M.D. et al v. Perry et al, No. 2:2011cv00084 - Document 213 (S.D. Tex. 2013)

Court Description: ORDER granting in part and denying in part 160 Motion to Certify Class; denying as moot 199 Motion for Partial Summary Judgment; granting 205 Motion for Leave to File; granting 212 Motion for Leave to File.(Signed by Judge Janis Graham Jack) Parties notified.(mserpa, )
Download PDF
M.D. et al v. Perry et al Doc. 213 UNITED STATES DISTRICT COURT SOUTHERN DISTRICT OF TEXAS CORPUS CHRISTI DIVISION M.D.; bnf STUKENBERG, et al, § § § § § § § § Plaintiffs, VS. RICK PERRY, et al, Defendants. CIVIL ACTION NO. 2:11-CV-84 ORDER Before the Court is Plaintiffs’ Motion for Class Certification and Appointment of Class Council. (D.E. 160). Plaintiffs have asked the Court to certify one General Class and four subclasses. Also currently before the Court is Defendants’ Motion for Partial Summary Judgment on the claims of named plaintiffs D.I., S.R.,1 M.R., and J.R., Plaintiffs’ Motion for Leave to File a Fourth Amended Complaint, and Plaintiffs’ Motion for Leave to File Reply in Support of that motion. (D.E. 199; 205; 212). For the reasons stated below, Plaintiffs’ Motion for Class Certification is GRANTED in part and DENIED in part. Defendants’ Motion for Partial Summary Judgment is DENIED as moot. Plaintiffs’ Motion for Leave to File Reply in Support of Motion for Leave to File a Fourth Amended Complaint and Motion for Leave to File a Fourth Amended Complaint are GRANTED. I. Jurisdiction Plaintiffs bring their claims under 42 U.S.C. § 1983. (D.E. 179 at 7). This Court hence has subject matter jurisdiction over this case per 28 U.S.C. § 1331. 1 S.R. is listed as S.T. in the Third Amended Complaint (D.E. 179) and Plaintiffs’ Brief in Support of Motion for Class Certification (D.E. 160). The initials are corrected in Plaintiffs’ Post-Hearing Brief (D.E. 195). Class Cert. Hr’g 1/23/13(Young) at 207-10. 1 / 107 Dockets.Justia.com II. Background A. Procedural History Plaintiffs are minor children in the custody of the Texas Department of Family and Protective Services (“DFPS”). They are all in DFPS’s Permanent Managing Conservatorship (“PMC”). On March 29, 2011 Plaintiffs filed suit through their next friends against Rick Perry, Governor of Texas, Thomas Suehs, Executive Commissioner of the Health and Human Services Commission of the State of Texas, and Anne Heiligenstein, Commissioner of the Department of Family and Protective Services of the State of Texas (collectively, “Defendants”), in their official capacities.2 (D.E. 1). On April 5, 2011 Plaintiffs filed a Motion for Class Certification. The Court granted their motion on June 2, 2011, holding that the requirements of Federal Rules of Civil Procedure Rule 23 had been met. (D.E. 49). Defendants appealed to the Fifth Circuit Court of Appeals pursuant to Rule 23(f). (D.E. 63). On March 23, 2012 the Fifth Circuit vacated the class certification order and remanded the case to this Court for proceedings in light of the intervening Supreme Court ruling in Wal-Mart v. Dukes, 131 S. Ct. 2541 (2011). M.D. ex rel. Stuckenberg v. Perry, 675 F.3d 832 (5th Cir. 2012) (“MD”). Plaintiffs filed a second motion to certify the class on October 5, 2012. (D.E. 160). B. Factual Background When the Child Protective Services (“CPS”) division of DFPS determines a child has been maltreated or is at risk and should not remain in her home it removes the child and petitions a court for Temporary Managing Conservatorship (“TMC”) over the child. If TMC is granted, 2 Kyle Janek has since replaced Suehs as Executive Commissioner of the Health and Human Services Commission and John J. Specia, Jr. has replaced Heiligenstein as Commissioner of DFPS. Plaintiffs’ pleadings have been amended to reflect these changes. (D.E. 179). 2 / 107 DFPS takes custody of the child and attempts to achieve reunification or, if this is not possible, place the child with a new, permanent family. TMC lasts one year unless a court orders that it be extended another six months. If at the termination of TMC the child has not been reunified or placed in a new family the child is transferred to the PMC. The child remains in the State’s PMC until he or she is adopted, custody is otherwise transferred, or they age out of the system. (D.E. 179 at 1-2). There are roughly 12,000 children in the PMC at a given time. (D.E. 195 at 18). Children in the PMC are placed in a variety of different settings. Many of these placements are managed by private child placing agencies (“CPAs”) that contract with the State. DFPS also directly places some children in homes and residential facilities. In both cases the Residential Child Care Licensing (“RCCL”) division of DFPS has the ultimate responsibility for inspections and investigations of placements, as well as licensing and monitoring CPAs. Foster family homes are traditional foster homes and contain fewer than 7 children. They are verified and the caregivers receive training and financial support. The verification process involves screening the home and conducting a variety of inspections to ensure the placement meets the requirements related to the type of care it will provide. (PX 50 at 204). Foster group homes contain 7 to 12 children but are otherwise almost identical to foster family homes. (PX 390 at 2). They are regulated, receive financial support, but are not required to have nighttime awake supervision. (PX 50). Facilities that contain 13 or more children are termed General Residential Operations (“GROs”). They are licensed by the RCCL, subject to extensive regulations, and receive financial support. (PX 49). Presently, GROs have capacities ranging from 13 to 437 residents. (PX 302). Residential Treatment Centers (“RTCs”) are a type of GRO that provide specialized treatment. (PX 390 at 2). 3 / 107 Texas also utilizes kinship placements. These are placements with someone who is related to the child or has a longstanding and significant relationship with the child or her family. All kinship placements must be approved by DFPS and have a home assessment. (PX 406 at 4). Kinship placements can be either verified or unverified. If verified, they undergo the same licensing requirements as a foster family home, the caregivers have the same training requirements as foster parents, and the homes receive the same financial assistance as foster homes. If they are unverified, then they are only eligible for limited monetary assistance and the caregivers are not required to complete the training provided for foster parents. (D.E. 179 at 2; PX 406 at 4). C. Plaintiffs’ Claims Plaintiffs generally allege that “Texas policies and practices result in structural deficiencies that put all children who are or will be in the State’s PMC at an unacceptable risk of harm.” (D.E. 179 at 2). Plaintiffs assert two counts in their Third Amended Complaint and ask the Court to certify a General Class and four subclasses. Count I alleges a violation of Fourteenth Amendment substantive due process rights to adequate care and a safe, secure, and suitable placement while in State custody. Count II alleges a violation of the First, Ninth, and Fourteenth Amendment rights to family integrity of the members of the putative Licensed Foster Care Subclass. (D.E. 179 at 8, 20). In each of their claims Plaintiffs assert that there is some structural deficiency that constitutes a policy or practice of Defendants that violates class members’ constitutional rights. On behalf of all children in the PMC, Plaintiffs bring a class action alleging that the State’s policies and practices deprive class members of their right to be free from an 4 / 107 unreasonable risk of harm, including psychological harm, which the State owes them by virtue of its decision to take them into its custody. Plaintiffs allege that DFPS has a practice of overburdening caseworkers with excessive workloads. Because of this practice, caseworkers are unable to fulfill their duties, putting class members at risk as caseworkers are critical to maintaining a foster child’s safety and well-being. Caseworkers simply do not have the time and resources to adequately monitor a child to ensure that her living conditions are safe. Plaintiffs allege, as well, that the overburdened caseworkers lead to a greater number of placement moves, which can inflict psychological harm on class members. On behalf of one subclass they claim that the State fails to exercise sufficient oversight over the residences foster care children are placed in. They also claim that the State maintains an insufficient array of placements, so that class members must either be placed in residences ill-suited to their needs or be uprooted and relocated considerable distances. Plaintiffs claim that Texas’ practices of placing children in the foster group homes described above and of placing children who need only basic child care services in GROs violate their respective subclass’ Fourteenth Amendment rights. In the course of this litigation Plaintiffs’ claims have changed substantially. In their original Complaint, (D.E. 1), Plaintiffs alleged a variety of problems with Texas’ management of foster care and claimed that the State’s mismanagement violated the rights of children in the PMC. In vacating the original certification of the class, the Fifth Circuit expressed doubt as to whether Plaintiffs could “advance a due process claim based on a bare finding that Texas has ‘organized or managed’ DFPS improperly.” MD, 675 F.3d at 841 n.3. Moreover, the Fifth Circuit noted that the original Complaint aggregated a number of individual claims into one “super-claim” on behalf of the class. Id. at 844. In response, Plaintiffs have proposed several subclasses and changed their allegations. 5 / 107 D. Named Plaintiffs Named plaintiff M.D. was removed from her home because she was neglected by her mother and had been abandoned by her father. (D.E. 1 at 6). At the age of 10 she was placed with her relatives, but was then removed from that home when her cousin sexually assaulted her. (PX 27 at DFPS #28240 (filed under seal)). After the incident, M.D.’s relatives decided that they could not care for her, and she was placed in the State’s custody. Id. at DFPS #28544. Since then, M.D. has been placed in a number of institutional facilities due to her therapeutic needs. These placements have involved moving M.D. far from her home community, her siblings (who are not in foster care), and necessitated repeatedly enrolling her in new schools. Id. at DFPS #28382, DFPS #28394, DFPS #28394. While placed in an RTC M.D. walked to a nearby retail establishment where she was raped. (D.E. 1 at 6). There have also been several reports that M.D. has been subjected to sexual abuse or misconduct while in the State’s custody. (PX 27 at 1 RFP 1 RCCL 00725, 1 RFP 1 RCCL 00863, 1 RFP 1 RCCL 00872-74, 1 RFP 1 RCCL 00009 (filed under seal)). Since being placed in Texas’ PMC, M.D. has been placed primarily in RTCs and Intensive Psychiatric Transition Programs (“IPTPs”) in RTCs that are designed to transition a child into less structured and restrictive environments.. E.g., id. at DFPS #28393-94, DFPS #28396-97, DFPS #28403, DFPS #282419-20, DFPS #28423-24, DFPS #28432-34. She has also been placed in juvenile detention on at least four occasions. Id. at DFPS #28544, DFPS #28574-75, DFPS #30210. Finally, her time in the State’s care has been marked by a number of hospitalizations, including as many as three hospitalizations in the course of a single year. E.g., id. at DFPS #28420. These hospitalizations were precipitated by M.D.’s “extreme behaviors” 6 / 107 including “severe self-harming behaviors” and cutting. Id. at DFPS #28420, DFPS #28428, DFPS #28399-400, DFPS #28407-08, DFPS #28415-16. M.D. currently resides in an RTC in Austin County. (D.E. 179 at 3). Excepting her hospital stays, M.D. has had 11 placements since being removed from her original home. D.I. is an eleven year old boy from Houston. He has been involved with DFPS since 2007 when DFPS received reports that his drug-addicted mother was neglecting him. He ran away from home in 2008 and DFPS instituted a safety plan but left him with his mother. In 2009 DFPS removed him from his home after the safety plan failed to end D.I.’s abuse and neglect. He entered the PMC in 2010. (D.E. 1 at 8-9). He was first placed with a relative in an arrangement that was intended to be permanent, but he was removed in June 2010 because that relative could not manage his behaviors. Id.; (PX 26 at DFPS #16449 (filed under seal)). At the age of eight he was placed in a foster group home. In his first month there two older boys in the home, aged 16 and 17, sexually abused him in at least three episodes involving oral and anal sex. The caregiver in the foster group home was unaware of this sexual abuse until one of the older boys confessed. (PX 26 at DFPS #17794, 1 RFP 2RCCL 00930 (filed under seal)). D.I. was removed to an emergency respite placement and then to another foster home. Id. at 1 RFP 2 RCCL 01021. He has subsequently been moved through several different foster homes and psychiatric hospitals. (D.E. 1 at 9-10). Excluding trips to hospitals, he has had four different placements while in care. In one of these he was placed with another child with a history as a victim of sexual abuse. While they were unsupervised, this child attempted to initiate sexual contact with D.I. (PX 26 at CCGH-000370, 1 RFP CPS 002150 (filed under seal)). Neither child was removed from the placement. On October 24, 2012 D.I. was adopted. (D.E. 199). 7 / 107 Z.H. was brought into foster care in February 2009 after being abused and neglected by his mother, who had left him malnourished and was insensitive to his medical and psychiatric needs. (PX 28 at DFPS #33580 (filed under seal)). He was separated from his siblings and placed in emergency shelters and hospitals where he was overmedicated; on at least one occasion heavy stimulation was required just to keep his eyes open. (D.E. 1 at 12; PX 28 at DFPS #33589 (filed under seal)). During his first six months in state care he was moved seven times, and has experienced nine placement moves total. He was first placed in an emergency shelter, then was moved to a foster group home. After the group home failed to meet his needs he was moved to a psychiatric hospital, then a foster family home, but then back to the psychiatric hospital, where he spent over a month. Z.H. then spent about three weeks in an emergency shelter before being placed in an RTC. (D.E. 1 at 12; PX 28 at DFPS #33187, DFPS #33190, DFPS #33193, DFPS #33196, DFPS #33199, DFPS #33202, DFPS #33580 (filed under seal)). He was in this RTC for almost a year but was removed when the facility gave notice that it could do no more for him. (PX 28 at DFPS #33205 (filed under seal)). He was then moved to his current placement, another RTC roughly 300 miles from his home in San Antonio. (D.E. 179 at 4; PX 38 at DFPS #33208 (filed under seal)). S.A. was removed from her parents’ custody in Florida after it was determined that she and her brother were at risk of physical abuse from their mother and that their father was unable to care for them. They were adopted by their grandmother, but more abuse was reported to Florida officials. Their grandmother relocated with them to Louisiana before the children could be removed by Florida officials. Child service authorities in Louisiana later removed S.A. from her grandmother’s care and placed her with a family friend. This family friend, however, returned her to her grandmother and they left the state. S.A. was removed by Texas in 2001 8 / 107 when her grandmother was arrested on outstanding warrants while traveling with a registered sex offender and attempting to cross the border into Mexico. S.A. was five years old at the time and was not with her brother, who has lived in Mexico since 2001. (D.E. 1 at 13; PX 32 at DFPS #48570, DFPS #49408 (filed under seal)). While in care she moved through foster homes, RTCs, and emergency shelters, often being moved after displaying aggressive and self-abusive behavior. (PX 32 at DFPS #48993, DFPS #48996, DFPS #48999, DFPS #49002, DFPS #49012 (filed under seal)). She has also spent time in several psychiatric hospitals. Id. at DFPS #49022, DFPS #49025, DFPS #49025, DFPS #49033, DFPS #49037, DFPS #49041, DFPS #49067, DFPS #49071, DFPS #49079, DFPS #49091, 1 RFP CPS 007960. During her time in care she has experienced over 24 different placements throughout the State and has had 12 different caseworkers. (D.E. 1 at 14-15). She is now 16 years old and is placed in an RTC in the Houston area. (D.E. 179 at 4; PX 32 at 1 RFP CPS 034267 (filed under seal)). A.M. is a 15 year old who has been in foster care since July 2004. She was removed at the age of six after calling the police to report a violent fight between her mother and two men. (D.E. 1 at 15-17). When she was taken into care her father was on death row and her mother was in the process of being tried for a homicide, which A.M. witnessed. (PX 25 at DFPS #3980 (filed under seal)). In one foster home she was punished by being made to drink vinegar and run laps. Id. at DFPS #4534; (D.E. 1 at 16). Later, A.M. was placed with her younger sister in a foster home that was supposed to lead to adoption. After over a year in the home, the foster parents told her that they were adopting her sister, but would not adopt her. A.M’s behavior and emotional state then deteriorated and she was sent to a psychiatric hospital. She has been hospitalized at least three times due to outbursts, violent behavior, and suicidal ideation. E.g., (PX 25 at DFPS #3915, DFPS #3980, DFPS #493, DFPS #4049 (filed under seal)). After being 9 / 107 discharged she was placed in an RTC far from that home and rarely has had contact with her sister or her former foster parents, who have no interest in bringing her back into the family. Id. at DFPS #313; (D.E. 1 at 15-17). She was moved to a foster group home in 2011, but was soon moved back to an RTC. She currently resides in another RTC almost three hours from her home community. (D.E. 179 at 4; PX 25 at DFPS #503, 1 RFP CPS 010074, 1 RFP CPS 045902, 1 RFP CPS 045903 (filed under seal)). During her time in the State’s care A.M. has had nine placements. In 2007 J.S. was removed at the age of five from his mother’s home due to neglect and his mother’s drug abuse. Soon after being removed his personality changed dramatically and he became very moody and aggressive. He was taken to a psychiatric hospital after having suicidal and homicidal ideations. He was then separated from his sister and niece, who were also in the State’s care, and moved to an RTC about 500 miles away. After making little progress at that facility, a court ordered him moved closer to home, and he then cycled through several foster homes and emergency shelters until being moved to another distant RTC. (D.E. 1 at 20; PX 29 at DFPS #38084, DFPS #38090, DFPS #38093, DFPS #38096 (filed under seal)). He was removed from one facility after it was placed under adverse action by DFPS and another after it was determined there was a risk of sexual abuse by another child. (PX 29 at DFPS #38099, DFPS #38104 (filed under seal)). He currently resides in yet another RTC, again over 300 miles from his home. In his first year in State care alone, he had five different placements. (D.E. 1 at 19; PX 29 at DFPS #37882, DFPS #37924 (filed under seal)). In total, he has had nine different placements. (D.E. 179 at 4; PX 29 at 1 RFP CPS 055447 (filed under seal)); see also Class Cert. Hr’g, 1/22/13 (Ricker) at 94). 