He would say something along that line. Q. Okay. But would he ever tell you what he believed needed to be done? A. No. Q. Would he ever ask you for your recommendations of what you believe could prevent it from happening? A. Not really, no. Id. at 531-533. Plaintiff deposed Al-Amin again in November 2018. [DE 50 at 1234]. During this deposition, they asked her questions about documents they received in discovery, specifically an October 9, 2017 email, sent nine months after Plaintiffs filed the lawsuit. Id. In the October 2017 email, Al-Amin wrote, in relevant part: I’ve been advised that we are requesting paperwork updates from the courts that they have already sent. I was informed that certain staff members are shredding or discarding the paperwork that is emailed from the court as “No bodies” and the paperwork is not being processed and/or the case number is being entered in XJail, but the paperwork is discarded and not placed inside the folder. All inmate paperwork coming from the courts is to be reviewed and processed according! [sic] I hope that this is not true. However, if it is please know that your actions will result [sic] severe consequences! [DE 50-12]. Al-Amin testified that she did not investigate who was “shredding or discarding” paperwork. [DE 49-2 at 615-616]. Nor did she discipline any LMDC employees for doing so. Id. c. Clark’s Deposition Testimony Clark was the Chief of Staff of LMDC from 2004 until spring of 2019. [DE 49-4 at 889]. 17 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 18 of 50 PageID #: 2104 Plaintiffs deposed Clark in April 2019. Id. at 871. Clark testified that Bolton never called a meeting to discuss the allegations in this suit against them. Id. at 891-892. He, like Al-Amin, testified that inmates should be released within four hours of a court order “all things being the same.” Id. at 896. Clark acknowledged that he has been aware since at least 2011 that issues with consistency between court orders resulted in over-detentions. Id. at 899. Clark testified that, in response to these issues, LMDC began receiving orders by email: Q. But you knew that court orders were inconsistent, correct? A. Yes. Q. And you’ve known about that for a long time, haven’t you? A. Yes. Q. Okay. And did you ever put together a committee or ask all your people to get together and come up with a solution to get this problem fixed with these court orders? A. Who would they get with? Q. You were chief of staff. I’m just asking if you ever did that? A. No, I didn’t do that. ... Q. Okay. Did Director Bolton ever say, “Hey, we’ve got to fix this problem. It’s affecting our jobs over here getting people released on time”? A. I do know that we did have monthly meetings with the clerk’s office. Now, how that began, I don’t know. ... Q. Okay. When did you begin having meetings? A. It had to be approximately 2011 or so. Q. So even back as far as 2011, there’s a problem? 18 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 19 of 50 PageID #: 2105 A. Yes. Q. Okay. And what became of these meetings? A. Some issues got fixed, some didn’t. Q. And what issues got fixed? A. Just for example . . . the way that the clerk’s office transferred court orders and court documents to us. Q. And? A. Rather than carry them over, we instituted through—with them an emailing system. ... Q. Okay. What else happened? You said some things weren’t fixed? A. Yeah, some things weren’t. Q. What didn’t get fixed? A. At this point, as I sit here, I can’t tell you what didn’t get fixed. Id. at 898-900. Clark believed that, at most, LMDC over-detains ten inmates a year. Id. at 933-934 (“[T]en people a year that gets over incarcerated. One is too many, don’t get me wrong, but ten people? That isn’t the issue. It isn’t this big issue as if [sic] you’re portraying to me. It’s not”). d. Simmons’ Deposition Testimony Simmons is the First Shift Supervisor in LMDC’s records department. [DE 49-6 at 1076]. Plaintiffs deposed her in March 2019. Id. at 1071. Simmons began working at LMDC in 2001 and was promoted to supervisor in 2011. Id. at 1076. She never participated in any discussions with LMDC management about trying to reduce the overall number of late releases. Id. at 1096. 19 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 20 of 50 PageID #: 2106 She mostly learned about over-detentions through outside sources, such as attorneys or family members who called to complain. Id. at 1103. The internal auditors made a series of recommendations in their final report, [DE 50-9 at 1602-1616], and Simmons testified about whether Defendants implemented them: Q. Now, let me ask you this: [The final report] came out in January of 2018, so about 14 months ago. A. Okay. Q. In that period of time . . . Are you aware of anything that’s been done to develop or make changes to streamline the method by which the courts and LMDC share essential information at the time the information is needed or in a format that's useful? Has there been any change in that? A. They’re working on it, as I previously stated earlier. Q. Okay. There’s not been a change, but they’re working on it? A. Uh-huh. Q. Okay. A. Yes. Q. Are you aware of any change that’s been made to improve LMDC’s and the court’s collaboration in developing court orders that are notated in a clear, consistent manner? A. The same thing, they working on them. Q. They’re working on it, but it’s not done? A. Right. Q. Same question as to whether there’s been any changes toward development and implementation of a protocol for periodic communication between your department and the courts. Same answer? A. Yeah. Q. Are you aware of anything that’s being done to improve collaboration 20 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 21 of 50 PageID #: 2107 between your department and the courts to effectively identify court orders that require special consideration? A. No. Q. You aware of anything that’s being done to develop a method to track and monitor court orders from the time of receipt through completion of processing? ... A. Not that I know of. ... Q. Okay . . . To your knowledge, has there been any effort undertaken to change the XJail system or significantly expand its functionality to help the records department do its job? A. Not that I know. I don’t know. [DE 49-6 at 1118-1121]. II. STANDARD The Court has broad discretion in deciding whether to grant class certification. Sterling v. Velsicol Chem. Corp., 855 F.2d 1188, 1197 (6th Cir. 1988). That said, “it must exercise that discretion within the framework of Rule 23.” Coleman v. General Motors Acceptance Corp., 296 F.3d 443, 446 (6th Cir. 2002). Under Rule 23 the moving party bears the burden of proving that the proposed class both satisfies Rule 23(a)’s requirements and fits into one of Rule 23(b)’s three subdivisions. Coleman, 296 F.3d at 446; Fed. R. Civ. P. 23(b); see also Ball v. Union Carbide Corp., 385 F.3d 713, 727 (6th Cir. 2004). “Rule 23 does not set forth a mere pleading standard. A party seeking class certification must affirmatively demonstrate his compliance with the Rule.” Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 350 (2011). Rule 23(a) has four requirements, which all must be satisfied: 21 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 22 of 50 PageID #: 2108 (1) the class is so numerous that joinder of all members is impracticable; (2) there are questions of law or fact common to the class; (3) the claims or defenses of the representative parties are typical of the claims or defenses of the class; and (4) the representative parties will fairly and adequately protect the interests of the class. Fed. R. Civ. P. 23(a). The Court’s Rule 23 analysis must be “rigorous.” Dukes, 564 U.S. at 350-51. “Frequently that ‘rigorous analysis’ will entail some overlap with the merits of the plaintiff's underlying claim.” Id. And “sometimes it may be necessary for the court to probe behind the pleadings before coming to rest on the certification question.” Gen. Tel. Co. of the Sw. v. Falcon, 457 U.S. 147, 160, (1982). That said, “the court should not turn the class certification proceedings into a dress rehearsal for the trial on the merits.” Glazer v. Whirlpool Corp. (In re Whirlpool Corp. Front-Loading Washer Prods. Liab. Litig.), 678 F.3d 409, 417-18 (6th Cir. 2012) (quoting Messner v. Northshore Univ. HealthSystem, 669 F.3d 802, 811 (7th Cir. 2012)). Thus, the court’s Rule 23 analysis must not include consideration of whether plaintiffs will ultimately prevail on the merits of their claims. Id. at 418 (citing Gariety v. Grant Thornton, LLP, 368 F.3d 356, 366 (4th Cir. 2004). III. DISCUSSION Based on the Court’s review and the parties’ briefs, the Sixth Circuit has not yet considered whether this type of over-detention class is certifiable. Thus, before turning to Plaintiffs’ proposed class definition and the Rule 23 factors, the Court considers the relevant precedent on overdetention class certification from various circuits as it informs the Court’s reasoning throughout this Opinion, especially as it pertains to Rule 23’s commonality and predominance requirements. 