10 / 107 P.O. and H.V. are members of a sibling group, aged 9 and 13 respectively. They were removed from their home in 2004 on the grounds of neglect by their mother who was found to be abusing drugs and providing inadequate housing. (PX 33 at 1RFP CPS 030253 (filed under seal); PX 34 at 1 RFP CPS 030255 (filed under seal)). They were both placed with a relative, which was intended to be a permanent home, but were removed a year later because of risk of abuse. (PX 33 at 1 RFP CPS 031087 (filed under seal); PX 34 at 1 RFP CPS 031176 (filed under seal)). They then spent several months with another relative, but were again removed as a precautionary measure due to allegations of abuse. (PX 33 at 1 RFP CPS 031090 (filed under seal); PX 34 at 1 RFP CPS 031179 (filed under seal)). About five years later, H.V. would be returned to this relative’s care, with the intention that the placement be permanent, but was again removed, this time due to his own behaviors. (PX 34 at 1 RFP CPS 031213 (filed under seal)). After being removed from his relatives’ care, P.O. was placed in an emergency shelter, where he remained for 30 days. (PX 33 at 1 RFP CPS 031090, 1 RFP CPS 031093 (filed under seal)). P.O. spent the next two years in a foster home away from his brothers, was then moved to the same city as his siblings, and was eventually placed in a home with them. Id. at 1 RFP CPS 031096, 1 RFP CPS 031099, 1 RFP CPS 031102. Reuniting P.O. with his siblings thus necessitated another placement move after being in a foster home for about eight months. He was then placed in several homes that were intended to lead to adoptions, but none of these were successful and he was placed in foster group homes and GROs, sometimes along with his siblings. Id. at 1 RFP CPS 031113, 1 RFP CPS 031117, 1 RFP CPS 031121, 1 RFP CPS 031225, 1 RFP CPS 031299. During this time P.O. had problems with aggression and destructive behavior, made suicidal statements that resulted in a short stay at a psychiatric hospital, and was on several medications to control his mood and behavior. Id. at 1 RFP CPS 11 / 107 030658, 1 RFP CPS 031804, 1 RFP CPS 032430, 1 RFP CPS 032437. P.O.’s case file documents his last placement over a year ago, in a foster group home. (PX 33 at 1 RFP CPS 059489 (filed under seal)). His next friend, Anna Ricker, testified that he is currently in a kinship placement, but was unsure if it was verified or not. Class Cert. Hr’g, 1/22/13 (Ricker) at 123-25; see also (D.E. 179 at 5). Since first coming into the State’s care, P.O. has had at least 11 placements. After being removed from his relatives’ home (for the first time), H.V. spent a month in an emergency shelter followed by several foster family homes and foster group homes. (PX 34 at 1 RFP CPS 031182, 1 RFP CPS 031185, 1 RFP CPS 031188, 1 RFP CPS 031191, 1 RFP CPS 031201, 1 RFP CPS 031205 (filed under seal)). In 2010 he was placed in an adoptive home, but adoption was disrupted and he was moved first to a GRO and then back to the kinship placement described above. After being removed from that home a second time, H.V. spent almost four months in an emergency shelter. Id. at 1 RFP CPS 031213, 1 RFP CPS 031217. In September 2011, H.V. was placed in a foster group home and has remained in there since. (D.E. 179 at 4; PX 34 at 1 RFP CPS 059587, 1 RFP CPS 059589 (filed under seal)). In total, he has been in nine different placements. Class Cert. Hr’g, 1/22/13 (Ricker) at 97. K.E. was removed from her home at the age of eight because of neglectful supervision and physical and emotional neglect. (D.E. 1 at 22; PX 31 at DFPS #44595, DFPS #44605 (filed under seal)). She was placed in a therapeutic foster family home, but was relocated to an emergency shelter, where she remained for two and a half months, as she required more intensive care. (PX 31 at DFPS #44877, DFPS #44880 (filed under seal)). She then spent two and a half years at an RTC until the facility determined it could do no more for her. Id. at DFPS #44883. Since then, she has been placed in a variety of foster family homes, foster group homes, 12 / 107 RTCs, and emergency shelters, often being removed from one placement due to behavioral problems, taken to an RTC, and then being moved to a less restrictive residence when her necessary level of care decreased, only for the pattern to begin again. Id. at DFPS #44886, DFPS #44889, DFPS #44892, DFPS #44895, DFPS #44898, DFPS #44901, DFPS #44907, DFPS #44910, DFPS #44916, DFPS #44919, DFPS #44934, DFPS #44938. She has also been placed in psychiatric hospitals and juvenile justice facilities. Id. at DFPS #44902, DFPS #44904, DFPS #44913, DFPS #44922, 1 RFP CPS 018759, 1 RFP CPS 018761. Since entering care she has had over 30 different placements, including 7 hospitals, 4 emergency shelters, and 11 different RTCs, at least 5 of which were in the Houston area, over 500 miles from her home. (D.E. 1 at 22; PX 71 at 1 RFP CPS 017395, 1 RFP CPS 018758, 1 RFP CPS 018759, 1 RFP CPS 018760, 1 RFP CPS 018762, 1 RFP CPS 038998 (filed under seal)). She is now 17 years old and resides at an RTC that is over a five hour drive from her home community. (D.E. 179 at 5 (filed under seal)). S.R., M.R., and J.R. (“the R. children”) are siblings from Houston aged seven, six, and three respectively. (D.E. 179 at 5). They were placed in the custody of the State in 2010. (DX 55 at 1 RFP CPS 072696, 1 RFP CPS 073453, 1 RFP CPS 108254 (filed under seal)). This occurred after then six month old J.R. was taken to the hospital with seizures caused by incorrectly mixed formula and was found to be malnourished and have several serious medical problems. She and her siblings were removed from their mother because their mother had mental health issues and could not take care of them or provide for J.R. and S.R.’s special needs. Id. at 1 RFP CPS 073470, 1 RFP CPS 073471, 1 RFP CPS 073473, 1 RFP CPS 108219. Shortly after entering care the R. children were placed with their maternal aunt due to a court order. Id. at 1 RFP CPS 072696, 1 RFP CPS 072703, 1 RFP CPS 072782. She is not a 13 / 107 licensed or verified caregiver and receives only very limited financial support from the State. Class Cert. Hr’g 1/23/13 (Young) at 215-16. Their aunt initiated adoption proceedings. Id. at 231, 233-34. The children were later removed from their aunt’s care due to health and safety concerns as well as the fact that there were seven children under the age of six in the home. (DX 55 at 1 RFP CPS 077572-78 (filed under seal)). The R. children’s next friend, Bobby Young, acting as their attorney ad litem, requested that a court return them to their aunt’s care, and the court issued an order to that effect. Id. at 1 RFP CPS 109781; Class Cert. Hr’g 1/23/13 (Young) at 217. According to Young, the foster parents had not been attentive to the needs of the children, J.R. had unexplained injuries, the children’s emotional state appeared to suffer, and the foster parents were reticent about her visits to check up on the children. Class Cert. Hr’g 1/23/13 (Young) at 236, 238-39. Custody for J.R. has been transferred to their aunt, and in the case of S.R. and M.R., jointly to their aunt and paternal grandmother. (D.E. 199, exh. 1). Based on the record before the Court, the R. children have had three placements since coming into the State’s care. L.H. and C.H. are two siblings out of eight that have been in foster care since 2008. Seven of the siblings are currently placed at a GRO. L.H. is 13 and C.H. is 10 and both are classified at the basic level of care. After a disruption in a previous placement they were placed in this GRO because that was the only way DFPS could keep the siblings together. Class Cert Hr’g. 1/23/13 (Vasquez) at 160-61. Estela Vasquez, their next friend, advocated for their placement together because the trauma the children were already going through would be worse if DFPS disrupted their tight sibling relationships, the only source of stability in their lives. Id. at 160-62. Due to high turnover, since coming into care they have had several different 14 / 107 caseworkers. Id. at 165. L.H. and C.H. have each been in two placements in the foster care system. III. Legal Framework for Class Certification: Rule 23 Class Certification is governed by Federal Rule of Civil Procedure 23. To warrant class certification, all of Rule 23(a)’s threshold requirements must be satisfied as well as one of the conditions contained in Rule 23(b). E.g., MD, 675 F.3d 832, 837 (5th Cir. 2012); Maldonado v. Ochsner Clinic Found., 493 F.3d 521, 523 (5th Cir. 2007). Rule 23(a) requires that: (1) the class is so numerous that joinder of all members is impracticable [numerosity]; (2) there are questions of law or fact common to the class [commonality]; (3) the claims or defenses of the representative parties are typical of the claims or defenses of the class [typicality]; and (4) the representative parties will fairly and adequately protect the interests of the class [adequacy of representation]. Fed. R. Civ. P. 23(a). Plaintiffs seek certification under Rule 23(b)(2), which allows for certification if: “the party opposing the class has acted or refused to act on grounds that apply generally to the class, so that final injunctive relief or corresponding declaratory relief is appropriate respecting the class as a whole.” Fed. R. Civ. P. 23(b)(2). Plaintiffs, as the party seeking class certification, bear the burden of establishing that Rule 23’s requirements have been met. Wal-Mart, 131 S.Ct. at 2551 (“Rule 23 does not set forth a mere pleading standard. A party seeking class certification must affirmatively demonstrate his compliance with the Rule—that is, he must be prepared to prove that there are in fact sufficiently numerous parties, common questions of law or fact, etc.”); MD, 675 F.3d at 837; In re Am. Med. Sys., Inc., 75 F.3d 1069, 1079 (6th Cir. 1996). In determining whether Rule 23 has been satisfied, courts are instructed to conduct a “rigorous analysis” and to “look beyond the pleadings to ‘understand the claims, defenses, relevant facts, and applicable substantive law in order to make 15 / 107 a meaningful determination of the certification issues.’” McManus v. Fleetwood Enters., Inc., 320 F.3d 545, 548 (5th Cir. 2003) (quoting Castano v. Am. Tobacco Co., 84 F.3d 734, 744 (5th Cir. 1996)); see also Castano, 84 F.3d at 740 (citing Gen. Tel. Co. v. Falcon, 457 U.S. 147, 161 (1982)); Unger v. Amedisys Inc., 401 F.3d 316, 319, (5th Cir. 2005). The same analytical approach is used for Rule 23(a) and Rule 23(b). Comcast Corp. v. Behrend, 133 S.Ct. 1426, 1432 (2013). This rigorous analysis will sometimes touch upon the merits of a plaintiff’s case as “class determination generally involves considerations that are enmeshed in the factual and legal issues comprising the plaintiff’s cause of action.” Wal-Mart, 131 S.Ct. at 2551 (citing Falcon, 457 U.S. at 147). In these cases, the plaintiff’s Rule 23 burdens are not relaxed. Id. at 2552; Funeral Consumers Alliance, Inc. v. Serv. Corp. Int’l, 695 F.3d 330, 346 (5th Cir. 2012); Ellis v. Costco Wholesale Corp., 657 F.3d 970, 981 (9th Cir. 2011); In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305, 307, 315-17 (3d Cir. 2008); In re Initial Pub. Offerings Secs. Litig., 471 F.3d 24 (2d Cir. 2006). A plaintiff may, then, need to prove an issue at class certification, and again at trial in order to prevail. Funeral Consumers Alliance, 695 F.3d at 346 (quoting Wal-Mart, 131 S.Ct. at 2552 n.6). However, any inquiry into the merits of a plaintiff’s case at this stage is limited to those issues germane to class certification. “Rule 23 grants courts no license to engage in free-ranging merits inquiries at the certification stage. Merits questions may be considered to the extent—but only to the extent—that they are relevant to determining whether the Rule 23 prerequisites for class certification are satisfied.” Amgen Inc. v. Conn. Ret. Plans & Trust Funds, 133 S. Ct. 1184, 1194-95 (2013) (internal citations omitted); see also Coastal Neurology, Inc. v. State Farm Mut. Auto. Ins. Co., 458 Fed. Appx. 793, 794 (11th Cir. 2012). Similarly, the Advisory Committee 16 / 107 Notes for Rule 23 provide: “an evaluation of the probable outcome on the merits is not properly part of the certification decision.” Advisory Committee’s 2003 Note on subd. (c)(1) of Fed. Rule Civ. P. 23; see also Daffin v. Ford Motor Co., 458 F.3d 549, 552 (6th Cir. 2006) (agreeing with the proposition that “whether class members will ultimately be successful in their claims is not a proper basis for reviewing a certification of a class action”). The Fifth Circuit recently considered this issue in Erica P. John Fund v. Halliburton, Co., 718 F.3d 423 (5th Cir. 2013) (“EPJ Fund”). The defendants claimed to have evidence negating a critical element of the plaintiffs’ case. The district court refused to hear this evidence, though, concluding that it did not bear on the Rule 23 analysis. Relying on Amgen, the court of appeals affirmed: regardless of the defendants’ evidence the issue at hand was a common one— the class members’ claims would, on this point, all stand or fall together. EPJ Fund, 718 F.3d at 433-34. Since Rule 23 was satisfied, the Fifth Circuit concluded that this evidence had no place at class certification, despite it being potentially dispositive of the class claims on the merits. In sum, a court should not shy away from merits inquiries when they are necessary to determine whether Rule 23’s requirements have been satisfied, but should go no further than that. A. Rule 23(a) Requirements 1. Numerosity Rule 23(a)(1) “requires examination of the specific facts of each case and imposes no absolute limitations.” Gen. Tel. Co. of Nw. v. EEOC, 446 U.S. 318, 329 (1980). There is no magic number: “The proper focus is not on numbers alone, but on whether joinder of all members is practicable in view of the numerosity of the class and all other relevant factors.” Phillips v. Joint Legislative Comm., 637 F.2d 1014, 1022 (5th Cir. 1981). Courts must consider 17 / 107 “the geographical dispersion of the class, the ease with which class members may be identified, the nature of the action, and the size of each plaintiff’s claim.” Zeidman v. J. Ray McDermott & Co., Inc., 651 F.2d 1030, 1038 (5th Cir. 1981). 2. Commonality Commonality was the main issue in the Supreme Court’s Wal-Mart decision, 131 S.Ct. at 2550 (“The crux of this case is commonality”), and is the most complex part of Rule 23(a) analysis. “Commonality requires the plaintiff to demonstrate that the class members ‘have suffered the same injury.’” Id. at 2551 (quoting Falcon, 457 U.S. at 157). It is not enough that class members suffer the same type of injury or have been subject to a violation of the same law. Id.; Lightfoot v. District of Columbia, 273 F.R.D. 314, 325 (D.D.C. 2011) (citing Love v. Johanns, 439 F.3d 723, 729 (D.C. Cir. 2006)). Rather, a plaintiff must identify a unified common policy, practice, or course of conduct that is the source of their alleged injury. For example, in Xuedan Wang, a class of former unpaid interns working at a number of Hearst’s magazines was not certified because the plaintiffs could not identify a uniform policy or set of policies regarding the way interns were treated, their responsibilities, and so on. The policies varied in significant ways among the 20 magazines and sometimes across departments within a single magazine. Xuedan Wang v. Hearst Corp., 2013 WL 1903787 (S.D.N.Y. May 8, 2013); see also, e.g., DL v. District of Columbia, 713 F.3d 120 (D.C. Cir. 2013); Luiken v. Domino's Pizza, LLC, 705 F.3d 370 (8th Cir. 2013); Bolden v. Walsh Const. Co., 688 F.3d 893 (7th Cir. 2012); Jamie S. v. Milwaukee Pub. Schs., 668 F.3d 481 (7th Cir. 2012). The policy or practice that a plaintiff identifies need not be formal or officially-adopted. Lightfoot v. District of Columbia sets out various ways plaintiffs can establish the existence of 18 / 107 government policies. The most straightforward of these is express adoption by the relevant entity, either through a legislative or administrative act (e.g., a local ordinance) or by an actor who possesses policymaking authority (e.g., a chief executive). Lightfoot, 273 F.R.D. at 320-21. Absent official sanction, a policy can be identified on the basis of custom or consistent practice. Id. at 321. Or, closely-related to custom, a uniform policy can be based on the defendant’s deliberate indifference: “The critical question here is whether the government has failed to respond to a need … in such a manner as to show ‘deliberate indifference’ to the risk that not addressing the need will result in constitutional violations.” Id. (internal quotation marks omitted) (citing Baker v. District of Columbia, 326 F.3d 1302, 1306 (D.C. Cir. 2003)). While using the same phrasing as the substantive due process deliberate indifference standard, see Section V. infra, the requirements for demonstrating the existence of a policy are less demanding. “The inquiry is an objective one … Mere negligence does suffice: the doctrine does not require the [official actor, in this case a city] to take reasonable care to discover and prevent constitutional violations, but simply means that, faced with actual or constructive knowledge that its agents will probably violate constitutional rights, the [official actor] may not adopt a policy of inaction.” Id. (emphasis in original) (internal quotation marks removed) (citing Warren v. District of Columbia, 353 F.3d 36, 39 (D.C. Cir. 2004)). A plaintiff also bears the burden of connecting the policy or practice to the alleged harm, especially in cases where this connection is not readily apparent. Cf. Ross v. RBS Citizens, N.A., 667 F.3d 900 (7th Cir. 2012) cert. granted, judgment vacated, 133 S. Ct. 