22 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 23 of 50 PageID #: 2109 A. Relevant Precedent 1. Supreme Court: McLaughlin In McLaughlin, the plaintiffs brought a class action under 42 U.S.C. § 1983 “challenging the manner in which the County of Riverside . . . provides probable cause determinations to persons arrested without a warrant.” Cty. of Riverside v. McLaughlin, 500 U.S. 44, 47 (1991). At issue in McLaughlin was the “County’s policy of combining probable cause determinations with its arraignment procedures . . . Under County policy, which tracks closely the provisions of Cal. Penal Code Ann. § 825 (West 1985), arraignments must be conducted without unnecessary delay and, in any event, within two days of arrest.” Id. After finding that the plaintiffs had standing to sue, the Court discussed Gerstein v. Pugh, 420 U.S. 103 (1975). In Gerstein, the Court “held that the Fourth Amendment requires a prompt judicial determination of probable cause as a prerequisite to an extended pretrial detention following a warrantless arrest.” Id. at 44. In reaching its holding in Gerstein, the Court sought to “reconcile important competing interests.” Id. at 52. On one hand, “States have a strong interest in protecting public safety by taking into custody those persons who are reasonably suspected of having engaged in criminal activity, even where there has been no opportunity for a prior judicial determination of probable cause.” Id. And, on the other, “prolonged detention based on incorrect or unfounded suspicion may unjustly ‘imperil [a] suspect’s job, interrupt his source of income, and impair his family relationships.’” Id. (quoting Gerstein, 420 U.S. at 114). The Court acknowledged, however, that some “delays are inevitable”: For example, where, as in Riverside County, the probable cause determination is combined with arraignment, there will be delays caused by paperwork and logistical problems. Records will have to be reviewed, charging documents drafted, appearance of counsel arranged, and appropriate bail determined. On weekends, when the number of arrests is often higher and available resources tend to be 23 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 24 of 50 PageID #: 2110 limited, arraignments may get pushed back even further. In our view, the Fourth Amendment permits a reasonable postponement of a probable cause determination while the police cope with the everyday problems of processing suspects through an overly burdened criminal justice system. Id. at 55. Yet, “flexibility has its limits,” “Gerstein is not a blank check,” and a “State has no legitimate interest in detaining for extended periods individuals who have been arrested without probable cause.” Id. As a result, the Court held “that a jurisdiction that provides judicial determinations of probable cause within 48 hours of arrest will, as a general matter, comply with the promptness requirement of Gerstein. For this reason, such jurisdictions will be immune from systemic challenges.” Id. at 56. But the Court added an important caveat: This is not to say that the probable cause determination in a particular case passes constitutional muster simply because it is provided within 48 hours. Such a hearing may nonetheless violate Gerstein if the arrested individual can prove that his or her probable cause determination was delayed unreasonably. Examples of unreasonable delay are delays for the purpose of gathering additional evidence to justify the arrest, a delay motivated by ill will against the arrested individual, or delay for delay’s sake. In evaluating whether the delay in a particular case is unreasonable, however, courts must allow a substantial degree of flexibility. Courts cannot ignore the often unavoidable delays in transporting arrested persons from one facility to another, handling late-night bookings where no magistrate is readily available, obtaining the presence of an arresting officer who may be busy processing other suspects or securing the premises of an arrest, and other practical realities. Id. at 56–57 (emphasis added). 2. The Ninth Circuit: Brass and Berry5 In Brass, the plaintiff was held for thirty-nine hours after the court ordered his release. 328 F.3d at 1194. He challenged the County’s policy of “process[ing] releases for prisoners pursuant The plaintiffs in Brass and Berry did not seek certification of over-detention classes. Yet Brass and Berry are instructive because they analyze §1983 claims premised on inadequate and inefficient release policies and procedures. 5 24 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 25 of 50 PageID #: 2111 to court order only after the completion of processing all inmates scheduled for release on that day.” Id. at 1198. He also challenged the County’s practice of “beginning the processing of releases only after all information relating to prisoners scheduled for release on a particular day had been received and entered into the computer system.” Id. The Brass panel affirmed the district court’s grant of summary judgment to the County. Id. at 1994. Although the panel recognized that over-detention after court-order release may violate constitutional rights, it found that the contested policies did not because the County presented compelling information about why its administrative processing system was reasonable. Id. at 1200-1201. The three plaintiffs in Berry sued the Los Angeles County Sheriff, in his official capacity, under 42 U.S.C. § 1983, for violating their Fourth and Fourteenth Amendment rights. 379 F.3d 764, 767 (9th Cir. 2004). They alleged that the “County’s unreasonably inefficient implementation of its administrative policies amounts to a policy of deliberate indifference to their constitutional rights.” Id. at 768. More specifically, they alleged that the County adopted a policy of inaction. Id. at 767. Relying on Brass, the district court granted summary judgment because “the County’s challenged policies did not result in a violation of the plaintiffs’ constitutional rights.” Id. (internal quotation marks omitted). The Ninth Circuit reversed. Id. at 766. It first noted that Brass was distinguishable because there the plaintiffs challenged specific procedures, but in Berry they were challenging the implementation of the County’s policies. Id. at 768. The Ninth Circuit found that “the County’s system of administrative processing cannot be immune from allegations that, in practice, it amounts to a policy of deliberate indifference.” Id. Although the County “submitted declarations from various county employees describing the administrative processing system and explaining why it reasonably takes twenty-four to forty-eight hours to process each release,” the Ninth Circuit 25 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 26 of 50 PageID #: 2112 denied summary judgment because there was a genuine issue of material fact about whether the “County’s implementation of its policies is in fact reasonably efficient.” Id. at 770. 3. The Third Circuit: Wharton In Wharton, plaintiffs moved to certify a class of: (a) Each person who has been, is, or in the future will be incarcerated in any Delaware prison from October 1, 2008, forward; and (b) who was not released, or, in the future, will not be released within 12 hours of the time that a court order has been forwarded to [Central Offender Records (“COR”) ] releasing the person, or is not released by midnight of the day his or her sentence has ended. Wharton v. Coupe, No. CV 12-1240-LPS, 2015 WL 5768936, at *1 (D. Del. Sept. 30, 2015), aff'd sub nom. Wharton v. Danberg, 854 F.3d 234 (3d Cir. 2017). The Wharton plaintiffs argued that they were over-detained because COR was “badly broken, causing or allowing the over-detention of as many as a thousand inmates a year.” 854 F.3d at 238. The plaintiffs identified two flaws with COR: 1) it is “under-staffed generally and particularly short-handed on nights and weekends”; and 2) “COR is unresponsive to inmates and those acting on their behalf . . . Prisoners’ families, friends, and bail bondsmen can call COR directly, but generally complain that COR is frequently unhelpful or indifferent, when it can be reached at all. Outside input allegedly falls on deaf ears.” Id. at 239. Affirming the district court’s grant of summary judgment, the Third Circuit noted the limitations of plaintiffs’ evidence of overdetention. Instead of statistical evidence, plaintiffs submitted “a disparate and somewhat disjointed assortment of affidavits from several witnesses whose work brings them in close contact with the correctional system. These affidavits . . . reported huge numbers of over-detentions, albeit in an impressionistic fashion based on the affiants’ own personal observations and estimates.” Id. at 238. The Third Circuit found that plaintiffs failed to establish a genuine issue of material fact about defendants’ deliberate indifference: 26 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 27 of 50 PageID #: 2113 As the District Court held, the record shows a variety of efforts by Appellees to improve COR and address the over-detention problem. For example, not only did McBride increase staffing levels at COR, she offered uncontradicted testimony that she did so specifically in response to delays in processing. Likewise, her efforts to improve the agency’s training system, to upgrade its technology, and to create special units to more efficiently handle certain types of release show, as she testified, that COR leadership was ‘always looking at ways to be more efficient.’ Id. at 243. The Third Circuit suggested that the plaintiffs’ deliberate indifference claim may have been strong enough to survive summary judgment if: “the crux of the evidence presented to the District Court had been that closed channels of communication caused particularly large numbers of overdetentions; that an alternative system had been presented to COR or was a best practice they should have known to adopt; or that changes to COR’s communications policy would have been easier or more efficacious than COR’s other reform efforts.” Id. at 245. But perhaps most damaging to plaintiffs’ case was the “substantial and uncontested record evidence that many overdetentions originate in the court system rather than at COR . . . To survive summary judgment, Appellants need more than speculation connecting any increase in overdetentions with the COR policies they deem ineffective.” Id. at 246. 4. The Seventh Circuit: Harper, Portis, and Driver In Harper, the plaintiff claimed that the Sheriff of Cook County was unconstitutionally holding new detainees after bond had been posted. Harper v. Sheriff of Cook Cty., 581 F.3d 511, 512 (7th Cir. 2009). The plaintiff successfully moved the district court to certify a class of “[a]ll persons processed into the Cook County Jail on and after May 2, 2005 while that person, or someone acting on his (or her) behalf sought to post cash bond.” Id. at 513. On appeal, the Seventh Circuit vacated the district court’s class certification. Id. at 512. The Seventh Circuit reasoned that the constitutionality of the detentions at issue depended on 27 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 28 of 50 PageID #: 2114 whether the length of the delay before the detainee’s release was reasonable in any given case, which, in turn, depended on “how long each detainee was held after bond was posted and what justifications there might be for the delay on that particular day or for that particular detainee.” Id. at 515. The Seventh Circuit noted various justifications that may exist for any given delay, including “the time of day, whether the jail was processing an unusually large number of detainees at that time, whether other events occurring at the jail legitimately slowed processing times, whether the detainee’s lack of cooperation delayed processing, etc.” Id. As a result, the Seventh Circuit determined that “[l]iability, to saying nothing of damages, would need to be determined on an individual basis. Thus, common issues do not predominate over individual issues, making this case inappropriate for class disposition.” Id. Likewise, in Portis, the Seventh Circuit reversed the district court’s decision to certify a class in a case challenging the Chicago Police Department’s allegedly unconstitutional detention of individuals arrested for non-jailable offenses. Portis v. City of Chicago, Ill., 613 F.3d 702, 703 (7th Cir. 2010). The plaintiffs in that case asserted that taking more than two hours to process and release those individuals made their detention unreasonable in violation of the Fourth Amendment. Id. at 703. The Seventh Circuit held that the district court erred in prescribing a two-hour limit to this process and, as a result, erred in certifying a class premised on the propriety of that limitation. Id. at 705. The Seventh Circuit concluded that “[b]ecause reasonableness is a standard rather than a rule, and because one detainee’s circumstances differs from another’s, common questions do not predominate and class certification [was] inappropriate.” Id. That said, the Seventh Circuit left open the possibility “that class treatment might be appropriate if the class sought to establish that a jurisdiction had adopted a policy of deliberate delay.” Id. In Driver, a group of inmates sued the Marion County Jail claiming the sheriff implemented 28 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 29 of 50 PageID #: 2115 release practices and procedures that caused their continued detention after they became eligible for release. Driver v. Marion Cty. Sheriff, 859 F.3d 489, 490 (7th Cir. 2017). The district court denied two of the plaintiffs’ proposed classes: [A]ll individuals who, from December 19, 2012 to the present, were held in confinement by the Sheriff after legal authority for those detentions ceased, due to: (1) the Sheriff’s practice of operating under a standard of allowing up to 72 hours to release prisoners who are ordered released; and (2) the Sheriff’s practice of employing a computer system inadequate for the purposes intended with respect to the timely release of prisoners. Id. Vacating the judgement of the district court, the Seventh Circuit held that class treatment was appropriate because the plaintiffs pointed to a specific “policy or practice [that] caused them to be detained for an unconstitutionally-unreasonable length of time.” Id. at 492. In doing so, the court distinguished Harper and Portis: The only common issue alleged by Harper was whether it was reasonable to assign a jail identification number before releasing a detainee on bond—with the concomitant delay that process entailed—but we held that the issue was not central to his claim because it could not cause the type of injuries asserted, and because it could not be unconstitutional unless it took an unreasonable amount of time in an individual case, thus again not supporting class disposition. Id. at 515–16. Similarly, in Portis, 613 F.3d at 703, we addressed a class challenge by persons subjected to custodial arrests for fine-only offenses, alleging that the failure to release them within two hours from generation of the central booking number was unreasonable in violation of the Fourth Amendment. We held that the Constitution forbids detentions that are unreasonable in length, but that an arbitrary inflexible time period such as two hours for release was not justified. Id. at 704. Accordingly, the claims required an individualized determination as to whether the delay in release was unreasonable, and class certification was improper. Id. The Seventh Circuit’s analysis of the over-detention sub-class based on the Sheriff’s use of an inadequate computer system is of particular relevance to Plaintiffs’ claims in this case and 29 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 30 of 50 PageID #: 2116 merits additional discussion. The plaintiffs in Driver presented evidence that the Sheriff “acquired a computer system, OMS, from Global Tel Link which offered the system to the Sheriff free of charge in return for the Sheriff’s extension of GTL’s contract for inmate collect call services.” Id. at 493. The plaintiffs further produced evidence that the Sheriff “bypassed the standard review process in choosing and retaining the computer system, disregarding the impact on the release times.” Id. OMS was “beset with technical issues from