1722 (2013) (certifying a class based on being denied overtime compensation due to unofficial company policies); In re Whirlpool Corp. Front-Loading Washer Prods. Liab. Litig., 678 F.3d 409, 414 (6th Cir. 2012) cert. granted, judgment vacated sub nom. Whirlpool Corp. v. Glazer, 133 S. Ct. 1722 (2013) 19 / 107 aff’d 2013 WL 3746205 (6th Cir. July 18, 2013) (certifying a class based on the purchase of washers with design defects); Butler v. Sears, Roebuck & Co., 702 F.3d 359 (7th Cir. 2012) cert. granted, judgment vacated, 133 S. Ct. 2768 (2013) and judgment reinstated, 2013 WL 4478200 (7th Cir. Aug. 22, 2013) (same);3 Gray v. Hearst Commc’ns, Inc., 444 F. App’x 698 (4th Cir. 2011) (certifying a class based on breach of a sufficiently similar contract). In doing so, “the party seeking class certification may rely on reasonable, common-sense assumptions and inferences to satisfy the requirements for class certification.” Kase v. Salomon Smith Barney, Inc., 218 F.R.D. 149, 152 (S.D. Tex. 2003) (citing Zeidman, 651 F.2d at 1039). McReynolds v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 672 F.3d 482 (7th Cir. 2012) cert. denied, 133 S.Ct. 338 (2012), involved a somewhat complicated connection between the defendant’s policies and the alleged harm. Plaintiffs had filed a Title VII class action on the basis of two company-wide policies: a “teaming” policy that permitted brokers in the same office to form their own teams, and an “account distribution” policy that transferred accounts from brokers who left the company largely on the basis of past performance. The plaintiffs alleged that these policies exacerbated racial discrimination. The theory of the claim was that, all else being equal, brokers tend to form teams of like members, so teams tend to be formed along racial lines. Members of a racial minority may therefore find it harder to find teams at all, 3 Ross, Butler v. Sears, and Whirlpool were remanded in light of Comcast Corp. v. Behrend, 133 S.Ct. 1426 (2013). Comcast added an additional consideration to class actions seeking certification pursuant to Rule 23(b)(3), which requires a demonstration that common questions of law and fact predominate over individual ones and that a class action is a superior method to adjudicate the controversy. Fed. R. Civ. P. 23(b)(3). Under Comcast, damage calculations must be measurable “on a classwide basis” and by a “common methodology,” and this methodology must encompass only those theories deemed amenable to classwide proof. Upon remand, the Sixth and Seventh Circuits reinstated the class certification orders in Butler v. Sears and Whirlpool. See Butler v. Sears, Roebuck & Co., 2013 WL 4478200, at *5 (“The concordance in reasoning and result of our decision and the Sixth Circuit’s decision averts an intercircuit conflict.”). The Seventh Circuit has not yet rendered a judgment regarding on Ross taking into account Comcast. These cases are cited here regarding Rule 23(a)’s threshold requirements, which were not in dispute in Comcast. Comcast, 133 S.Ct. at 1432; see also id. (“If anything, Rule 23(b)(3)’s predominance criterion is even more demanding than Rule 23(a)”); accord Ahmad v. Old Republic Nat’l Title Ins. Co., 690 F.3d 698, 702 (5th Cir. 2012); Semenko v. Wendy’s Int’l, Inc., 2013 WL 1568407 (W.D. Pa. Apr. 12, 2013). 20 / 107 or at least good ones, and are thus denied the advantages of the teaming policy. The account distribution policy is alleged to aggravate this disadvantage; brokers on good teams do better in the competition for account distributions, leading to a “vicious cycle” for minority (in this case African-American) brokers. McReynolds, 672 F.3d at 488-90. It is important to note, as the Seventh Circuit did, that class certification does not require the plaintiffs to establish that the harm actually occurred, i.e., they do not need to prove that the policies they identified did, in fact, cause the harm they are alleging. Consistent with Amgen and EPJ Fund, the only consideration at the class certification stage is whether the issues are appropriate for classwide litigation. That is, whether the plaintiffs have satisfied Rule 23’s requirements. If the teaming policy causes racial discrimination and is not justified by business necessity, then it violates Title VII as ‘disparate impact’ employment discrimination—and whether it causes racial discrimination and whether it nonetheless is justified by business necessity are issues common to the entire class and therefore appropriate for class-wide determination. … We are not suggesting that there is in fact racial discrimination at any level within Merrill Lynch, or that management’s teaming and account distribution policies have a racial effect. The fact that black brokers have on average lower earnings than white brokers may have different causes altogether. Id. at 489-90 (emphasis added). As the Lightfoot court indicated, failure to act can also constitute a policy or practice. Young v. Nationwide Mutual Insurance, Co. affirmed the certification of a class based on the defendants’ miscalculations of local taxes that they were authorized to collect. 693 F.3d, 532, 535 (6th Cir. 2012). Despite the defendants’ contentions that they had no uniform policy or practice that resulted in these errors—asserting instead that each was the product of unique circumstances like misspelling a name—the Sixth Circuit concluded that there was a common policy since defendants opted not to use available geocoding verification procedures that would 21 / 107 have, plaintiffs alleged, avoided the errors. Id. at 542-43. The Young plaintiffs were, therefore, able to point to a discrete policy or practice of the defendants—not using these procedures—and the availability of something like geocoding technology served to highlight the defendants’ choice. Cf. Wal-Mart, 131 S.Ct. at 2554, 2555-56 (“respondents have identified no ‘specific employment practice’—much less one that ties all their 1.5 million claims together.”); In re Countrywide Fin. Corp. Mortg. Lending Practices Litig., 708 F.3d 704, 708 (6th Cir. 2013). The common practice or policy does not have to injure every class member or injure them in exactly the same manner: “Class certification is appropriate ‘if class members complain of a pattern or practice that is generally applicable to the class as a whole. Even if some class members have not been injured by the challenged practice, a class may nevertheless be appropriate.’” Whirlpool, 678 F.3d at 420 (citing Gooch v. Life Investors Ins. Co. of Am., 672 F.3d 402, 428 (6th Cir. 2012)). In Whirlpool plaintiffs complained of defective front-loading washers, and though not all members of the class had suffered the actual injuries of mold and mildew growing in their washers, they shared a common legal injury in having purchased a defective washer. Id.; accord Butler v. Sears, 702 F.3d at 362 (“If, as appears to be the case, the defect in a Kenmore-brand washing machine can precipitate a mold problem at any time, the defect is an expected harm, just as having symptomless high blood pressures creates harm in the form of an abnormally high risk of stroke.”) (emphasis added); Daffin v. Ford Motor Co., 458 F.3d 549 (6th Cir. 2006). In addition, a plaintiff must demonstrate that there are common questions of fact or law. Not any questions will do, however. The plaintiff must show that the class’ claims are based on a common contention that 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 of one of the 22 / 107 claims in one stroke.” Wal-Mart, 131 S.Ct. at 2551; see also Ealy v. Pinkerton Gov’t Servs., Inc., 514 Fed. Appx. 299, 303-04 (4th Cir. 2013). In short, the common questions identified must be dispositive of the claims. If they are not—if, for instance, after the common questions are answered one way or another individualized inquiries to determine liability would be needed— then commonality has not been established. Ahmad v. Old Republic Nat. Title Ins. Co., 690 F.3d 698, 704 (5th Cir. 2012); see also Luiken, 705 F.3d at 377; Forte v. Wal-Mart Stores, Inc., 2012 WL 2886711, at *2 (S.D. Tex. July 13, 2012). The number of common questions does not matter, one will suffice, provided it is critical to resolving the class members’ claims. Wal-Mart, 131 S.Ct. at 2556; In re Am Med. Sys., Inc., 75 F.3d 1069, 1080 (6th Cir. 1996) (internal citations omitted). The Fifth Circuit emphasized this requirement in its MD decision. MD, 675 F.3d at 837; 841. Similarly, the Fifth Circuit held that the potential defenses against the plaintiff’s claims should also be considered as they may undermine the dispositive, “one stroke” nature of the proposed common questions. Id. at 843-44; see also Witt v. Chesapeake Exploration, L.L.C., 276 F.R.D. 458, 468-69 (E.D. Tex. 2011); cf. Schumacher v. AK Steel Corp. Ret. Accumulation Pension Plan, 711 F.3d 675, 683-84 (6th Cir. 2013). In order to determine whether there are common questions of law or fact, a court must trace the class claims and conclude that the common questions will resolve them without the need for additional extensive individualized inquiry. The existence of some variations within the class or individualized defenses do not necessarily make class certification improper. “It is not necessary that members of the proposed class ‘share every fact in common.’” Evon v. Law Offices of Sidney Mickell, 688 F.3d 1015, 1030 (9th Cir. 2012) (quoting Rodriguez v. Hayes, 591 F.3d 1105, 1122 (9th Cir. 2010)). In re 23 / 107 Heartland Payment Sys., Inc. Customer Data Sec. Breach Litig., 851 F.Supp. 2d 1040, 1054 (S.D. Tex. 2012). So long as the common questions linking the putative class members are dispositive of their claim and the claim arises out of a single course of conduct and on a single theory of liability, Rule 23(a)’s commonality requirement is satisfied. Young, 693 F.3d at 543; Evon, 688 F.3d at 1029-30; Wang v. Chinese Daily News, Inc., 709 F.3d 829, 834 (9th Cir. 2013). Post-Wal-Mart, establishing commonality entails two things: that there exists a common policy or practice, possibly an implicit one, that is the alleged source of the harm to class members, and that there are common questions of law or fact that will be dispositive of the class members’ claim. 3. Typicality There is significant overlap between typicality and commonality. Wal-Mart, 131 S.Ct. at 2551 n.5 (quoting Falcon, 457 U.S. at 157-58 n.13). Like commonality, ‘[t]ypicality does not require a complete identity of claims. Rather the critical inquiry is whether the class representative’s claims have the same essential characteristics of those of the putative class. If the claims arise from a similar course of conduct and share the same legal theory, factual differences will not defeat typicality.’ James v. City of Dallas, 254 F.3d 551, 571 (5th Cir. 2001), abrogated on other grounds by MD, 675 F.3d at 832 abrogation recognized In re Rodriguez, 695 F.3d 360, 367 n.9 (5th Cir. 2012) (quoting 5-23 Moore’s Federal Practice – Civil § 23.24); In re Heartland Payment Sys., 851 F.Supp. 2d at 1055 (holding that typicality was satisfied “[d]espite possible state-by-state variations in the elements of these claims,” because “they arise from a single course of conduct by [the defendant] and a single set of legal theories.”); Danvers Motor Co., Inc. v. Ford Motor Co., 543 F.3d 141, 150 (3d Cir. 2008); Marcus v. BMW of N. Am., LLC, 687 F.3d 583, 599 (3d 24 / 107 Cir. 2012); Wright & Miller, 7A Fed. Prac. & Proc. Civ. § 1764 (3d ed.) (“a number of courts have noted that typical claims need not be identical to one another”). Typicality requires a showing that the claims of the named plaintiffs are in fact those asserted as the common class claims. In this sense, typicality is commonality addressed from the perspective of the named plaintiffs. Commonality requires showing that, in fact, all members of the proposed class share a common claim, the validity of which can be determined on a classwide basis. Typicality requires showing that, in fact, the proposed representatives have that claim. Often, once commonality is shown typicality will follow as a matter of course. See, e.g., Frey v. First Nat’l Bank Sw., 2013 U.S. Dist. LEXIS 37153 (N.D. Tex. Feb. 20, 2013). 4. Adequacy of Representation Like typicality, adequacy or representation also tends to merge with commonality, though it “also raises concerns about the competency of class counsel and conflicts of interest.” WalMart, 131 S.Ct. at 2551 n.5 (quoting Falcon, 457 U.S. at 157-58 n.13). Adequacy of representation “‘encompasses class representatives, their counsel, and the relationship between the two.’” Stirman v. Exxon Corp., 280 F.3d 554, 563 (5th Cir. 2002) (quoting Berger v. Compaq Computer Corp., 257 F.3d 475, 479 (5th Cir. 2001)). In evaluating this requirement, a court “must consider ‘[1] the zeal and competence of the representative[s’] counsel and . . . [2] the willingness and ability of the representative[s] to take an active role in and control the litigation and to protect the interests of absentees[.]’” Id. (alterations in original) (quoting Berger, 257 F.3d at 479). Class counsel are fiduciaries of the class and the court must be satisfied that they are prosecuting the case in the interest of the class. See, e.g., Culver v. City of Milwaukee, 277 F.3d 908, 913 (7th Cir. 2002); In re Pharm. Indus. Average Wholesale Price Litig., 588 F.3d 24, 36 n. 12 (1st Cir. 2009). 25 / 107 In the present case Plaintiffs are minors represented by next friends pursuant to Rule 17 of the Federal Rules of Civil Procedure. In such a case courts “consider the individual’s familiarity with the litigation, the reasons that move her to pursue the litigation, and her ability to pursue the case on the child’s behalf” to determine whether or not the next friend is dedicated to the minor’s best interests. Sam M. v. Carcieri, 608 F.3d 77, 92 (1st Cir. 2010). B. Rule 23(b)(2) In the Fifth Circuit, Rule 23(b)(2) has been interpreted to have three components. First, “class members must have been harmed in essentially the same way.” Maldonado, 493 F.3d at 524 (citing Bolin v. Sears, Roebuck & Co., 231 F.3d 970, 975 (5th Cir. 2000)); see also MD, 675 F.3d at 845 (citing Maldonado). This understanding of the Rule is reinforced by Wal-Mart: Rule 23(b)(2) applies only when a single injunction or declaratory judgment would provide relief to each member of the class. It does not authorize class certification when each individual class member would be entitled to a different injunction or declaratory judgment against the defendant … the relief sought must perforce affect the entire class at once. 131 S.Ct. at 2557-58. Given the threshold requirement of commonality, if a plaintiff has satisfied the requirements of Rule 23(a), then the conditions of Rule 23(b)(2) will tend to follow. In establishing commonality the plaintiff will have identified a common practice or policy that is the source of the class members’ harm. So, if she prevails on the merits, a single injunction barring or modifying that course of that behavior will, in the ordinary course of things, provide relief to the members of the class. In contrast, if the requested relief requires particular relief tailored to each class member, then certification under Rule 23(b)(2) is inappropriate. MD, 675 F.3d at 846-47; Shook v. Bd. of Cnty. Comm’rs, 543 F.3d 597, 604 (10th Cir. 2008) (“Shook II”). 26 / 107 Second, “injunctive relief must predominate over monetary damage claims.” Maldonado, 493 F.3d at 524 (citing Bolin, 231 F.3d at 975); accord Wal-Mart, 131 S.Ct. at 2558-59; id. at 2561 (Ginsburg, J., concurring in part and dissenting in part). Damages can be awarded under Rule 23(b)(2), but they must be “incidental,” meaning that they do “not require additional hearings to resolve the disparate merits of each individual’s case,” nor do they “introduce new legal or factual issues, nor entail complex individualized determinations.” WalMart, 131 S.Ct. at 2560 (citing Allison v. Citgo Petroleum Corp., 151 F.3d 402, 415 (5th Cir. 1998)). Third, the injunctive relief must be specific. Maldonado, 493 F.3d at 524 (citing Bolin, 231 F.3d at 975). This requirement is derived from Rule 65(d), which requires that all injunctions be stated specifically and describe in reasonable detail the act or acts restrained or required. Fed. R. Civ. P. 65.(d). In MD, the court of appeals held that a plaintiff must “make an effort” to describe the injunctive relief they request “so that final injunctive relief may be crafted to describe in reasonable detail the acts required.” 675 F.3d at 848 (internal quotation marks omitted) (citing Shook II, 543 F.3d at 605-06). The precise terms of the injunction need not be decided at class certification, only that the class members’ claim is such that a sufficiently specific injunction can be conceived; a plaintiff must present evidence and arguments “sufficient to allow the district court to see how it might satisfy Rule 65(d)’s constraints and thus conform with Rule 23(b)(2)’s requirement.” Shook II, 543 F.3d at 605 n.4 (emphasis added); accord Morrow v. Washington, 277 F.R.D. 172, 198 (E.D. Tex. 2011) (“Plaintiffs have set forth facts suggesting that Defendants’ behavior was generally applicable to the class as a whole, making injunctive relief appropriate. The precise terms of the injunction need not be decided at this stage, only that the allegations are such that injunctive and declaratory relief are appropriate and 27 / 107 that the class is sufficiently cohesive that an injunction can be crafted that meets the specificity requirements of Rule 65(d).”) (emphasis added). IV. Proposed Class Definitions Plaintiffs propose a General Class and four subclasses defined as follows, (D.E. 179 at 78): 1. General Class: all children now, or in the future, in the Texas PMC. 2. Licensed Foster Care (“LFC”) Subclass: all members of the General Class who are now or will be in a licensed or verified foster care placement, except those children now, or in the future, placed in a verified kinship placement.4 3. Foster Group Home Subclass: all members of the General Class who are now or will be in a foster group home. 4. Basic Care GRO Subclass: all members of the General Class who are now or will be in a GRO and who are now or will be receiving solely non-emergency, basic child care services. 