the start; it could not interface with DEXTER, the computerized transfer system that allowed agencies such as the Sheriff, the Public Defender, the Prosecutor’s Office, Community Corrections, and the Indianapolis Police Department to exchange information with each other on Odyssey.” Id. at 493-494. As a result, the Sheriff could not receive electronic court information and the Sheriff’s office was required to “manually update and process the codes received from the court and to rely on emails, paper records, faxes, and telephone calls to gather information to make release decisions.” Id. at 494. Despite “significant delays in release times,” the Sheriff continued to use the OMS system and did not take “efforts to measure the magnitude of the problem.” Id. The plaintiffs presented evidence that from “June to December 2014, 38,000 extra days were spent in jail by inmates pending release compared to the practice before OMS.” Id. The delays “resulted in complaints to the Sheriff’s department by judges, defense counsel and family members.” Id. But, despite knowing about “widespread excessive delays,” the Sheriff chose to continue “with OMS rather than implement the compatible system recommended by its IT people.” Id. Based on the plaintiffs’ evidence, the Seventh Circuit found that the proposed sub-class could be properly certified: Given the evidence of a dramatic increase in detention times in correlation with the implementation of the computer system, the evidence that release information could not be properly transmitted electronically in that system and the court having 30 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 31 of 50 PageID #: 2117 to resort to alternative channels of communication, and the absence of any evidence that delays of that length could be attributable to individual factors, the class is capable of definition both by the timing and the length of the delay in release. Id. B. Synthesis of Precedent From its survey of the relevant precedent on this issue, the Court can distill four general principles applicable to this case. First, courts have been reluctant to certify classes based on whether the amount of delay is unreasonable. See Harper, 581 F.3d at 515; Portis, 613 F.3d at 705. This theory of over-detention liability is generally not amenable to class treatment because determining whether a given detention rises to the level of a constitutional violation requires individualized determinations about the reason the inmate’s release was delayed. Second, courts have been willing to certify over-detention classes where the class can answer questions about whether a municipality’s policy (or lack of one) for releasing inmates caused over-detentions. Put differently, an over-detention class may be certified if it alleges a systemic issue—which impacts all inmates—with the municipality’s release process. See Driver, 859 F.3d at 494; see also Berry, 379 F.3d at 768 (“They claim that the County’s unreasonably inefficient implementation of its administrative policies amounts to a policy of deliberate indifference . . . As a matter of law, the County’s system of administrative processing cannot be immune [from such allegations]”). Third, courts that have certified over-detention classes have done so based on statistical, not anecdotal, evidence. See Driver, 859 F.3d at 494; see also Wharton, 854 F.3d at 238. And fourth, the class definitions of over-detention classes premised on a municipal policy, practice, or custom should contain the cause of the over-detention. See Driver, 859 F.3d at 494. These general principles are largely distilled from the trilogy of cases from the Seventh 31 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 32 of 50 PageID #: 2118 Circuit. As a result, the Court finds persuasive the Seventh Circuit’s reasoning in Harper, Portis, and Driver. The Seventh Circuit has carefully considered and explored the contours of overdetention classes in three separate published opinions. Of most assistance to this Court, the Seventh Circuit has provided examples of the types of over-detention claims that are certifiable and those that are not. With these general principles in mind, the Court turns to the proposed class definition and the Rule 23 factors. C. Class Definition “Although not specifically mentioned in [Rule 23], the definition of the class is an essential prerequisite to maintaining a class action.” Adams v. Fed. Materials Co., 2006 WL 3772065 at *3 (W.D.Ky.2006) (quoting Gevedon v. Purdue Pharma, 212 F.R.D. 333, 335 (E.D.Ky.2002)); see also Bentley v. Honeywell Int’l, Inc., 223 F.R.D. 471, 477 (S.D.Ohio 2004) (“Before delving into the ‘rigorous analysis’ required by Rule 23, a court first should consider whether a precisely defined class exists and whether the named plaintiffs are members of the proposed class”). At a minimum, the description must be “sufficiently definite that it is administratively feasible for the court to determine whether a particular individual is a member.” 7A Charles Alan Wright, Arthur R. Miller, & Mary Kay Kane, Federal Practice and Procedure, § 1760, at 120–21 (2d ed.1986). The Court’s duty is to ensure the proper exercise of Rule 23. That duty includes the obligation to create a new definition sua sponte if the parties’ own proposals are not adequate or accurate. See Powers v. Hamilton Cty. Pub. Def. Comm’n, 501 F.3d 592, 619 (6th Cir. 2007) (“[C]ourts must be vigilant to ensure that a certified class is properly constituted. More to the point, district courts have broad discretion to modify class definitions”); see, e.g. Schorsch v. Hewlett–Packard Co., 417 F.3d 748, 750 (7th Cir. 2005) (“Litigants and judges regularly modify class definitions”); In re Monumental Life Ins. Co., 365 F.3d 408, 414 (5th Cir. 2004) (“District 32 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 33 of 50 PageID #: 2119 courts are permitted to limit or modify class definitions to provide the necessary precision.”). Rule 23(c) “empowers courts to define an appropriate class, whether by accepting the proposed class, limiting the class to certain issues, or creating subclasses.” Kensu v. Michigan Dep’t of Corr., No. 18-CV-10175, 2020 WL 1698662, at *9 (E.D. Mich. Apr. 8, 2020); see Fed. R. Civ. P. 23 (“When appropriate, a class may be divided into subclasses that are each treated as a class under this rule”); Randleman v. Fid. Nat. Title Ins. Co., 646 F.3d 347, 355 (6th Cir. 2011) (“District courts also have broad discretion in determining whether to . . . divide a class action into subclasses”). Plaintiffs request certification of a class of: All persons who, since February 3, 2016, were imprisoned, detained or incarcerated more than four hours after: (a) entry of an order directing their release; and/or (b) satisfaction of their term of incarceration set by prior court order. [DE 50 at 1222]. Detention, imprisonment, and incarceration are distinct, albeit related, concepts. Detention is the “act or instance of holding a person in custody.” Detention, Black’s Law Dictionary (11th ed. 2019); see also Pretrial Detention, Black’s Law Dictionary (11th ed. 2019) (“The holding of a defendant before trial on criminal charges either because the established bail could not be posted or because release was denied”). Incarceration is the “act or process of confining someone.” Incarceration, Black’s Law Dictionary (11th ed. 2019). And imprisonment is “[t]he act of confining a person, esp. in prison,” “[t]he quality, state, or condition of being confined,” and “the period during which a person is not at liberty.” Imprisonment, Black’s Law Dictionary (11th ed. 2019). In their proposed class definition, Plaintiffs distinguish between two types of inmates: 1) those who were held more than four hours after “satisfaction of their term of incarceration set by prior court order” and 2) those who were held more than four hours after “entry of an order 33 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 34 of 50 PageID #: 2120 directing their release.” [DE 50 at 1222]. The first type of inmate is “imprisoned”6 (i.e., Plaintiff Healey)—he or she is “confin[ed]” in LMDC and is serving a sentence “set by prior court order.” [DE 37 at 349]. On the other hand, the second type of inmate is “detained” (i.e., Plaintiffs Jarvis, Yates, and Melloan)—he or she is held in LMDC custody, but has been ordered released. Id. at 350. LMDC’s release procedure for imprisoned inmates appears to be different than its release procedure for detained ones. [Compare DE 49-1 at 392-393 with DE 49-1 at 408-410]. Defendants have not explained why releasing an imprisoned inmate who has completed a sentence should take longer than their goal of four hours. [DE 49-1 at 408-410]. Defendants argue that issues with unclear court orders—and not the Defendants’ deliberate indifference—is the cause of some late releases. [DE 67 at 1978]. But, because the order for an imprisoned inmate is received when the inmate is committed to serve his or her sentence, issues with unclear orders can be resolved long before the date he or she will be released. Id. at 408. For example, if an imprisoned inmate is sentenced to serve sixty days in LMDC, Defendants have sixty days to prepare for a timely release and clarify any issues with the court order. Furthermore, Defendants’ own data reveals that their inadequate process for releasing imprisoned inmates (and not issues with court orders) is the cause of a substantial majority (42 of 43, or 98%) of the late releases of imprisoned inmates. [See DE 50-2 and DE 50-3]. On the other hand, there is evidence that issues with interpreting court orders can delay release for detained inmates. [DE 50-10 at 1626]. While Defendants argue that they cannot control defects in the orders they receive, they concede that once they receive the orders it is their responsibility to timely implement them. [DE 49-3 at 724-725]. The Court acknowledges that Although “incarcerated” and “imprisoned” are nearly synonymous, “imprisoned” more precisely describes the condition of the first type of inmate in this case. 6 34 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 35 of 50 PageID #: 2121 Defendants process a large volume of orders each day (and year) and that sometimes there are delays in the system outside their control. [DE 50-10 at 1626]. But the Court also recognizes that inmates have a right to be released “promptly” after LMDC receives the release order. See McLaughlin, 500 U.S. at 55; Shorts v. Bartholomew, 255 F. App’x 46, 51 (6th Cir. 2007) (It “is beyond dispute: when a prisoner’s sentence has expired, he is entitled to release. Perhaps more to the point, an incarcerated inmate has a liberty interest in being released at the end of his term of imprisonment. This liberty interest is most often attributed to the Due Process Clause of the Fourteenth Amendment”) (internal citation and quotation marks omitted); Davis v. Hall, 375 F.3d 703, 713 (8th Cir. 2004) (“[E]ven a thirty-minute detention after being ordered released could work a violation of a prisoner’s constitutional rights under the Fourteenth Amendment”). As the Supreme Court noted in Gerstein, “[t]he consequences of prolonged detention may be more serious than the interference occasioned by arrest. Pretrial confinement may imperil the suspect’s job, interrupt his source of income, and impair his family relationships.” 420 U.S. at 114; see also DE 50-1 at 1255 (“The errors enumerated herein have caused the loss of employment, the loss of treatment opportunities, missed medical and mental health appointments, and they have placed Defendants and others at risk of harm”)]. Defendants sworn testimony is that, barring extenuating circumstances, releases should be accomplished within two to four hours. [See DE 49-1 at 427]. Because the release process appears to be different for imprisoned and detained inmates, the Court will create two subclasses: one for imprisoned inmates and one for detained ones. [Compare DE 49-1 at 392-393 with DE 49-1 at 408-410]. The Court will also increase the overdetention threshold for detained inmates from four hours to twelve hours. Defendants have produced no evidence that delays of more than twelve hours could be because of individual issues. See Driver, 859 F.3d at 491-492 (“Evidence in the record indicates that the average time period to 35 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 36 of 50 PageID #: 2122 effect such a release is 2–4 hours in counties in general, and up to 6 hours if problems are encountered, but even if we doubled those times, release still would be accomplished within 12 hours . . . At some point well short of the 24-plus hours alleged here, there is no reason to believe that individual issues would account for that delay.”). Twelve hours is a “practical compromise” between the rights of detained inmates and the realities of processing release orders for detained inmates in LMDC. McLaughlin, 500 U.S. at 53. By tripling the maximum time allowed under Defendants’ goal, the new class definition provides time to resolve issues that may arise, such as clarification of court orders. In addition, the Court will edit the class definition to include what Plaintiffs have affirmatively demonstrated is the cause of a substantial majority of overdetentions7: Defendants’ failure to implement and maintain an adequate process for timely releasing detained inmates and imprisoned inmates. See Driver, 859 F.3d at 494; see also Berry, 379 F.3d at 768. As a result, the class definition is: Subclass A: All persons who from February 3, 2016 to present were imprisoned in the Louisville Metro Department of Corrections for more than four hours after satisfaction of their term of imprisonment set by prior court order due to the failure of Defendants to implement and maintain an adequate process for timely releasing imprisoned inmates. Subclass B: All persons who from February 3, 2016 to present were detained in the Metro Government Department of Corrections for more than twelve hours after receipt of an order directing their release due to the failure of Defendants to implement and maintain an adequate process for timely releasing detained inmates. Having modified the class definition, the Court now turns to the factors for class certification. Although the Court has created a subclass of detained inmates and one of imprisoned inmates, the Court uses the term “over-detained” to refer to both groups. 7 36 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 37 of 50 PageID #: 2123 D. Requirements of Fed. R. Civ. P. 23(a) 1. Numerosity Rule 23(a)(1) requires that a class be “so numerous that joinder of all members is impracticable.” Fed. R. Civ. P. 23(a)(1). “There is no strict numerical test for determining impracticability of joinder.” In re Am. Med. Sys., Inc., 75 F.3d 1069, 1079 (6th Cir. 1996). “However, sheer number of potential litigants in a class, especially if it is more than several hundred, can be the only factor needed to satisfy Rule 23(a)(1).” Bacon v. Honda of America Mfg., Inc., 370 F.3d 565, 570 (6th Cir. 2004) (citing 1 Herbert B. Newberg & Alba Conte, Newberg on Class Actions, § 3:5, at 243-45 (4th ed. 2002)). In making this determination, the Court may consider “reasonable inferences drawn from facts before it.” Senter v. General Motors Corp., 532 F.2d 511, 523 (6th Cir.1976). But, the Court may not rely on speculation or the conclusory allegations of the proposed representatives. Instead, Plaintiffs must “show some evidence of or reasonably estimate the number of class members.” Brashear v. Perry Cty., Ky., No. CIV.A. 6:06143-DCR, 2007 WL 1434876, at *4 (E.D. Ky. May 14, 2007) (internal quotation marks and citation omitted). In support of their argument that they have satisfied the numerosity requirement, Plaintiffs assert: [A] review of just four months of LMDC records showed that out of 8,437 releases occurring during those periods: 27 took more than a day after receipt of the court order, with many being several days or even weeks; 342 took between 10 and 24 hours; and 3,066 took between 4 and 10 hours. Extrapolating those figures over the more than three-year period currently covered by the class yields: 37 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 38 of 50 PageID #: 2124 243 took more than a day after receipt of the court order; 6156 took between 10 and 24 hours; and 55,188 took between 4 and 10 hours. [DE 50 at 1241]. Defendants argue that Plaintiffs’ data is unreliable: “Plaintiffs [sic] methodology for its proposed class is also skewed by the fact that it does not contain any expert analysis or preparation of the data it bases its argument [sic]. The Plaintiffs [sic] prepared there [sic] own analysis which equates to a bare bone [sic] assertion. There [sic] methodology does not contain the root cause of every claimed overdention [sic].” [DE 53 at 1909]. Defendants do not support this argument with legal authority. Defendants did not hire their own expert to refute Plaintiffs’ methodology. They also apparently did not conduct their own review of the records. Plaintiffs have provided evidence that over a three-year period thousands of imprisoned inmates were held for more than four hours and hundreds of detained inmates were held for more than twelve hours. [See DE 50-7]. At this point, the subclasses will cover more than four years. When recalculated over the four-year period, the subclasses will be even larger. Based on Plaintiffs’ “reasonable inferences drawn from the facts,” the Court finds that Plaintiffs have shown that there are enough persons who would be potential members of the subclasses to satisfy the numerosity requirement. Senter, 532 F.2d at 523; see Phillips v. Philip Morris Companies Inc., 298 F.R.D. 355, 362 (N.D. Ohio 2014) (“Generally, the numerosity requirement is fulfilled when the number of class members exceeds forty. In addition, the numerosity requirement is satisfied where the exact size of the class is not known, but general knowledge and common sense indicate that the class is large”) (internal citation omitted). 