5. Unverified Kinship Subclass: all members of the General Class who are now or will be in an unverified kinship placement. The subclasses overlap: all members of the Foster Group Home subclass and the Basic Care GRO Subclass are also members of the Licensed Foster Care Subclass. The membership of the proposed class and each proposed subclass is easily ascertainable. They are all defined in terms of categories used by DFPS. The PMC is well-defined at any given 4 Plaintiffs’ proposed definition of the LFC Subclass could be misleading. As they word it, it sounds as if a future placement in a verified kinship placement somehow removes the child in question from the subclass. A clearer presentation would be: all members of the General Class who are or will be in a licensed or verified foster care placement, excluding verified kinship placements. 28 / 107 time based on the legal status of the child in state custody. Each subclass is defined by the type of DFPS placement the child is currently in along with other categorizations—like being classified as needing only basic child care services—employed by DFPS. V. Plaintiffs’ Fourteenth Amendment Claim Plaintiffs’ claims are all based on the same legal theory, save one claim asserted on behalf of the Licensed Foster Care Subclass discussed below (see Section VIII. infra). They assert a Fourteenth Amendment substantive due process right to a duty of care while in the custody of the State. They allege that Defendants have instituted policies or practices that deprive them of their Fourteenth Amendment right to be free from an unreasonable risk of harm while in the State’s custody. The variation in the General Class and subclass claims consists of the policy that the Plaintiffs allege is causing the harm and the group of children in the PMC they allege are being harmed by that policy. The Supreme Court has held that the Fourteenth Amendment does not generally create an affirmative duty of care on the part of the State to protect citizens from private harm. DeShaney v. Winnebago Dep’t of Soc. Servs., 489 U.S. 189, 195-96 (1989) (“[The Fourteenth Amendment’s] purpose was to protect the people from the State, not to ensure that the State protected them from each other.”). But in so ruling the Court recognized “that in certain limited circumstances the Constitution imposes upon the State affirmative duties of care and protection with respect to particular individuals.” Id. at 198. The first context in which such a right was identified was prisoner litigation. Estelle v. Gamble, 429 U.S. 97 (1976), held that the Eighth Amendment prohibition of cruel and unusual punishment imposed an obligation on the State to provide medical care for those it imprisons. 29 / 107 Id. at 103. While incarcerated, a prisoner is forced to rely on prison officials for her medical needs, and failure to provide for those needs can, in extreme cases, amount to torture and death. Id. at 103-04 (quoting Spicer v. Williamson, 191 N.C. 487 (N.C. 1926)). This right has been extended beyond prisons and the Eighth Amendment to individuals who are involuntarily committed to a mental institution. Youngberg v. Romeo, 457 U.S. 307, 315 (1982). In Youngberg, Pennsylvania had conceded that committed individuals possessed rights to “food, shelter, clothing, and medical care.” Id. The Court reasoned that they also had the right to safe living conditions: “If it is cruel and unusual punishment to hold convicted criminals in unsafe conditions, it must be unconstitutional to confine the involuntarily committed—who may not be punished at all—in unsafe conditions.” Id. at 315-16. The Court has also determined that individuals in police custody who were injured while being apprehended have a due process right to medical care. City of Revere v. Mass. Gen. Hosp., 463 U.S. 239, 244 (1983). The Eighth Amendment had no application in that case as the injured party was not a convicted prisoner. But, the Court paralleled its reasoning in Youngberg: if prisoners have such a right then surely others in State custody have such a right as well. Id.; see also Loe v. Armistead, 582 F.2d 1291, 1294 (4th Cir. 1978) (recognizing protection for pretrial detainees under the due process clause of the Fifth Amendment as at least coextensive with the Eight Amendment protection afforded prisoners). Summarizing this line of cases, DeShaney explained the nature of this right: when the State exercises its power to take custody of an individual it undertakes a constitutional duty to provide for her basic needs because it has taken control over her. DeShaney, 489 U.S. at 200. This right is based on a special relationship between the State and the individual in its custody; the State’s duty arises from its decision to limit the individual’s capacity to act on her own 30 / 107 behalf. The State’s obligations are therefore not to the public at large, but limited to those individuals it exercises control over. Pertaining to the instant case, a number of courts have concluded that this sort of special relationship is formed when the State takes a child away from his or her parents and takes over custody of the child. The same reasoning applies: the State exercises control over children in State custody and has a commensurate duty of care to them. In 1990 the Fifth Circuit stated that while in State custody, the child is entitled to “constitutionally adequate care.” Griffith v. Johnston, 899 F.2d 1427, 1439 (5th Cir. 1990). Defendants point out that Griffith involved determining whether the State has a constitutional duty to children after they leave foster care and concluded that there is no such duty. (D.E. 198, at 12). But, the discussion in Griffith makes sense only presuming a Fourteenth Amendment right exists. While dicta, the Griffith court stated: “TDHS [Texas Department of Human Services] created a ‘special relationship’ with the Griffiths’ children when it removed them from their natural homes and placed them under state supervision. At that time, TDHS assumed the responsibility to provide constitutionally adequate care for these children.” Griffith, 899 F.2d at 1439. The Fifth Circuit in Griffith merely declined to recognize an ongoing special relationship that persists after a child who was once in the State’s care has been adopted. Id. at 1439-40. Indeed, the Fifth Circuit has cited Griffith for just this proposition: that when State social agencies remove a child from her natural home and place her under State supervision a special relationship is created. Doe v. Taylor Indep. Sch. Dist., 975 F.2d 137, 146 (5th Cir. 1992) reh’g granted and opinion vacated, 987 F.2d 231 (5th Cir. 1993) and on reh’g, 15 F.3d 443 (5th Cir. 1994). DeShaney, which Griffith’s logic is based on, is similar: the Court held that the State had no obligation to protect the child from his father because, though it had taken custody of the child before, it had since returned him to his father’s 31 / 107 custody. DeShaney, 489 U.S. at 201. The Court recognized that if DeShaney had been in the custody of the State rather than that of his parents it would be a different case and might be “sufficiently analogous to incarceration or institutionalization.” Id. at 201 n.9. The Fifth Circuit recognized this right more explicitly in Hernandez v. Texas Department of Protective & Regulatory Services, 380 F.3d 872, 880 (5th Cir. 2004). Hernandez held that foster children enjoy a substantive due process right “to personal security and reasonably safe living conditions” based on the special relationship between them and the State. Id. The existence of the right was not disputed on appeal, and the district court had found that its existence was clearly established for the purposes of qualified immunity. Id. (citing Hernandez v. Tex. Dep’t of Protective & Regulatory Servs., 2002 WL 31689710 (N.D. Tex. Nov. 22, 2002)); see also Doe ex rel. Magee v. Covington Cnty. Sch. Dist. ex rel. Keys, 675 F.3d 849, 859 (5th Cir. 2012) (recognizing foster care as one of the “strictly enumerated” situations where the State assumes a duty of care sufficient to create a special relationship). The Fifth Circuit is not alone. While noting that there are important differences between prisoners and foster children, the Third Circuit found the analogy persuasive. Consistent with DeShaney, it held that the salient feature that creates duties on the part of the State is that it “by affirmative act, renders the individual substantially ‘dependent upon the state … to meet [his or her] basic needs.’” Nicini v. Morra, 212 F.3d 798, 808 (3d Cir. 2000) (quoting D.R. v. Middle Bucks Area Vocational Technical Sch., 972 F. 2d 1364, 1372 (3d Cir. 1992)). The Second Circuit has held that there is a constitutional right not to be placed in a setting known to be unsafe. Doe v. N.Y. Dep’t of Soc. Servs., 649 F.2d 134, 143 (2d Cir. 1981), after remand, 709 F.2d 782. On the basis of Youngberg, the Eleventh Circuit also has recognized a substantive due process liberty interest in reasonably safe living conditions for foster children. Taylor v. 32 / 107 Ledbetter, 818 F.2d 791, 795 (11th Cir. 1987). Similarly, the Sixth Circuit has recognized that “due process extends the right to be free from the infliction of unnecessary harm to children in state-regulated foster homes.” Lintz v. Skipski, 25 F.3d 304, 305 (6th Cir. 1994) (quoting Meador v. Cabinet for Human Res., 902 F.2d 474, 476 (6th Cir. 1990)). Distinguishing a case where a child was injured by a foster parent the State placed her with from DeShaney, the Seventh Circuit concluded that when the State had taken the child away from her parents and into its custody, it could no more place a foster child in a situation of danger than it could a prisoner. K.H. through Murphy v. Morgan, 914 F.2d 846, 848-49 (7th Cir. 1990). Indeed, “courts have found, with apparent unanimity, that [a special] relationship exists in the foster-care context.” Connor B. v. Patrick, 771 F.Supp. 2d 142, 160 (D. Mass. 2011). Children in the State’s care have a right to be free from an unreasonable risk of harm that would compromise their personal security or deprive them of reasonably safe living conditions. Hernandez, 380 F.3d at 880. This substantive due process right “encompasses a right to protection from psychological as well as physical abuse.” R.G. v. Koller, 415 F.Supp. 2d 1129, 1156 (D. Haw. 2006). Similarly, Marisol A. held that children in foster care “have a substantive due process right to be free from unreasonable and unnecessary intrusions into their emotional well-being.” Marisol A. by Forbes v. Giuliani, 929 F.Supp. 662, 675 (S.D.N.Y. 1996) aff’d sub nom. Marisol A. v. Giuliani, 126 F.3d 372 (2d Cir. 1997) (citing B.H. v. Johnson, 715 F.Supp. 1387, 1395 (N.D. Ill. 1989)); accord K.H., 914 F.2d at 848 (“The extension to the case in which the plaintiff’s mental health is seriously impaired by deliberate and unjustified state action is straightforward.”); Aristotle P. v. Johnson, 721 F.Supp. 1002, 1010 (N.D. Ill. 1989) (“The fact that the plaintiffs’ injuries are psychological rather than physical is of no moment.”). 33 / 107 Harm, both psychological and physical, can be inflicted in a variety of ways, including neglect, physical abuse, sexual abuse, and psychological maltreatment. See, e.g., (PX 14 at vii; PX 435 § 6000-2). The K.H. court, relying on Youngberg, provided the following general formulation for the State’s obligations: “the Constitution requires the responsible state officials to take steps to prevent children in state institutions from deteriorating physically and psychologically.” 914 F.2d at 852; see also id. at 849 (“Once the state assumes custody of a person, it owes him a rudimentary duty of safekeeping no matter how perilous his circumstances when he was free.”). In the context of the present case, this characterization of substantive due process rights is especially important as children in the PMC are practically guaranteed to have experienced significant trauma before entering the State’s care. As one presentation by DFPS summarized: “Because children in the conservatorship of DFPS are victims of abuse and/or neglect, the trauma they have endured can result in significant treatment needs. These children experience not only trauma, but great loss by being removed from their homes, schools and friends. Also, the children often have serious underlying medical and developmental needs.” (PX 322 at 11). The right described in the previous paragraphs is much more limited than those rejected in Griffith. The Griffith plaintiffs asserted that the State had an obligation “to have maximized their personal psychological development” and to place them in “unrestrictive circumstances.” Griffith, 899 F.2d at 1438-40. The cases cited above only hold the State cannot treat the marked deterioration of those in its care with cavalier indifference. Griffith also differs from the instant case as the plaintiffs in Griffith had been adopted, and so were no longer in the State’s care. Id. at 1439-40. 34 / 107 Those entitled to this duty of care do not need to wait to suffer an actual harm in order to obtain relief. This has been made most explicit in the prison context, leading the right to be characterized as the right “not to be subjected to the unreasonable threat of injury.” Hoptowit v. Spellman, 753 F.2d 779, 784 (9th Cir. 1985) (“Prisoners have the right not to be subjected to the unreasonable threat of injury or death by fire and need not wait until actual casualties occur in order to obtain relief from such conditions.”); see also Leeds v. Watson, 630 F.2d 674, 675-76 (9th Cir. 1980) (noting that the infrequent use of a room that poses a fire hazard “does not lessen the hazard it presents to the lives of persons incarcerated there.”); Battle v. Anderson, 564 F.2d 388, 395 (10th Cir. 1977); Coniglio v. Thomas, 657 F.Supp. 409 (S.D.N.Y. 1987) (finding that an absence of smoke barriers and an effective system of smoke management in a prison unit violated the right of the class of prisoners housed there not to be subjected to an unreasonable risk of death or injury by fire). Prisoners have a right to be free from an unreasonable threat of injury and they can suffer a legal injury if that right is violated by, for example, unreasonably unsafe fire safety conditions, without the risk becoming actualized and causing physical injuries. Having suffered the legal injury they can obtain relief even if, in the fire safety cases, there has been no fire at the prison.5 The principle here is similar to the one in Whirlpool and Butler v. Sears described in Section III.A.2., supra. The legal injury is the risk of harm. So, the Fourteenth Amendment right upon which Plaintiffs base their claim can be characterized as a right to be free from an unreasonable risk of harm while in the State’s custody. Again, in the words of Hernandez, children in foster care have “a constitutional right to personal security and reasonably safe living conditions.”6 Hernandez, 380 F.3d at 880; see also Hall v. 5 Though the occurrence of a fire or fires at the prison could be used as evidence that the risk exists. Defendants argue that the harm Plaintiffs base their claim on is too attenuated to give rise to a cause of action. E.g., (D.E. 163 at 28); (D.E. 198 at 11 n.4). Defendants’ argument misconstrues the contours of the Fourteenth Amendment right possessed by an individual in the State’s custody. The right itself is to be free from the 6 35 / 107 Smith, 497 Fed.Appx. 366, 377 (5th Cir. 2012). Plaintiffs do not need to wait until they are actually abused or suffer from unsafe living conditions. They have a right to be free of the unreasonable risk of harm, and if they suffer that risk now, they have suffered a legal injury. See D.G. v. Yarbrough, 278 F.R.D. 635, 638 (N.D. Okla. 2011) (“Thus, the court is mindful that the issue here is not what percentage of children have actually suffered abuse while in state custody, but whether and to what extent foster children are at imminent risk of serious harm.”) (emphasis in original). The same right described a different way is that Plaintiffs are entitled to safeguards that prevent them from deteriorating either physically or psychologically. To succeed in a substantive due process claim of this sort a plaintiff must show that the State has acted in a way such that it has breached the duty of care that it owes to those in State custody, and that the State action that did so rises to the requisite level of culpability. In making this determination courts sometimes ask whether the complained of conduct “shocks the conscience.” Cnty. of Sacramento v. Lewis, 523 U.S. 833, 846 (1998); Rochin v. California, 342 U.S. 165, 172 (1952). In this case Plaintiffs challenge policies rather than discrete executive actions, so on this test they would need to show that those policies are such that they shock the conscience. This standard, however, can be difficult to apply as it is hard to flesh out a more concrete rule to apply uniformly. See, e.g., Rochin, 342 U.S. at 179 (Douglas, J., concurring). Courts have articulated culpability standards to further define the test, two of which—deliberate indifference and absence of professional judgment—the parties here have disputed. (D.E. 195, at 48-49). In the Fifth Circuit, a plaintiff generally must prove deliberate indifference, which requires showing that there was conscious disregard of a known and excessive risk to the unreasonable risk of harm. That risk need not be realized in order for a claim to be actionable—a plaintiff must rather show that the policy in question creates a risk that rises to the level of an unreasonable risk of harm. Whether or not Plaintiffs here can succeed in showing this is a question for the merits. 36 / 107 victim’s health or safety. Hernandez, 380 F.3d at 880 (citing Farmer v. Brennan, 511 U.S. 825, 837 (1994); McClendon v. City of Columbia, 305 F.3d 314, 326 n.8 (5th Cir. 2002)); see also Hall, 497 Fed. Appx. at 377 n.16; Smith v. Tex. Dep’t of Family & Protective Servs. Child Protective Servs., 2009 WL 2998202 (W.D. Tex. Sept. 15, 2009). Deliberate indifference is determined by a subjective standard of recklessness: the defendant “‘must both be aware of facts from which the inference could be drawn that a substantial risk of serious harm exists, and he must draw the inference.’” Smith v. Renoettsy, 158 F.3d 908, 912 (5th Cir. 1998) (quoting Farmer, 511 U.S. at 837). “Deliberate indifference is a degree of culpability beyond mere negligence or even gross negligence; it must amount to an intentional choice, not merely an unintentional oversight.” Hall, 497 Fed. Appx. at 377 (quoting James v. Harris Cnty., 577 F.3d 612, 617-18 (5th Cir. 2009)). The state of mind can be inferred, however, if the risk of harm is obvious. Hernandez, 380 F.3d at 881 (citing Hope v. Pelzer, 536 U.S. 730, 738 (2002)); see also Farmer, 511 U.S. at 842. To succeed on the deliberate indifference standard Plaintiffs would have to show that the systematic deficiency causes an unreasonable risk of harm and that Defendants had actual or constructive knowledge of that risk, or alternatively that such knowledge should be inferred from the obviousness of the risk. Defendants can defeat the claims by showing there is no such deficiency, there is no such risk, or they have no actual or constructive knowledge of that risk. The question on the merits would be whether or not the policy or practice Plaintiffs allege creates an unreasonable risk of harm is such that Defendants, in continuing it by action or inaction, are deliberately indifferent to class members’ rights to personal security and reasonably safe living conditions. 