38 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 39 of 50 PageID #: 2125 2. Commonality Rule 23(a)(2) requires that “there are questions of law or fact common to the class.” Fed. R. Civ. P. 23(a)(2). “Commonality requires the plaintiff to demonstrate that the class members ‘have suffered the same injury.’” Dukes, 564 U.S. at 349-50 (quoting Falcon, 457 U.S. at 157). Even so, this does not mean that plaintiffs may show commonality by asserting that they have all suffered a violation of the same provision of the law. Id. at 350. Instead, their claims must rely on a common contention “of such a nature that it is capable of classwide resolution—which means that determination of its truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke.” Id. Thus, commonality depends not upon the “raising of common ‘questions’—even in droves—but, rather the capacity of a classwide proceeding to generate common answers apt to drive the resolution of the litigation.” Id. (quoting Nagareda, 84 N.Y.U. L. REV. 97, at 132). “The mere fact that questions peculiar to each individual member of the class remain after the common questions of the defendant’s liability have been resolved does not dictate the conclusion that a class action is impermissible.” Sterling, 855 F.2d at 1197. Here, the subclasses can “generate common answers apt to drive the resolution of litigation.” Dukes, 564 U.S. at 350. The respective subclasses can answer common questions such as: whether LMDC held inmates for more than four hours, or twelve hours, longer than they legally could; whether Defendants failed to implement and maintain an adequate process for timely releasing detained inmates or imprisoned ones; whether Defendants knew that detained and imprisoned inmates were not being timely released; whether Defendants failed to implement the recommendations of the internal audit; whether Defendants’ failure to implement and maintain an adequate process constitutes deliberate indifference; and whether Defendants caused the constitutional injuries of the members of the subclasses. See Monell v. Dep’t of Soc. Servs. of City 39 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 40 of 50 PageID #: 2126 of New York, 436 U.S. 658, 694 (1978); D’Ambrosio v. Marino, 747 F.3d 378, 387–88 (6th Cir. 2014) (quoting Doe v. Claiborne Cnty., 103 F.3d 495, 508 (6th Cir.1996) (internal quotation marks omitted); see also Shorts v. Bartholomew, 255 F. App’x 46, 55 (6th Cir. 2007). Defendants first argue that Plaintiffs have failed to show commonality because the “proposed class definition is overly broad. The named class representatives or at least some of these class representatives are not proper because any over detainment did not result in a constitutional violation. For example, two of the so-called named Plaintiffs, Louis Hibbs, Jr. and Allen Goodflesch, were not over-detained at all and these individuals are merely asserting claims alleging they were denied work release.” [DE 67 at 1978]. But, this is a moot issue because Plaintiffs did not include Hibbs and Goodfleisch in their proposed class. [DE 50 at 1218 (“Movants do not include all of the named Plaintiffs in this Motion. The claims of Plaintiffs, Larry Louis Hibbs, Jr. and Allen Goodfleisch relate solely to Defendants’ failure to effect their release for work or home incarceration”)]. Defendants next contend that “Plaintiffs have . . . failed to establish the requisite of commonality in the instant case by their overly broad definition of the class set as four hours and the failure to exclude Court mistakes that caused over-detention.” [DE 67 at 1978]. Defendants further assert that “Plaintiffs have made no effort to identify the cause of any of the over-detentions except to incorrectly argue that they are caused by deliberate indifference” and that a lack of commonality is evidenced by “the inclusion of the named Plaintiffs whose claims do not include over-detention and the inclusion of claims that were not the result of the Defendants or attributable to Louisville Metro Department of Corrections (“LMDC”) actions.” Id. The Court disagrees. First, Plaintiffs have affirmatively demonstrated that Defendants’ failure to implement and maintain an adequate process for timely releasing inmates is the cause of 40 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 41 of 50 PageID #: 2127 a substantial majority of over-detentions. [See DE 50-10 at 1626]. Second, Plaintiffs asserted during oral argument (and Defendants did not rebut them) that LMDC does not include information about the cause of a late release in an inmate’s file. [DE 69 at 2044-2045]. Based on the Plaintiffs’ record and Defendants’ statements during oral argument, it appears as though Defendants do not systematically document over-detentions. At best, it appears that they document some over-detentions on an ad hoc basis; at worst, it seems like they fail to document a substantial percentage of them.8 Third, the Court is unpersuaded by Defendants’ arguments about Ms. Melloan and Mr. Smith. While Defendants argue these two inmates were not over-detained as a result of their actions, they are silent about the scores of other inmates Plaintiffs identified. Defendants argue: Ms. Melloan indicated that she was going to commit suicide and therefore on 1/7/18 she was placed on suicide. [sic] Part of the release procedure in addition [sic] processing the release is to clear from medical and complete a warrant check which took additional time. Ms. Melloan was scheduled to be released at 10:21 pm on January 8th, the same day as her order was issued when she asked an officer on the booking floor if she could be released in the morning. It was 37 degrees out and she had no ride home. There was no bus service at that time so her request was given the ok. This does not fall in the pattern of commonality as there is a separate defense to this claim. 8 Al-Amin testified about LMDC’s use of incident reports to document late releases [DE 49-1 at 499-450]: Q. Is an incident report supposed to be prepared every time somebody is not released when they should be? A. It should be. But not—I'm not going to say that it is always . . . Q. If it was a court error or you don’t do an incident report? A. We don’t do a—an incident report. Q. Okay. If it was a records error you—you’re supposed to? A. Generally, we do. Yeah. Generally, we do unless told otherwise. Q. Who makes the determination of whether it was a court error or a records error? A. I’ll . . . well, the [LMDC employee] sees it first . . . in their review and they’ll pass on to the . . . direct supervisor their findings. And they . . . can see, you know, if it was court error or if was records error. And, in either case, the supervisor will bring it to me and we’ll review it. And I’ll give my determination as to whether it was a court error . . . or a records error. And then . . . the supervisor and I will inform Chief of Staff Clark. Q. And who decides whether or not an incident report will be prepared? A. Chief of Staff Clark. 41 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 42 of 50 PageID #: 2128 [DE 53 at 1915-1916]. Plaintiffs vigorously dispute Defendants’ characterization of the cause of Ms. Melloan’s delayed release: LMDC Records allegedly received the release order at 10:58 a.m. on 1/8/18. The release order was not even started to be processed until 18:31 (6:31 p.m.) – 7 hours later. Then it took Defendants until 10:21 p.m. to complete the paperwork for her release – almost 12 hours after the release order was entered. Then, they did not attempt to go get Ms. Melloan to release her until 1:31 a.m. – almost 14 hours after the release order was entered. At this time of the morning when it was too cold and she did not have a ride, Ms. Melloan had no choice but to ask to be kept until the next morning. Then, Defendants’ own timeline states “wait until 8:00 a.m.” – yet they still did not release her until 10:45 a.m. the next morning– almost three more hours after 8:00 a.m. Id. Defendants ignore entirely that it was their failure to effect her release within a reasonable period of time after the order was received which forced her to spend another night in jail. Of note, is that this happened to Ms. Melloan almost a year AFTER this action was filed. [DE 56 at 1961] (formatting in original). As to Justin Smith, Defendants included him in the spreadsheets they provided to Plaintiffs as part of the discovery process. [DE 50-2 at 1259]. Based on these records, Justin Smith was the only inmate out of the 43 identified by Defendants whose release was delayed because of an error with a court order. At this point, the Court is not persuaded—based on Defendants’ representations about these isolated cases—that Plaintiffs have failed to affirmatively demonstrate commonality between members of the subclasses. Finally, Defendants’ reliance on Wharton is misplaced. [DE 67 at 1979-1981]. First, unlike in Wharton, the Court is not considering Plaintiffs’ evidence under the summary judgment standard. Second, Plaintiffs have submitted more than “speculation” about what caused their untimely releases. Along with sworn testimony and several audits, Plaintiffs have submitted statistical evidence that thousands of inmates were not timely released during the class period. [See DE 50-5 and DE 50-6]. Plaintiffs’ data—which, for purposes of class certification, 42 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 43 of 50 PageID #: 2129 Defendants have not convincingly refuted with their own—shows that late releases increased after the filing of this lawsuit. [See DE 50-7]. Moreover, unlike in Wharton, there are not “substantial and uncontested records that many over-detentions originate in the court system.” Wharton, 854 F.3d at 246. Based on CGL’s audit, a substantial percentage of delays are not caused by issues with unclear court orders. [DE 50-10 at 1626]. Third, in highlighting how defendants in that case were not deliberately indifferent—by increasing staffing levels, improving the agency’s training system, upgrading its technology, and creating special units—Wharton shines a harsh light on Defendants’ actions (or inactions) in this case. 3. Typicality To certify a class under Fed. R. Civ. P. 23(a), a plaintiff must also prove that “the claims or defenses of the representative parties are typical of the claims or defenses of the class.” Fed. R. Civ. P. 23(a)(3). “Typicality determines whether a sufficient relationship exists between the injury to the named plaintiff and the conduct affecting the class, so that the court may properly attribute a collective nature to the challenged conduct.” Sprague v. Gen. Motors Corp., 133 F.3d 388, 399 (6th Cir. 1988) (internal quotation marks omitted) (quoting In re Am. Med. Sys., 75 F.3d at 1082). “A claim is typical if ‘it arises from the same event or practice or course of conduct that gives rise to the claims of other class members, and if his or her claims are based on the same legal theory.’” Beattie v. CenturyTel, Inc., 511 F.3d 554, 561 (6th Cir. 2007) (quoting In re Am. Med. Sys., 75 F.3d at 1082). “The commonality and typicality requirements of Rule 23(a) tend to merge. Both serve as guideposts for determining whether under the particular circumstances maintenance of a class action is economical.” Falcon, 457 U.S. at 158 (1982). “Under the commonality prong, a court must ask whether there are sufficient factual or legal questions in common among the class 43 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 44 of 50 PageID #: 2130 members’ claims to make class certification economical and otherwise appropriate. In contrast, under the typicality prong, a court must ask whether, despite the presence of common questions, each class member’s claim involves so many distinct factual or legal questions as to make class certification inappropriate.” Brashear, 2007 WL 1434876 at *6 (internal quotation marks and citation omitted). Plaintiffs argue that “[t]here is really no disputing the fact that Movants’ claims are typical of those of the class. Movants were overdetained for periods of between 24 hours and six days. There is no evidence in this case that Movants were singled out individually for their overdetention, which is really the only thing that could distinguish their claims from those of the class generally.” [DE 50 at 1244]. Defendants do not directly dispute this requirement for class certification. Here, the Named Plaintiffs’ claims against Defendants are typical of the respective subclasses because they stem from the same legal theory as the class members: Defendants violated their constitutional rights by over-detaining them. If the fact finder determines that Defendants have an inadequate process for timely releasing detained inmates and imprisoned ones, such a determination would further the interest of each member of the subclasses. Thus, Rule 23(a)(3)’s typicality requirement is met. 4. Adequacy of Representation The final requirement is that Plaintiffs show that “the representative parties will fairly and adequately protect the interests of the class.” Fed. R. Civ. P. 23(a)(4). The Sixth Circuit has articulated two criteria to satisfy this requirement: “1) [t]he representative must have common 44 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 45 of 50 PageID #: 2131 interests with unnamed members of the class, and 2) it must appear that the representatives will vigorously prosecute the interests of the class through qualified counsel.” Senter, 532 F.2d at 525. Plaintiffs argue that they have satisfied this requirement: “There is no evidence of anything that distinguishes Movants’ interests in this litigation from that of the class as a whole . . . Movants have retained experienced and competent counsel who will fairly and adequately protect the interests of the proposed class.” [DE 50 at 1244-1245]. Although Defendants assert that Plaintiffs have failed to satisfy this requirement, they do not cite any authority nor make any arguments in support of their position. As to the first consideration, Plaintiff Healey, who was imprisoned, has the same interest as the imprisoned members in Subclass A; Plaintiffs Jarvis, Yates, and Melloan, who were all detained, have the same interest as the detained members in Subclass B. [DE 37 at 349-350]. Proposed class counsel also appear adequate to represent the class in this case. Attorneys James D. Ballinger and Gregory A. Belzley are both experienced and competent litigators who are capable of adequately representing the members of the proposed subclasses. [DE 50-14; DE 5015]. Moreover, both attorneys have experience in class-action cases specifically. Id. After considering all four elements of Fed. R. Civ. P. 23(a), Plaintiffs have satisfied their burden. Attorneys Ballinger and Belzley appear adequate to represent the proposed subclasses. E. Fed. R. Civ. P. 23(b) The Plaintiffs must also show that the action fits one of Rule 23(b)’s subdivisions. Plaintiffs seek certification of the proposed class of plaintiffs under Rule 23(b)(1), (2), and (3). [DE 50 at 1245-1248]. Certification is proper under Rule 23(b)(3). As a result, the Court need not address Rule 23(b)(1) or (b)(2). 