37 / 107 The alternative standard, absence of professional judgment, is described in Youngberg. A decision “if made by a professional, is presumptively valid; liability may be imposed only when the decision by the professional is such a substantial departure from accepted professional judgment, practice, or standards as to demonstrate that the person responsible actually did not base the decision on such a judgment.” Youngberg, 457 U.S. at 323. This standard has been employed by other courts in the foster care context. See, e.g., Yvonne L. v. N.M. Dep’t Human Servs., 959 F.2d 883, 894 (10th Cir. 1992); K.H. 914 F.2d at 853-54; Brian A. by Brooks v. Sundquist, 149 F.Supp. 2d 941, 953 (M.D. Tenn. 2000); T.M. ex rel. Cox v. Carson, 93 F.Supp. 2d 1179 (D. Wyo. 2000); Wendy H. By & Through Smith v. Philadelphia Dep’t of Human Servs., 849 F.Supp. 367 (E.D. Pa. 1994); LaShawn A. v. Dixon, 762 F.Supp. 959 (D.D.C. 1991) aff’d and remanded on other grounds sub nom. LaShawn A. by Moore v. Kelly, 990 F.2d 1319 (D.C. Cir. 1993). Plaintiffs also argue that it should be employed in this case for injunctive relief, since other courts have employed it in lieu of the deliberate indifference standard when no money damages are sought. (D.E. 164 at 17-18); see, e.g., LaShawn A., 762 F.Supp. at 996 n.29. The standard described above is often referred to as simply the professional judgment standard, but such labeling can be misleading. It is not the case that the United States Constitution compels Texas to adopt a particular standard to organize its foster care system around. Instead, under the professional judgment standard Plaintiffs must show that an alleged systematic deficiency is such a departure from professional standards and practice that it cannot have been based on professionals exercising their professional judgment. This is not to say that various professional standards may have no bearing on the case: compliance or lack of compliance with a professional standard, if admissible, can be evidence for or against finding a constitutional violation. Such standards are just not the substance of a 38 / 107 constitutional duty. A substantial departure from all relevant professional standards can also indicate that a professional was subjectively aware of and opted to disregard a risk. Put another way, accepted professional standards can make a risk of harm obvious in the manner Hernandez refers to. Indeed, though the two standards are different in form, in practice they are not so different. Yvonne L., 959 F.2d at 894 (“As applied to a foster care setting we doubt that there is much difference in the two standards”). Plaintiffs argue that it is not necessary at this point in the litigation to determine which standard to apply. (D.E. 195, at 48-49). The Court agrees as the issue has not been fully briefed, and it is not necessary to decide the appropriate standard to employ on the merits in order to determine the appropriateness of class certification. In considering class certification the relevant question is whether the Plaintiffs’ claim, under either standard, will turn on common questions of fact or law. In order to conduct a rigorous analysis of the Motion for Class Certification it has been necessary to flesh out the Plaintiffs’ constitutional claim. Plaintiffs allege systematic deficiencies related to the General Class and each subclass. Their claims are based on class members’ Fourteenth Amendment rights as foster children in State custody to personal security, to reasonably safe living conditions, and to have steps taken to prevent their physical or psychological deterioration. As the claims under consideration here are based on these Fourteenth Amendment rights, those rights define, and to the extent relevant limit, the scope of Plaintiffs’ claims. For Plaintiffs to be entitled to class certification with regards to these claims the policies or practices they identify must allegedly impair the Fourteenth Amendment rights described above. For example, Plaintiffs can make out a claim that a given policy instituted by Defendants subjects children in foster care to an unreasonable risk of psychological harm. 39 / 107 VI. Evidence Before the Court In the course of its rigorous analysis of Plaintiffs’ motion for class certification, the Court has reviewed numerous exhibits submitted by the parties and held a three day Class Certification Hearing on January 22-24, 2013. The evidence falls into three broad categories. First, the Court heard testimony and received studies prepared by experts on child welfare management, administration, and evaluation of child welfare systems, which are described in more detail below. Second, the Court heard testimony and excerpts of depositions from persons who are involved in Texas foster care, such as employees of DFPS, private contractors who provide monitoring and review services to the State, and next friends of the named plaintiffs. Third, the Court was presented with numerous DFPS documents, including excerpts of the named plaintiffs’ case files, internal reports, appropriations requests to the Texas legislature, and communications amongst DFPS personnel. Viola Miller was called as an expert witness by Plaintiffs at the class certification hearing. Dr. Miller has a doctorate in special education. From 2003 to 2011 she served as Tennessee’s Commissioner of the Department for Children Services, a state cabinet-level position where she reported to the governor. Prior to that post, from 1995 to 2003, she was the Cabinet Secretary for the Kentucky Department of Families and Children, a comprehensive human services agency in that state. (PX 9B at 64). Dr. Miller thus has experience running two state-wide child welfare agencies. She has also been accepted as an expert in litigation similar to the instant case. D.G. ex rel. Strickland v. Yarbrough, 278 F.R.D. 635 (2011). For the purposes of this litigation Dr. Miller reviewed numerous documents produced by DFPS, datasets, and depositions. Class Cert. Hr’g 1/22/13 (Miller) at 147; (PX 9A). Dr. Miller had access to an 40 / 107 array of sources, though she did not have access to a sample of case reviews. Class Cert. Hr’g 1/22/13 (Miller) at 157. Defendants objected to Dr. Miller’s status as an expert on child welfare management, administration, and reform on two grounds. First, they objected to Dr. Miller drawing conclusions on the basis of her experiences in Tennessee and Kentucky, arguing that these observations should not be applied to Texas. E.g., (D.E. 198 at 48-50); Class Cert. Hr’g 1/22/13 (Miller) at 158, 164-66. But, Defendants identified no special characteristics of Texas—such as size or demography—that would render Dr. Miller’s extensive experience with child welfare systems inapplicable. Second, Defendants objected on the basis that Dr. Miller’s opinions are informed by the national standards articulated by the Child Welfare League of America (“CWLA”) and the Council on Accreditation (“COA”), and that since those standards are not, as a matter of law, the ones that determine liability under the Fourteenth Amendment, the testimony was not relevant to the issues before the Court. Class Cert. Hr’g 1/22/13 (Miller) at 177. As explained above, professional standards such as these do not control liability in a Fourteenth Amendment case. Although it would also be inaccurate to say that such standards are irrelevant; a finding that the State adhered to one of these standards would shield it from liability under Youngberg’s absence of professional judgment standard, for instance. That Dr. Miller relied, to some extent, on accepted professional standards in assessing Texas’ foster care system does not eliminate her years of experience in this field or render her observations irrelevant, especially since she was called upon to evaluate the performance of the Texas child welfare system and not its constitutionality. The Court accepted Dr. Miller as an expert in the fields offered. Id. at 167, 177. 41 / 107 Daryl Kennedy Chansuthus has a Master of Science in Social Work and also has extensive experience in the child welfare systems. She has held several posts in Tennessee where she evaluated that state’s child welfare agencies and the services they provide. She monitored Tennessee’s compliance with the Brian A. Settlement Agreement, which touched upon all aspects of child welfare services, including caseworker workloads and training. Class Cert. Hr’g 1/23/13 (Chansuthus) at 259; see also (PX 21A). In particular, she developed protocols that determine what the agency’s actual practices are from reports and raw data. Class Cert. Hr’g 1/23/13 (Chansuthus) at 260. She later expanded these practices as the Tennessee Department of Children’s Services’ Director of Quality Assurance and Quality Improvement, creating comprehensive systems to monitor and evaluate child welfare management and practice. Id. at 262-63; (PX 21A at 2). From 2007 to 2011 she served as the Executive Director of the Tennessee Center for Child Welfare, where she continued to develop training and content analysis programs. Class Cert. Hr’g 1/23/13 (Chansuthus) at 265-66; (PX 21A at 1). Ms. Chansuthus thus has expertise in evaluating child welfare systems. For this litigation Ms. Chansuthus prepared an expert report focusing on whether RCCL can adequately and consistently perform its oversight duties. Her conclusions were based on an analysis of documents provided by DFPS, including communications between DFPS personnel, workload data, and DFPS’s own assessments, drawing from her years of experience in the field. See (PX 21B; PX 21 at 3); Class Cert. Hr’g 1/23/13 (Chansuthus) at 268-69. Defendants incorrectly assert that Chansuthus’ expert report, Plaintiffs’ Exhibit 21, is not part of the record for this case. (D.E. 198 at 7). Plaintiffs’ Exhibit 21 was entered into evidence. Class Cert. Hr’g 1/23/13 (Chansuthus) at 277. 42 / 107 Plaintiffs offered the testimony of two other experts. Jill Duerr Berrick was accepted as an expert on child welfare research with a focus on foster care. Class Cert. Hr’g 1/23/13 (Berrick) at 15; see also (PX 8A (Dr. Berrick’s curriculum vitae)). She prepared a review of the literature concerning the effects of caseload size on child welfare outcomes. Class Cert. Hr’g 1/23/13 (Berrick) at 11. The Court granted Defendants’ motion to exclude Dr. Berrick’s testimony on Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993), grounds; the Court is capable of reviewing the literature, had it been included in the record, and discerning trends in the research. Class Cert. Hr’g 1/23/13 (Berrick) at 51-52; (D.E. 173 at 6 (citing United States v. Paul, 175 F.3d 906, 912 (11th Cir. 1999)). Plaintiffs’ expert Donald Warren calculated weighted caseloads for DFPS caseworkers based on data provided by Defendants. Texas records caseloads idiosyncratically, though, see Section VII.B., infra, and Dr. Warren’s analysis did not adequately take this into account. Class Cert. Hr’g 1/23/13 (Warren) at 124-25, 318. So, the caseloads that Dr. Warren calculated do not offer much insight. In addition, Plaintiffs have not established any connection between a particular caseload number—either CWLA’s recommended maximum of 15 cases, COA’s of 18, or the State’s target caseload of about 25, see (D.E. 195 at 9; PX 234 at DFPS002413318)—and a derogation of class members’ Fourteenth Amendment rights that are the basis of the claims in this case. VII. General Class At this point, Plaintiffs’ allegations on behalf of the General Class are not based on the particular, tragic, experiences of the named plaintiffs. “[E]ven a ‘catalog of disquieting events is not sufficient to demonstrate a pervasive pattern,’” Lightfoot v. District of Columbia, 273 F.R.D. 43 / 107 314, 321-22 (D.D.C. 2011) (quoting Carter v. District of Columbia, 795 F.2d 116, 123-24 (D.C. Cir. 1986)), in and of itself. Instead, Plaintiffs assert that the Texas PMC has a structural deficiency of overburdening caseworkers with excessive caseloads. (D.E. 195 at 8). Plaintiffs contend that this subjects children in the PMC to a risk of being assigned a caseworker who has a caseload that is so high that she cannot fulfill her duties to the child, leading to too few visits to monitor the child’s well-being, inappropriate placements in the foster care system, failure to meet the child’s needs, and failure to develop and implement a plan for a permanent home. These conditions are alleged to combine to deny the members of the General Class their constitutional right to personal security and reasonably safe living conditions. A. Numerosity On March 31, 2011 there were a total of 12,196 children in the PMC. (PX 2 at APPX 2077). This easily satisfies the numerosity requirement of Rule 23 described above and is not disputed by Defendants. And, as the children in the class are located throughout the State and their locations will change as placements change, which happens frequently among children in the PMC, Class Cert. Hr’g 1/23/13 (Gonzalez) at 71-72; id. (Liebgott) at 99-100; (PX 250), joinder is impractical. The Court finds the General Class satisfies the numerosity requirement of Rule 23(a)(1). Similar observations apply to three of Plaintiffs’ proposed subclasses.7 On August 11, 2011 there were 8,174 children in the Licensed Foster Care Subclass. (PX 2 at APPX 2077). On May 31, 2012 there were roughly 900 children in the Foster Group Home Subclass. (PX 298 (filed under seal)). And, on May 31, 2012 there were nearly 400 children in the Basic Care GRO Subclass. (PX 299 (filed under seal)). Like the members of the General Class, the members of 7 Numerosity is not discussed as to the Unverified Kinship Subclass, considered later. 44 / 107 these subclasses are located throughout the State, making joinder impractical. The Court concludes that these subclasses are also sufficiently numerous under Rule 23(a). B. Commonality In order to establish commonality, Plaintiffs must establish that there is a policy or practice on the part of the Defendants that is the source of the General Class’ alleged injury, and also that the claim arising out of that injury depends on common questions of law and fact. Even if Texas caseworkers are overworked, class certification would not be appropriate if these workloads were, for instance, the product of disparate, uncoordinated decisions by supervisors. This was the problem the Supreme Court identified with the putative class in Wal-Mart. That case involved a far-flung class action where the plaintiffs’ claims were based on sexual discrimination, namely denial of equal pay and promotions. Wal-Mart v. Dukes, 131 S. Ct. 2541, 2547 (2011). Pay and promotion decisions, however, were made by local store managers, and the overarching company gave them broad discretion. The class was held to not satisfy commonality and therefore denied certification because this discretion was the only corporate policy the plaintiffs could establish. As the Supreme Court pointed out, this is essentially the opposite of having any uniform policy, let alone a discriminatory one. Id. at 2554. Plaintiffs’ claim here is different. On behalf of the General Class, Plaintiffs make out a claim that caseworkers carry excessive caseloads, which results in class members being deprived of their Fourteenth Amendment rights. The workload of the State’s caseworkers—whether excessive or not—is the product of deliberate choices made by State actors. Naturally, the number of caseworkers the State employs affects their respective workloads, and this is a policy adopted by the State. Similarly, Texas has opted not to establish any guidelines regarding 45 / 107 maximum or optimum caseloads, as Colleen McCall, Director of Field Operations for Child Protective Services, explained when she was deposed. Class Cert. Hr’g 1/22/13 (McCall) at 334 (“We do not have caseload limits.”); id. at 335 (stating that there are no DFPS guidelines for an optimal caseload). DFPS also does not have a practice of taking into account a child’s special needs when considering caseloads, or at least there is no evidence presently before the Court that it does so. Id. (Miller) at 186. A policy or practice need not be officially adopted. As described above, a policy can be established through custom or through failing to act in the face of actual or constructive knowledge that there is likely to be a violation of constitutional rights. See Section III.A.2., supra. There is ample evidence on the record presently before the Court, described below, that caseworkers are overburdened, that this might pose risks to the children in the PMC, and that DFPS and other State officials had actual or constructive knowledge of these risks and have not acted to cap or otherwise limit caseloads. In addition, this case differs from Wal-Mart as DFPS caseworkers do not act with the sort of free-wheeling discretion that Wal-Mart managers enjoy in making employment decisions. A caseworker’s placement decisions, to take one example, must be approved and explained, see (PX 27), and caseworkers are closely supervised, Class Cert. Hr’g 1/22/13 (McCall) at 354 (stating that the ratio of caseworkers to supervisors is about one to six). The Court therefore concludes that the workloads of DFPS caseworkers is a policy or practice as contemplated by Lightfoot. The Court now turns to the more complex issue of whether there is evidence in the record indicating that the workloads of caseworkers are excessive and whether those workloads are connected to the alleged harm. There is some inevitable overlap between these inquiries as whether or not caseworkers can carry out their 46 / 107 tasks, especially safeguarding and monitoring the children in their care, is key evidence as to whether they are overburdened or not. The workloads of Texas caseworkers is still something of an open question, even with the fairly extensive record before the Court. Texas calculates caseloads in terms of “stages,” each of which represents an aspect of the work that needs to be done with a child or her family. E.g., Class Cert. Hr’g 1/24/13 (Deckinga) at 30-32; (PX 430 at 9; DX 24). Texas stages include Intake, Investigation, Family Preservation, Child Substitute Care (relating to children removed from the home), Family Substitute Care (a stage created for families when a child has been removed from the home), Foster and Adoptive Home Development, Kinship, and Adoption. (PX 43). The same child, then, could represent several stages simultaneously. She could be in the Child Substitute Care and Adoption stages while her family was in the Family Substitute Care one. As Audrey Deckinga, Assistant Commissioner of the Child Protective Services Division of DFPS noted: “There could be conservatorship sub care and adoption stages for many of these children.” Class Cert. Hr’g 1/24/13 (Deckinga) at 32. It is difficult, then, to understand just what the caseloads for Texas caseworkers are and to compare them to professional standards and leading research. Although, as explained in Section V., such standards do not constitute the constitutional standard for liability, a point which Defendants have urged, see e.g., (D.E. 198 at 14; D.E. 163 at 44 (citing Braam v. State, 81 P.3d 851, 861 (Wash. 2003)); id. at 66; Class Cert Hr’g 1/22/13 at 57, 72 (opening statements). At the Class Certification Hearing, the State was able to produce rough average caseload numbers in terms of children rather than stages. The average caseload on August 31, 2011 was 20.59 children and the average on August 31, 2010 was 18.83. Class Cert. Hr’g 1/24/13 (Stephens) at 272. However, these numbers do not separate out TMC and PMC children, and 47 / 107 there is extreme variation in caseworker workloads. In February 2012 the State-wide average was 33.9 stages per caseworker, but caseworkers ranged from carrying 1 to over 100 stages. (PX 243); see also (PX 309 (as of June 30, 2012 caseworker caseloads ranged from 6 to 84 stages) (filed under seal)). This range is the rule rather than the exception: in the data before the Court, since 2009 some caseworkers have carried less than 10 stages while others have simultaneously carried more than 60. (PX 230). Extreme variation makes an average caseload measure less useful. The class member whose caseworker is responsible for only a handful of stages is not at risk due to caseworker overwork while the one whose caseworker is responsible for 95 stages might very well be. An average masks these differences. That being said, there is considerable evidence in the record indicating that DFPS caseworkers carry excessive caseloads. Dr. Miller offered expert testimony concluding that Texas caseloads “are quite high and they are certainly out of compliance with national standards. And work that the State itself has done, DFPS has done, has indicated their awareness” of this. Class Cert. Hr’g 1/22/13 (Miller) at 183. The record contains numerous reports from DFPS and other State agencies indicating that caseworkers cannot handle the volume of work they are responsible for. In 2009, Texas created the Adoption Review Committee to, inter alia, conduct “an extensive review of the foster care system … and to develop ways to improve the foster care system by reducing the time a child is in the conservatorship of [DFPS] before being permanently placed.” (PX 323 at PLTF0019500). In its report, the Adoption Review Committee found that while those involved in the foster care system are well-intentioned and hard-working “the system in which they are operating is unable to effectively handle the volume and deeply complex natures of children brought into state care.” Id. at PLTF0019499. Among the Committee’s findings were “CPS caseworkers carry extremely high caseloads, often twice what 48 / 107 is deemed best practice.” Id. at PLT0019506; see also id. at PLTF0019507 (recommending a reduction in caseloads). Unfortunately, these conditions have been long-standing ones in Texas. In 1996 a similar committee made the same recommendation regarding reducing caseloads. Id.; Class Cert. Hr’g 1/24/13 (Deckinga) at 24-26. And, since the Adoption Review Committee’s report was published in late 2010, caseloads have increased. Class Cert. Hr’g 1/24/13 (Deckinga) at 26. Further, Gail Gonzalez, the Director of Placement, Foster Adoptive Home Development, agreed that caseworkers carried high caseloads. Class Cert. Hr’g 1/23/13 (Gonzalez) at 70. There is evidence that DFPS is also significantly understaffed. A DFPS work measurement time study conducted in 2010 found that the agency needed approximately 200 additional conservatorship stage workers. (PX 234 at 4). In fiscal year 2012, roughly 650 DFPS positions that were authorized went unfilled. (PX 39). Likewise, in its budget requests to the Texas legislature DFPS has expressed a need for additional staff, lest caseloads increase. “Without additional staff, caseloads would increase, which results in significant child and adult safety issues.” (PX 12 at Administrator’s Statement 2) (emphasis added); see also (PX 13 at 652 (noting that unless 50 full-time staff were added to conservatorship workers caseloads would increase from the projected 28.8 per worker to 29.5 per worker)). Workers involved in kinship care seem to be especially understaffed. In its August 16, 2012 appropriations request, setting the budget for fiscal years 2014 and 2015, DFPS stated that “staffing to support kinship placement has not grown since [fiscal year] 2009 and is now insufficient to properly assist kinship caregivers.” (PX 13 at 662). The agency requested additional funding to add more caseworkers, reducing the “very high” caseload of 47.2 per worker (projected to increase to 53.8 in fiscal year 2015) to 40. Id. In addition, supervisors working in kinship placements were 49 / 107 responsible for 12 workers, and DFPS asked for funds so that it could reduce this number to 7. Id. The increasing staffing and funding needs of DFPS seem to be driven by population increase, there has been an 18% increase in the Texas child population over the past 10 years, Class Cert. Hr’g 1/24/13 (Deckinga) at 39, and decreasing federal support, (PX 13 at 649-50; PX 319 at Administrator’s Statement 1-2). There is also evidence of substantial turnover in DFPS personnel. The agency has lost, on average, 28% of its caseworkers annually. (PX 107 at 4). The turnover rate has ranged from 23.6% in fiscal year 2009 to a staggering 34.1% in fiscal year 2007. Id. In fiscal year 2012, the latest data presented to the Court, DFPS caseworkers had a turnover rate of just over 24%. (PX 387). Turnover is especially acute with entry-level caseworkers at approximately 38% in fiscal year 2012. (DX 49 at DFPS005158063; PX 387); Class Cert. Hr’g 1/22/13 (McCall) at 346 (agreeing with the statement that “almost 36 percent of all your entry-level workers leave each year”); Class Cert. Hr’g 1/24/13 (Deckinga) at 63-64 (stating that entry-level turnover was about 37%). Turnover has been a serious area of concern for DFPS, as Ms. McCall explained: “Turnover is something that we’re always concerned about what we can do to impact it …” Class Cert. Hr’g 1/22/13 (McCall) at 347. In fall 2012 DFPS completed a Human Resources Management Plan as required by the Texas legislature to attempt to improve employee retention. (DX 49). The Plan noted that “agency-wide turnover has increased in [fiscal year] 2012 … we continue to experience high rates of turnover.” Id. at DFPS005158067; see also (PX 319 at PLTF0007007 (noting the importance of staff retention)). High rates of staff turnover are both indicative of and exacerbate overworked personnel. In the words of a report by the U.S. Department of Health and Human Services: “Heavy caseloads and workloads have been cited repeatedly as key reasons that workers leave the child 50 / 107 welfare workforce.” (PX 340 at JDB0000354-55 (internal citations omitted)); see also id. at JDB0000357 (describing turnover as “both a consequence and a cause of high workloads”). More specifically applicable to Texas, the Adoption Review Committee found that being overworked was by far the most popular reason for quitting DFPS—approximately 45% listed it as the reason for quitting, with supervisors being the next most popular at about 28%. Decreased workload was the most popular suggestion for improvement—making up 50% of the sample as compared to work environment, the next most popular suggestion, making up only 29%. (PX 323 at PLTF0019527). In a 2012 survey 38% of DFPS employees said that the amount of work they are asked to do is unreasonable and 35% did not feel that the work environment supported a balance between work and personal life. (DX 49 at DFPS005158064). Along similar lines, DFPS explained in its August 2010 Legislative Appropriations Request that “Protective services is a stressful job, made even more so when caseloads are high,” (PX 319 at PLTF0007007), and that “High caseloads lead to high worker turnover, further exacerbating high caseloads,” id. at PLTF0007562. Dr. Miller had analogous experiences with the Tennessee and Kentucky child welfare systems: “as we added caseworkers and reduced caseloads, we were able to significantly reduce those turnover rates.” Class Cert. Hr’g 1/22/13 (Miller) at 191-92. Newly-hired caseworkers must be trained and cannot take on the full responsibilities of an experienced caseworker, so turnover exacerbates the existing workload. E.g., (PX 320 at PLTF0007562); Class Cert. Hr’g 1/22/13 (McCall) at 347 (“as folks are newer in their careers they’re probably not as skilled as the folks who have been around … longer … we do believe the longer people stay with us, the better they get at what they do.”). DFPS caseworkers begin with a 12-week training program, and it takes two years, during which time they are learning about the job, for them to carry a full caseload. (DX 49 at DFPS005158064). Thus, even if DFPS 51 / 107 were able to easily replace all of the personnel it lost due to turnover, the agency’s effectiveness would still be impaired as the novice caseworkers would need to acquire the necessary skills. And, that assumes a best-case scenario that the position does not remain vacant for long. While there is a vacancy, of course the caseloads for all the remaining caseworkers increase correspondingly. E.g., (PX 162 (e-mail stating that due to leave, turnover, vacant positions, and an employee being new the remaining caseworkers would have “high” workloads)). Further, turnover makes a caseworker’s job, in particular, more difficult. Monitoring a child’s safety and well-being requires the caseworkers to build a trusting relationship with the child. Class Cert. Hr’g 1/22/13 (Miller) at 191; see also (PX 323 at PLTF0019506 (“Children have a better chance of achieving permanency quickly if they have only one caseworker during their stay in foster care.”)). A “revolving door” of caseworkers makes this rapport difficult to establish, especially because all PMC children have had their lives destabilized by being removed from their homes and, in some cases, their communities. Named plaintiff S.A. provides an extreme example: since entering care she has had twelve different caseworkers. (D.E. 1 at 15). As in McReynolds, there is a “vicious cycle” here. Caseworkers with too little time, either by dint of being too new to the job or being overburdened, will find it difficult to build the necessary rapport, which, in turn, increases their effective workload and encourages staff turnover. Plaintiffs have established a policy or practice on the part of Defendants to assign or allow caseworkers to have these high workloads. What remains is to connect that policy or practice to the alleged harm, namely a violation of class members’ Fourteenth Amendment right to personal security and reasonably safe living conditions. The Court first notes that Plaintiffs’ argument on this point is intuitive: a higher caseload means that a caseworker can devote less 52 / 107 time and energy to each case, and naturally class members’ safety is affected by the amount of time and energy their caseworkers can devote to their respective cases. This is especially true since PMC children often represent complex or difficult cases. (PX 323 at PLTF0019499). As stated previously, Plaintiffs are allowed to rely on common sense inferences at the class certification stage. Kase v. Salomon Smith Barney, Inc., 218 F.R.D. 149, 152 (S.D. Tex. 2003) (citing Zeidman v. J. Ray McDermott & Co., Inc., 651 F.2d 1030, 1039 (5th Cir. 1981)). In addition, Plaintiffs have presented considerable evidence that high caseloads can threaten the safety of children in the PMC. DFPS documents state: “Caseworkers are responsible for making sure children are safe and their needs are being met.” (DX 13 at 4). Yet, a number of DFPS documents also explain that high caseloads make it difficult for caseworkers to succeed at this task: “High caseloads cause quality casework to suffer, thus putting children in our system further at risk of harm,” (PX 319 at PLTF0007109), and “The higher the caseload, the harder it is to make the required monthly face-to-face contact with children in foster care and their parents. National data shows that regular, meaningful contact is directly related to … increase[d] child safety,” (PX 13 at 652). DFPS has repeatedly acknowledged the importance of regular visits from a caseworker: “Caseworker visits are a critical element in maintaining the safety and well-being of children in foster care.” (PX 382 at VM0008077); see also (PX 324 at PLTF0027552 (“Regular visits allow the caseworker to ensure the safety and well-being of the child …”); PX 234). In spite of the importance of these visits, in 2010 almost 20% of children in the State’s care were not receiving monthly visits from their caseworkers, and that value represents a substantial improvement from preceding years. (PX 15 at 314). Texas has been improving in this regard; in fiscal year 2011 90% of foster care children were visited by their 53 / 107 caseworkers on a monthly basis. (DX 28 at DFPS005458049); see also (PX 235 (listing average monthly face-to-face contacts with child in care completed in fiscal year 2011 at 95.2%)). None of the foregoing depends on the Court endorsing or adopting a particular caseload number, such as those recommended by COA or CWLA. The evidence that DFPS caseworkers are overworked comes from expert testimony, the agency’s own documents, and inferences from facts in the record. It further bears noting that there is absolutely no evidence that DFPS intentionally maintains high caseloads in order to undermine the safety or well-being of children in the PMC. The instant case is thus similar to both McReynolds—where seemingly benign policies were alleged to cause the harm—and Young—where the defendants’ failure to act constituted a policy or practice sufficient to establish commonality. Plaintiffs also allege that overburdened caseworkers cause class members to experience repeated placement moves. E.g., (D.E. 195 at 15-16). Children in the PMC frequently change placements. Six of the named plaintiffs have experienced more than 10 different placements in the Texas foster care system. (D.E. 160 at 4). Named plaintiff M.D. had three foster placements in just a six month period. (D.E. 1 at 5-6). D.I. was in one placement for only 10 days, id. at 9, while S.A. has had a staggering 21 placements since entering the PMC, id. at 13-14. While in the custody of the state, D.P. and T.C., who were once named plaintiffs in this case but have since been replaced, were moved 28 (including 16 times in a three and a half year period) and 20 times, respectively. Id. at 23, 25, 27-28. These named plaintiffs are indicative of a broad trend. Over 40% of children who have been in the State’s care for between 12 and 24 months have had more than two placements. That is, they are averaging more than one placement per year. For children in care for more than 24 months, nearly 80% have had more than two placements. Class Cert. Hr’g 1/23/13 (Gonzalez) at 54 / 107 71-72; (PX 15 at 318; PX 48 at 12). Children in the PMC average 4.5 placements, with more than one placement move per year, and the average number of placements increases the longer a child is in the State’s care. (PX 250); see also Class Cert. Hr’g 1/23/13 (Liebgott) at 99-100. Plaintiffs’ causal theory seems to be that a caseworker with sufficient time and resources can do a better job finding a placement suited to the child. The result would be an increased likelihood that any given placement is permanent, or at least longer-term and stable, because it fits the needs of the child. Defendants’ practices and policies regarding caseworker workload, then, contribute to the considerable number of placements that each class member experiences. This sort of shuttling of foster care children through a number of placements can cause psychological harm. Class Cert. Hr’g 1/23/13 (Gonzalez) at 65-66; (PX 12 at 17 (Administrator’s statement) (listing foster care moves as a factor that lead to poorer outcomes for the children)). For this reason, frequent placement moves have sometimes been recognized as giving rise to a Fourteenth Amendment claim. A Seventh Circuit panel was divided over this issue in K.H. The majority held that a right to be free from a “general practice of shuttling children among foster parents” was not clearly-established for the purposes of qualified immunity, which left open the possibility that such a right could be established in later litigation, especially in an equity action such as this one. K.H. through Murphy v. Morgan, 914 F.2d 846, 853 (7th Cir. 1990). Writing for the majority, Judge Posner opined that frequent placement moves are not a harm in and of themselves—such moves may be necessary to safeguard the child’s safety—but they indicate systematic failings in foster care. Id. Such failings can inflict harm on class members, so on that understanding placement moves could form the basis of a Fourteenth Amendment claim—they would be a symptom that provides evidence that there are failings and Fourteenth Amendment violations in the foster care system. Concurring in part, 55 / 107 Judge Coffey wrote: “I am of the opinion that the allegations regarding K.H.’s ‘9 placements in a period of approximately three and a half years,’ together with the serious traumatic experiences (physical and sexual assaults) K.H. suffered during these myriad placements might conceivably establish a violation of the due process clause as to care and judgment as Youngberg recognized.” Id. at 865 (Coffey, J., concurring in part and dissenting in part). More recently in a class action on behalf