45 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 46 of 50 PageID #: 2132 1. Fed. R. Civ. P. 23(b)(3) Rule 23(b)(3) requires “that the questions of law or fact common to the members of the class predominate over any questions affecting only individual members, and that a class action is superior to other available methods for the fair and efficient adjudication of the controversy.” Fed. R. Civ. P. 23(b)(3). As a result, to qualify for certification under Rule 23(b)(3), the proposed class “must satisfy a two-part test of commonality and superiority and should only be certified if doing so would ‘achieve economies of time, effort, and expense.’” Cochran v. Oxy Vinyls LP, No. 3:06CV-364-H, 2008 WL 4146383, at *11 (W.D. Ky. Aug. 29, 2008) (quoting Sterling, 855 F.2d at 1196). a. Predominance “Subdivision (b)(3) of Rule 23 parallels subdivision (a)(2) of Rule 23 in that both require that common questions exist, but subdivision (b)(3) contains a more stringent requirement that common issues ‘predominate’ over individual issues.” Whitlock v. FSL Mgmt., LLC, No. 3:10CV00562-JHM, 2012 WL 3274973, at *11 (W.D. Ky. Aug. 10, 2012). “To satisfy the predominance requirement in Rule 23(b)(3), a plaintiff must establish that the issues in the class action that are subject to generalized proof . . . predominate over those issues that are subject only to individualized proof.” Beattie, 511 F.3d at 564 (internal quotation marks and citation omitted). “There are no bright lines for determining whether common questions predominate.” In re Cardizem CD Antitrust Litigation, 200 F.R.D. 297, 307 (E.D. Mich. 2001) (citing In re Potash Antitrust Litig., 159 F.R.D. 682, 693 (D. Minn. 1995)). Instead, the Court must determine whether the questions common to the class are “at the heart of the litigation,” Powers, 501 F.3d at 619, while keeping in mind “the mere fact that questions peculiar to each individual member of the 46 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 47 of 50 PageID #: 2133 class action remain after the common questions of the defendant’s liability have been resolved does not dictate the conclusion that a class action is impermissible.” Sterling, 855 F.2d at 1197. Defendants contend that “[t]he proposed class does not meet the predominance requirements of Fed. R. Civ. P 23 (b)(3). Plaintiffs claim questions of law and fact are common to the class thus making the class satisfy the predominance requirement. However, individualized issues in Plaintiffs[’] claims predominate over common ones.” [DE 67 at 1982]. Citing Harper, the Defendants argue that determining liability would require individualized determinations about why each inmate was over-detained and whether the delay was reasonable. Id. at 1983. The Court disagrees. Plaintiffs’ subclasses are not premised on the unreasonableness of each over-detention. Rather, they are based on common questions about whether Defendants’ inadequate process for timely releasing inmates caused respective subclass members to be over-detained for four hours or twelve hours. See Driver, 859 F.3d at 492 (“Neither Portis nor Harper preclude class certification in a case such as this one, in which the plaintiffs assert that the defendants’ policy or practice caused them to be detained for an unconstitutionally unreasonable length of time”); Berry, 379 F.3d at 768 (“Stated another way, the plaintiffs in this case challenge the implementation of the County’s policies, rather than the specific policies themselves. They claim that the County’s unreasonably inefficient implementation of its administrative policies amounts to a policy of deliberate indifference to their constitutional rights”); Thompson v. Jackson, No. 1:16-CV-04217, 2018 WL 5993867, at *8 (N.D. Ga. Nov. 15, 2018) (“But Plaintiffs do not claim that their overdetentions were unreasonable given their particularized (or individual) circumstances, so that each class member’s circumstances become relevant. Instead, Plaintiffs allege that they were all intentionally detained for the same reason and 47 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 48 of 50 PageID #: 2134 by the same event: Defendants’ decision to detain inmates eligible for release during the November 2014 outage simply because they wanted to run GCIC checks, and the GCIC system was down”). As the respective subclasses have been defined, Defendants’ potential liability arises from the common question of over-detention. Even if the amount of damages may vary slightly among the proposed members of the subclasses, the operative facts are the same—all proposed subclass members were over-detained beyond their court-ordered release at a facility operated by LMDC. As Plaintiffs have alleged, the rights of the proposed class members were being violated: [by] a continuing pattern of misconduct and were the result of policies, procedures, customs and practices (or the lack of same), of the LMDC, either written or unwritten, that are systematically applied whenever an arrestee is admitted to LMDC, and/or an egregious failure to train and supervise LMDC staff to prevent such misconduct. [DE 37 at 352]. While the amount of damages incurred by each proposed subclass member is individualized, the significant and common issue of Defendants’ alleged constitutional violations outweighs any issues relating to each proposed subclass member’s individual damages. Thus, the predominance requirement of Rule 23(b)(3) has been satisfied. b. Superiority Rule 23(b)(3) requires that the proposed “class action is superior to other available methods for fairly and efficiently adjudicating the controversy.” Fed. R. Civ. P. 23(b)(3). The Court must consider the following factors to determine whether a class action is the superior method for adjudicating the controversy: (A) the class members’ interests in individually controlling the prosecution or defense of separate actions; (B) the extent and nature of any litigation concerning the controversy already begun by or against class members; 48 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 49 of 50 PageID #: 2135 (C) the desirability or undesirability of concentrating the litigation of the claims in the particular forum; and (D) the likely difficulties in managing a class action. Id. These factors support certification. First, the members of the subclasses have no interest in individually controlling the prosecution of their claims. In fact, individual claims might be abandoned, given the relatively meager individual damages at stake for some potential class members. See Zinser v. Accufix Research Institute, Inc., 253 F.3d 1180, 1190 (9th Cir. 2001) (“Where damages suffered by each putative class member are not large, this factor weighs in favor of certifying a class action.”). Second, the Court is unaware of any other lawsuits in this district in which any of the potential subclass members have asserted similar claims against Defendants or that any other forum is more desirable or appropriate than this District. Third, deciding whether LMDC had an inadequate process for timely releasing inmates would “‘both reduce the range of issues and promote judicial economy.’” Whitlock, 2012 WL 3274973 at *13 (quoting Dodge v. County of Orange, 226 F.R.D. 177, 184 (S.D.N.Y. 2005)). Finally, although there are inherent difficulties in managing a class action, those difficulties do not render class action inappropriate. Thus, Rule 23(b)’s superiority is met. IV. CONCLUSION For all these reasons, IT IS ORDERED as follows. 1. Plaintiffs’ Motion to Certify Class [DE 50] is GRANTED IN PART and DENIED IN PART consistent with this OPINION. 2. Counts I, II, III, IV, V, VII, and VIII of the Third Amended Complaint shall be maintained as a class action under Fed. R. Civ. P. 23(b)(3) by Plaintiff Class Representatives on behalf of the subclasses, defined as: 49 Case 3:17-cv-00071-RGJ-RSE Document 72 Filed 01/15/21 Page 50 of 50 PageID #: 2136 Subclass A: All persons who from February 3, 2016 to present were imprisoned in the Louisville Metro Department of Corrections for more than four hours after satisfaction of their term of imprisonment set by prior court order due to the failure of Defendants to implement and maintain an adequate process for timely releasing imprisoned inmates. Subclass B: All persons who from February 3, 2016 to present were detained in the Metro Government Department of Corrections for more than twelve hours after receipt of an order directing their release due to the failure of Defendants to implement and maintain an adequate process for timely releasing detained inmates. 3. Attorneys James D. Ballinger and Gregory A. Belzley are appointed as co-class counsel for Plaintiffs as they meet all of the requirements of Fed. R. Civ. P. 23(g). 4. Plaintiffs Larry Louis Hibbs, Jr. and Allen Goodfleisch are not part of the proposed class. Plaintiffs’ counsel shall a file a status report as to Plaintiffs Hibbs and Goodfleisch within thirty (30) days of this Order. cc: counsel of record 50

Some case metadata and case summaries were written with the help of AI, which can produce inaccuracies. You should read the full case before relying on it for legal research purposes.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.