of Massachusetts foster children, the Connor B. court listed “shuttling [class members] among foster families without any hope of finding a permanent home” among the alleged systemic failures of that state’s child welfare system, holding that those deficiencies together were sufficient to give rise to a Fourteenth Amendment claim. Connor B. ex rel. Vigurs v. Patrick, 771 F.Supp. 2d 142, 163 (D. Mass. 2011). To prevail on the merits with this claim, Plaintiffs will have to prove a causal connection between the State’s caseworker workload and an unconstitutional risk of harm. At class certification, though, they do not need to prove that they are entitled to relief based on their claims; they need only make out a claim on behalf of the class. They have done so here. The Court finds the relationship between caseworkers’ workloads and class members’ safety persuasive. Caseworkers are, in effect, these children’s fire alarms. They are the first and best mechanism to ensure the class members’ safety, i.e., to ensure that the child is not suffering any harm or abuses. A caseworker that is so overburdened that she cannot visit the children she is responsible for to keep apprised of their well-being, or if she cannot build enough of a rapport to do so effectively, cannot fulfill this function. The same is true if the caseworker cannot identify suitable, stable placements for the children in her care, necessitating frequent placement moves. In such circumstances, a caseworker is analogous to a faulty fire alarm or other safety precaution, and courts have considered claims based on shoddy or inadequate safety measures on 56 / 107 behalf of those in the State’s custody. See Section V., supra. The Court’s reasoning here parallels D.G. ex rel. Strickland v. Yarbrough, 278 F.R.D. 635 (N.D. Okla. 2011), which in a similar case concluded that excessive caseloads for caseworkers, as well as high turnover rates in child welfare agencies, could pose a threat to foster care children. Citing ample evidence, the D.G. court held that “manageable caseloads for child welfare workers are important to ensure that children are kept safe from harm.” Id. at 641. The LaShawn A. court reached a similar conclusion, holding: “To the extent that certain services, such as appropriate placements and case planning, are essential to preventing harm to the children in the District’s custody, this Court holds that the children have a constitutional liberty interest in those services.” LaShawn A. v. Dixon, 762 F. Supp. 959, 993 (D.D.C. 1991) aff’d and remanded sub nom. LaShawn A. by Moore v. Kelly, 990 F.2d 1319 (D.C. Cir. 1993); see also Wendy H. By and Through Smith v. City of Philadelphia, 849 F.Supp. 367 (E.D. Pa. 1994) (citing LaShawn A. as persuasive). To what extent caseworkers are overworked, whether this overwork is significant enough to subject the members of the General Class to an unconstitutionally unreasonable risk of harm, and whether the State has sufficient mechanisms in place to mitigate those risks8 are the issues central to the Plaintiffs’ claim. Resolving them will determine the validity of the common General Class Fourteenth Amendment claim in “one stroke.” If, for example, it is proven at trial that caseworkers are not so overworked that they fail to provide constitutionally adequate care, then that will resolve all of the class members’ claims at once; the State’s policies regarding caseworker workloads would be found not to violate the Fourteenth Amendment.9 8 DFPS may have informal means to help alleviate high caseloads, such as relying on local supervisors. Class Cert. Hr’g 1/22/13 (McCall) at 344, 342, 336. 9 Of course, some members of the class may still be able to bring individual claims on the grounds that their caseworkers acted so as to subject them to an unreasonable risk of harm or unsafe living conditions. The actions of those caseworkers rather than the State’s policies would be at issue in those cases. Although both cases would be based on their respective plaintiffs’ Fourteenth Amendment rights, they would be different claims. 57 / 107 Defendants contend that Plaintiffs’ Fourteenth Amendment claim necessitates an individualized inquiry into the circumstances of each class member, defeating commonality. (D.E. 163 at 18-20; D.E. 198 at 13). This claim is not persuasive. Under either the deliberate indifference or professional judgment standard, whether an identified policy violates a class’ due process rights does not entail individualized inquiries. To prevail under deliberate indifference, the Plaintiffs must show that the State officials disregarded a known or obvious risk to the class members’ safety. This showing would hinge on the information available to the State as well as the effects of the policies and practices it opted to implement (i.e., did they actually pose a risk?). It does not hinge on the individual circumstances and characteristics of class members. Similarly, the absence of professional judgment standard holds State officials liable only when the policy or practice that causes harm has no basis in professional judgment. These culpability standards turn on the Defendants’ mental states and justifications for their decisions, respectively, not on the circumstances or characteristics of class members. Defendants’ argument also mistakes the legal harm that is the basis of the General Class claim. Plaintiffs allege that these policies lead to an unreasonable risk of harm. As explained above, that risk alone is the legal injury. Individuals in the State’s custody have a constitutional right to be free from an undue risk of harm, or, the same thing stated another way, a right to reasonably safe living conditions. Whether some members of the putative class have actually suffered physical or psychological harm provides some evidence that this risk exists, but the actual harms suffered by some class members is not constitutive of the legal injury at the heart of Plaintiffs’ claim. As in the prison fire safety cases cited above, a plaintiff does not need to wait until actually harmed, until the risk of harm is realized, in order to bring a claim for inadequate safety precautions while in the State’s care. The fact of whether these policies subject class 58 / 107 members to an unreasonable risk of harm, and whether that risk is so unreasonable as to rise to a constitutional violation, can be proven on the basis of classwide evidence without individualized inquiries. The risk of harm is general; a particular realization of that risk may entail an individualized inquiry. After conducting a rigorous analysis of Plaintiffs’ General Class claim the Court finds that Plaintiffs have satisfied Rule 23(a)’s commonality requirement. Plaintiffs have alleged a deprivation of a legally-protected right (class members’ Fourteenth Amendment rights), have identified a common policy or practice on the part of Defendants (those relating to caseworker workloads), have presented evidence that shows that this policy or practice may impair that right, and have identified common questions of law or fact that will resolve their claim. C. Typicality The State’s caseworker policies and practices affect all children in the Texas PMC, including the named plaintiffs.10 Their claims are therefore typical of the class. Defendants press a version of the argument above, maintaining that Plaintiffs have failed to show typicality because they have not provided detailed proof regarding the named plaintiffs’ individual claims such that the Court could compare those claims with those of the class as a whole. (D.E. 198 at 22-23). Rule 23(a)(3) does not, however, require named plaintiffs to have individual claims identical to the class claims—i.e., they need not have suffered identical harms or even have harms that were caused the exact same way. It is sufficient that their claims arise from the same or a similar course of conduct and share the same legal theory. In Daffin v. Ford Motor Co., 458 F.3d 549 (6th Cir. 2006), the plaintiff alleged that a design defect caused the accelerator to stick. 10 D.I., S.R., M.R., and J.R. are not included in this discussion, or in the adequacy of representation section below, due to their changed circumstances—they are discussed in Section XI., infra. 59 / 107 The court of appeals affirmed certification of a class consisting of all state residents who leased or owned certain models of a van while under warranty, regardless of whether the class members had experienced the specific accelerator sticking problem that the named plaintiff had. Daffin and the other class members’ claims arise from the same practice (delivery of non-conforming vehicle), the same defect (the allegedly defective throttle body assembly), and are based on the same legal theory (breach of express warranty). Typicality is satisfied despite the different factual circumstances regarding the manifestation of the accelerator sticking and Ford’s attempts to remedy manifested sticking. Daffin, 458 F.3d at 553; accord Butler v. Sears, Roebuck & Co., 702 F.3d 359, 362 (7th Cir. 2012) cert. granted, judgment vacated, 133 S. Ct. 2768 (2013) and judgment reinstated, 2013 WL 4478200 (7th Cir. Aug. 22, 2013); In re Whirlpool Corp. Front-Loading Washer Prods. Liab. Litig., 678 F.3d 409, 420 (6th Cir. 2012) cert. granted, judgment vacated sub nom. Whirlpool Corp. v. Glazer, 133 S. Ct. 1722 (2013) aff’d 2013 WL 3746205 (6th Cir. July 18, 2013). The instant case is the same. The named plaintiffs’ claims arise from the same practice (the State’s caseworker workload practices), the same “defect” (caseworkers cannot devote sufficient time to safeguard class members’ well-being), and are based on the same legal theory (violation of the State’s duties under the Fourteenth Amendment for those in its care). As above, the Defendants’ argument appears to mistake the nature of the class members’ alleged injury. Plaintiffs do not allege, at this point in the litigation, that policies or practices in the Texas foster care system cause a psychological harm that is common to all class members. Defendants are correct that such a claim would be difficult to sustain as a class action; variations in the particular circumstances and psychological states of each class member would abound. Plaintiffs’ claim, as the Court understands it, is much more general. They assert that these 60 / 107 policies and practices subject class members to an unreasonable risk of harm, a legal injury that is common to all class members. The named plaintiffs’ claims are thus typical. D. Adequacy of Representation Adequacy of representation requires that the class representatives fairly and adequately protect the interests of the putative class. Aspects of adequate representation also tend to merge into typicality and commonality. Those aspects of Rule 23(a)(4) have already been discussed at length above. Plaintiffs have submitted exhibits detailing the extensive experience of Plaintiffs’ counsel in complex and class action litigation. (D.E. 160-A-160-D). Defendants do not challenge the competence and zeal of Plaintiffs’ counsel of record. (D.E. 163 at 59). Accordingly, the Court finds that Plaintiffs’ counsel can adequately represent the class. Class representatives must satisfy the Court that they, and not their attorneys, are directing litigation. Unger v. Amedisys Inc., 401 F.3d 316, 321 (5th Cir. 2005). In their original Brief Opposing Class Certification, Defendants argued that the class representatives were not sufficiently informed to direct the litigation. (D.E. 163 at 59). They promised to “present evidence at the certification hearing that the class representatives have demonstrated woefully inadequate knowledge about the case and that their lack of knowledge and understanding has seriously impeded progress in the case.” Id. No such evidence was proffered. Instead, the Court heard testimony from Anna Ricker, next friend for J.S., H.V., and P.O.; Estella Vasquez, next friend for L.H. and C.H.; Bobbie Young, next friend for S.R., J.R., and M.R.; and Carla Morrison, next friend for Z.H. The Court approved their status as next friends if it had not already done so. They demonstrated knowledge of and engagement with the litigation. 61 / 107 Additionally, the Court had previously approved many of the next friends and nothing has changed in the posture of the case that would undermine those findings. Thus, the Court finds that the next friends adequately represent the interests of the General Class. These observations about adequacy of representation apply to the Licensed Foster Care Subclass, the Foster Group Home Subclass, and the Basic Care GRO Subclass as well. The Court finds that Rule 23(a)(4) has been satisfied with respect to those subclasses. E. Rule 23(b)(2) Plaintiffs request that the Court enter a permanent injunction “requiring Defendants to ensure that all children in the General Class are assigned DFPS caseworkers whose individual caseloads do not exceed caseload standards established by the Child Welfare League of America and/or the Council on Accreditation.” (D.E. 179 at 31). Further, the Court may consider equitable remedies other than those suggested by Plaintiffs as it is well-established that equitable relief is flexible and intended to be tailored to the circumstances. E.g., Lemon v. Kurtzman, 411 U.S. 192, 199-201 (1973); Hecht Co. v. Bowles, 321 U.S. 321, 329-30 (1944). The Court need not, at this stage, determine what remedy Plaintiffs would be entitled to if they prevailed on the merits of their claim. Rather, the Court must determine that the Plaintiffs’ claim is one that is susceptible to common, specific relief. Plaintiffs have alleged a unified harm to class members—an unreasonable risk of harm due to ineffective monitoring and planning because caseworkers are overburdened. The Court can conceive of a number of appropriate injunctions that would cure this injury, such as: setting maximum caseloads, hiring more caseworkers, or some overflow procedure that distributes cases so as to ensure that no caseworker is especially overburdened. See D.G. ex rel. Strickland v. Yarbrough, 278 F.R.D. 62 / 107 635, 636 (N.D. Okla. 2011) (holding that an injunction setting limits on the caseloads of caseworkers and their supervisors met the requirements of Rules 23(b)(2) and 65(d)). Critically, injunctions of this sort would require no additional individual assessments. In addition, Plaintiffs have suggested some standards by which to fix a caseworker’s workload—namely those articulated by CWLA and COA—as well as a system by which to weight cases according to the challenges they represent to DFPS staff. Class Cert. Hr’g 1/23/13 (Warren) at 113-15; cf. Shook II, 543 F.3d 596, 605-06 (10th Cir. 2008) (“The problem in our case is that plaintiffs have eschewed any effort to give content to what it would mean to provide adequate mental health staff, adequate screening …”). While these standards do not define Defendants’ duties under the Fourteenth Amendment, they may be useful in constructing an appropriate remedy. The Court finds that the General Class satisfies Rule 23(b)(2). VIII. Licensed Foster Care Subclass On behalf of the LFC Subclass, Plaintiffs present two Fourteenth Amendment Claims. First, they argue that licensed foster care placements are not subject to sufficient monitoring and oversight. In short, the RCCL division of DFPS,11 charged with the task of regulating and licensing all CPAs and child care placements in the State, is not performing adequately to secure the LFC Subclass’ constitutional rights. Poor oversight of these placements, they allege, subjects members of this subclass to an unreasonable risk of harm and unsafe living conditions. (D.E. 179 at 14). Second, Plaintiffs argue that the State maintains “an inadequate placement array,” meaning that the only appropriate placements for a member of the LFC Subclass will frequently involve relocating from their home community, sometimes great distances, and that appropriate placements may not even be available. (D.E. 179 at 11). 11 RCCL is part of the Child Care Licensing division of DFPS. 63 / 107 Plaintiffs also allege that this inadequate placement array violates the subclass’ right to family integrity, derived primarily from the First Amendment. (D.E. 179 at 20). However, they cite no legal authority in support of this claim. See (D.E. 160 at 40; D.E. 195 at 44). Moore v. City of East Cleveland, 431 U.S. 494, 498-99 (1977), extends constitutional recognition and protection to the family generally. Aristotle P. v. Johnson, 721 F.Supp. 1002 (N.D. Ill. 1989), seems especially relevant to this claim: plaintiffs were siblings in foster homes who alleged that the State’s policies separated them and made them unable to visit or communicate with each other. The Aristotle P. court held that the plaintiffs’ relationships with each other are “the sort of ‘intimate human relationships’ that are afforded ‘a substantial measure of sanctuary from unjustified interference by the State,’” id. at 1005-06 (quoting Roberts v. U.S. Jaycees, 468 U.S. 609, 618 (1984)), and that “[t]he plaintiffs’ relationships with their siblings are even more important because their relationships with their biological parents are often tenuous or nonexistent.” Id. at 1006. Accordingly, the court held that plaintiffs had stated a claim based on their First and Fourteenth Amendment rights. There are some important differences between Aristotle P and the instant case, though. Notably, Aristotle P did not involve a class action. Also, the plaintiffs there did not dispute the State’s placement decisions, id. at 1006, challenging instead policies that did not facilitate sibling visitation, id. at 1007-08. Here Plaintiffs have not provided sufficient legal basis for the Court to conduct a rigorous analysis of this claim and determine to what extent it—and the potentially common questions of fact and law it entails—differs from Plaintiffs’ Fourteenth Amendment claim based on an unreasonable risk of harm and arising from Defendants’ same conduct. So, the Court restricts attention to the latter claim. 64 / 107 A. Commonality Plaintiffs identify the State’s monitoring, licensing, and oversight of licensed foster care placements as the relevant policies or practices in their first Fourteenth Amendment claim. As with the General Class, staffing and operational decisions, such as allowing for follow-up inspections via phone or e-mail, Class Cert. Hr’g 1/23/13 (Chansuthus) at 287, are made by a central authority. State officials and agents make deliberate choices about how to approach these tasks and how to structure and staff the RCCL. The same is true for Plaintiffs’ second Fourteenth Amendment claim—the array of placements within the State is a function of State policy as to how operations are licensed and contracted for. Even if the distribution and array as they are now are not what the policy makers intended, it is a result of the way the State manages the system and is something that the State is clearly aware of and can control since the licensing and operation of facilities is heavily regulated. See (PX49; PX 50). Further, as explained above, failure to act can constitute a policy or practice that can be properly litigated as a class action. See Lightfoot v. District of Columbia, 273 F.R.D. 314, 321 (D.D.C. 2011); Young v. Nationwide Mut. Ins. Co., 693 F.3d 532, 543 (6th Cir. 2012). State actors are aware of the situations regarding RCCL and the placement array as a substantial portion of the evidence Plaintiffs have presented to the Court comes from State records. See, e.g., (PX 295 (2011 Annual Assessment by the Performance Management Unit of DFPS’s Child Care Licensing Division); PX 296 (2012 Annual Assessment by the Performance Management Unit); PX 148 (internal DFPS document expressing concern about geographic distribution of services and providers); PX 255 (DFPS document listing children placed outside their region)). State officials also gave depositions in which they acknowledged the problems with Texas’ oversight and array of placements. In short, both the oversight and array of foster care 65 / 107 placements in the State—whether constitutionally adequate or inadequate—constitute policies or practices that can potentially form the basis of a class action. As with the General Class, Plaintiffs must connect the policy or practice to the alleged harm, though they do not have to prove that class members were, in fact, harmed at the class certification stage. In order to conduct a rigorous analysis of the first LFC Subclass claim, the Court must consider whether there is evidence that Defendants do not engage in sufficient oversight of foster care placements and whether that policy or practice might impair class members’ right to reasonably safe living conditions. In other words, the Court must analyze the evidence to determine not just whether the State has some policy or practice of oversight (which it obviously does), but whether it has a policy or practice of inadequate oversight. The same analysis must then be done for the array of foster care placements in Texas. The Court considers each claim in turn. RCCL’s responsibilities are extensive. The department is charged with regulating the provision of services by licensing, inspecting, and investigating residential child-care operations. (DX 3). The Child Care Licensing Department, of which RCCL is a part, has four general responsibilities. First it must regulate “all child-care operations and child-placing agencies to protect the health, safety, and well-being of children in care.” Second it must “[e]stablish and monitor operations and agencies for compliance with licensing standards, rules, and law.” Third it keeps parents and the public informed about child care and the histories of various facilities and their compliance with minimum standards of care. Fourth it provides technical assistance to providers to help them meet licensing standards, rules, and laws. (PX 434 at 8). The licensing standards RCCL must enforce are voluminous and relate to nearly all aspects of the operation of a facility and the provision of care. See, e.g., (PX 49 (Minimum 66 / 107 Standards for GROs, in excess of 300 pages); PX 50 (Minimum Standards for Child-Placing Agencies, the same)). Licensing and inspection are extensive processes, requiring a great deal of time in order to complete correctly and thoroughly. In Texas, the provision of care is done mostly by private CPAs that contract with the State, and RCCL is the State’s mechanism for ensuring that these providers meet minimum standards, keep children safe, and provide adequate care. This task is daunting. In 2011 there were 10,744 24-hour care facilities in Texas that RCCL was responsible for, and there were 11,577 inspections and 4,824 investigations, 2,065 involving cases of abuse or neglect. (PX 325); see also (PX 70; PX 244). Between September 2010 and August 2011 RCCL had to complete an average of 1098.3 inspections per month and process an average of 84.8 applications for licenses per month. (PX 226). Despite these extensive and important responsibilities RCCL had only 124.4 full-time equivalent investigative caseworkers and 177 total staff in fiscal year 2011. (PX 325 at PLTF0062239). Recently the number of investigators was cut by a small amount due to budget constraints, though there was no corresponding reduction in RCCL’s duties. Class Cert. Hr’g 1/23/13 (Bateman) at 244-45. And, RCCL’s duties have only increased in the past few years. The number of placements the department is responsible for has generally increased, although in 2012 there was a decrease in both the foster care population and the number of licensed foster care facilities. (PX 21 at 7). Also, in 2011 RCCL changed its policies so that it inspects RTCs more frequently, resulting in another increase in its workload without a corresponding increase in personnel. (PX 206 at DFPS001283841 (increasing inspections of RTCs to once every six months rather than annually and requiring nighttime or weekend inspections of RTCs)). The effect of these conditions is that RCCL staff, like caseworkers, carry high workloads. Caseloads in RCCL have been rising: from 7.9 in 2009 to 9.5 in 2012, a 20% increase in just 67 / 107 three years. (PX 21 at 8). Similarly, Charlane Bateman, Director of Policy and Program Operations in the Child Care Licensing Division of DFPS, stated that RCCL staff carried high workloads. Class Cert. Hr’g 1/23/13 (Bateman) at 251. In her review of RCCL operations, Ms. Chansuthus concluded that RCCL lacked adequate staff “to perform their regulatory work, essentially to comply with their own policies to handle what seemed to me to be a heavy and complex workload to do quality work and to enforce provider compliance with minimum standards.” Id. (Chansuthus) at 282-83; see also (PX 21 at 9 (“it is my opinion that RCCL staffing does not adequately accommodate the scope of the critical regulatory work to be completed”)). She also found that in the two regions with the highest concentration of placements, Region 3 and Region 6, RCCL staff could not complete their tasks within a standard number of workdays in a month. Workers in Region 6 would need 26 days in a month, or six more than the number of workdays in a standard month, to complete their investigations and inspections while workers in Region 3 would need an impossible 34 days per month. Class Cert. Hr’g 1/23/13 (Chansuthus) at 281-82; (PX 21 at 8). These calculations assume an eight hour workday and that tasks are distributed equally amongst all investigating and inspecting staff. (PX 21 at 8). As Bateman explained, a higher workload means a less effective staff generally. More cases either means fewer inspections, less probing inspections, or both. The ability of RCCL to enforce compliance decreases and problems at facilities persist longer. Class Cert. Hr’g 1/23/13 (Bateman) at 244-50. Chansuthus’ research corroborates that assessment and provides evidence that deficiencies of this sort presently exist. The Performance Management Unit (“PMU”) of DFPS ensures that the agency is performing according to policies and procedures. See Class Cert. Hr’g 1/23/13 (Chansuthus) at 283 (describing the PMU as analogous to a continuous 68 / 107 quality assurance unit that monitors performance and makes recommendations). In its January 2011 Data and Trend Assessment (“2011 Assessment”) the PMU found that RCCL failed to perform follow-up inspections to check whether the placement was complying with standards in a timely fashion 35.7% of the time. PMU’s February 2012 Data and Trend Assessment (“2012 Assessment”) reported that follow-up inspections were not timely performed in 36.4% of the cases reviewed. (PX 21 at 9-10). Even when follow-up inspections have been conducted, they are sometimes done in a manner that is ineffective, such as via e-mail, fax, or telephone. The PMU has recommended that only certain placement deficiencies, such as those involving paperwork, be verified in this manner, but the 2011 Assessment reported that in 41.5% of the placements reviewed the “‘[t]ype of follow-up chosen was not the best method to reduce risk (e.g. not making an in-person inspection in the case of health and safety related issues).’” Id. at 10 (quoting the 2011 Assessment). There has been significant improvement in this area; the 2012 PMU report found that inappropriate follow-up was used in only 22% of the placements reviewed, though that is still a substantial failure rate. Id. at 10. Inappropriate follow-up methods, delays in performing follow-ups, or failing to perform them at all are connected to workload. As Ms. Chansuthus—accepted as an expert in evaluating child welfare systems—noted, staff with heavy workloads will try to take what shortcuts they can to complete their work, which can include substituting the less time-consuming phone follow-up for an on-site investigation. Class Cert. Hr’g 1/23/13 (Chansuthus) at 287-88. Emails between DFPS staff support this observation: if RCCL workloads increase then “follow ups at RTCs may be seriously impacted. These are very time consuming …” (PX 21 at 11 n.75). Looking to the effects of RCCL’s efforts, its oversight does not seem effective at curtailing deficiencies in placements, leading to a pattern of repeat deficiencies. The 2011 69 / 107 Assessment found that 65.6% of the placements surveyed had repeat deficiencies, a number that rose to 77.6% in the 2012 Assessment. Id. at 11-12. There is additional evidence of widespread non-compliance with minimum standards. The latest data before the Court states that there was an average of 2.3 deficiencies per inspection and only 6.2% of facilities had no deficiencies in the last two years. (PX 235 at 3). The foregoing all supports Chansuthus’ conclusion that “follow-up inspections conducted by RCCL were untimely and ineffective.” Class Cert. Hr’g 1/23/13 (Chansuthus) at 284. Chansuthus’ expert report also describes three other areas where RCCL appears deficient. One of RCCL’s jobs is to ensure that the proper background checks are performed. Texas minimum standards require background checks for anyone who works, volunteers, or lives at a placement. E.g., (PX 21 at 12-13). In addition to ordinary monitoring, RCCL is supposed to evaluate a placement’s compliance with these rules as part of an investigation. Id. at 13. However, PMU reviews indicate that from March 2011 to September 2011, in cases designated “Priority 1,” meaning they involved allegations of physical or sexual abuse or the death of a child, the investigator’s documentation did not indicate that background checks were evaluated in an average of 52.8% of the cases. Id. at 13 n.92. The percentage of evaluations that were either not performed or not documented ranged from 24% in September 2011 to 67% of them that March. Id. Over the same period of time, a safety plan was documented for less than half of the Priority 1 cases where one was called for. Id. at 18 (the average between March 2011 and September 2011 was 42%). RCCL also monitors the use of emergency behavior intervention (“EBI”), which includes personal restraints, seclusion, and emergency medication. (PX 183 at DFPS000968400). EBI is sometimes necessary to prevent a child in care from harming herself or others, and the standards 70 / 107 concerning its use are detailed. Id. at DFPS000968401-405. There are concerns, however, that EBI is over- and inappropriately used. As one DFPS report put it: Although minimum standards prohibit these practices, EBI is often used to control children’s behavior and/or as a result of inadequate treatment programs. Particularly in facilities serving the children with the most severe emotional and behavioral challenges, the current trend is to overuse EBI in lieu of more progressive interventions. Id. at DFPS000968406. Placements are required to report their EBI use quarterly, but many fail to do so or do so only inconsistently: between 2008 and 2010 an average of 32% of placements did not properly report their EBI use. (PX 21 at 16 n.109). Similarly, in a review of EBI evaluations that surveyed 79 RTCs, RCCL determined that “the vast majority of RTCs either are not conducting annual evaluations or are doing so only in partial compliance with minimum standards requirements.” (PX 208 at DFPS001380734); see also Class Cert. Hr’g 1/23/13 (Chansuthus) at 293-94. Finally, Chansuthus also cites a number of worrying statistics from a 2011 PMU review of RCCL abuse and neglect investigations in Region 3, where RCCL workers are particularly overworked. An assessment of the immediate safety of the children, a required part of any abuse or neglect investigation, was conducted 83% of the time. (PX 21 at 18). “‘[a]lthough this number could be considered high, a 17% rate of failure should be improved upon.’” Id. (quoting Residential Child Care Abuse/Neglect Investigations Region 3). Chansuthus cites the PMU’s comments, relating a number of anecdotes about foster children who had suffered, inter alia, repeated injuries from EBI or unexplained bruises and where there was no documentation indicating that an assessment relating to the child’s immediate safety was conducted. Id. at 1819. Along similar lines, the PMU found that risk was “thoroughly assessed” only 53% of the time, and a safety plan was actually implemented in only 64% of cases where one was called for. 71 / 107 Id. at 20. Finally, 20% of the cases surveyed by the PMU were inappropriately closed by supervisors. Id. In these cases, the investigator’s determinations were insufficiently explained, not all parties were interviewed, risk was not thoroughly assessed, the determination was not consistent with policy definitions, or law enforcement personnel were not contacted when they should have been. Id. It bears reiterating that Region 3 contains among the most foster care placements and RCCL workers in that region appear to be especially taxed. So, these statistics and anecdotes represent something of a worst case scenario for Texas foster care. Combined with all the other evidence presented to the Court, there is sufficient reason to believe that there are substantial deficiencies in RCCL’s oversight of licensed foster care placements. The connection between monitoring and oversight of placements and risks to class members or unsafe living conditions is a natural one—the point of RCCL’s oversight is to protect members of the Licensed Foster Care Subclass from harm and to ensure that they live in a safe environment. Plaintiffs have provided evidence that the State’s policies and practices lead to fewer and less thorough inspections and reduce RCCL’s ability to enforce compliance with minimum standards. An effect of these policies and practices is that more violations of State standards will go unnoticed and uncorrected and will be able to persist longer, subjecting foster children in these placements to an increased risk of harm. Of the placements with repeat standards violations identified in the PMU’s 2011 Assessment, 47% of the deficiencies were classified as High or Medium-High, and in the 2012 Assessment 65.5% of the deficiencies were so classified. (PX 295 at DFPS004490505; PX 296 at DFPS004490540). Deficiencies are classified according to the seriousness of the violation, and “higher weighted standard[s] have greater risk attached to them.” (PX 21 at 11 n.78 (quoting deposition of Charlane Bateman)). High level deficiencies include failure to supervise and failure to provide a safe environment, 72 / 107 while medium-high deficiencies include not being aware of a child’s special needs or not taking into account surrounding circumstances, hazards, and risks in deciding how closely to supervise a child. See, e.g., (PX 50 at 12; PX 49 at 49). Background checks are likewise designed to protect children by screening out those that should not be caring for them, and the 2011 Assessment listed operations with repeated background check deficiencies as one of the four riskiest elements for CPAs. (PX 295 at DFPS004490497). The failure to implement safety plans when they are called for clearly threatens class members’ right to personal security and to a reasonably safe living environment. EBI, while sometimes necessary, is also inherently dangerous. “All forms of EBI present certain risks to the child in care … Even when properly applied, restraint can result in death.” (PX 183 at DFPS000968402); see also (PX 316 at HHSC000000855 (child died due to improperly initiated restraint); PX 317 at HHSC00001015 (child died due to improper restraint)). The 2012 Assessment found that 60% of all serious injuries at GROs and RTCs surveyed were related to behavioral intervention. (PX 296 at DFPS004490546). Defendants counter that Chansuthus’ analysis is based only on documentation, meaning that her work cannot distinguish between something actually not happening and something happening and not being documented.12 (D.E. 198 at 39); Class Cert. Hr’g 1/23/13 (Chansuthus) at 299-300. So, it might be the case that the safety plan was, for instance, implemented; it was just not documented correctly. By her account, Chansuthus would need access to the original files or be able to speak to RCCL personnel to be able to make this distinction. Class Cert. Hr’g 1/23/13 (Chansuthus) at 299-300. The same observation may apply to PMU reports and assessments, which are based on surveys and “readings” of operations. This might be especially 12 Defendants also wrongly assert that the expert report produced by Ms. Chansuthus was not admitted into evidence. See Section VI., supra. 73 / 107 plausible given how overworked RCCL staff appear to be: given limited time and resources, they might choose to eschew rigorous documentation in favor of actually working to safeguard class members’ safety and well-being. As Chansuthus explained, however: “effective child welfare and licensing practice demands that work completed be documented timely, because, as any experienced child welfare worker knows, ‘if you don’t document it, it didn’t happen.’” (PX 21 at 14); Class Cert. Hr’g 1/23/13 (Chansuthus) at 292 (“sort of a general axiom in child welfare practice is ‘if it’s not documented, then it’s not done.’ So, you can’t make assumtions when you are in a unit like PMU.”). Although Defendants’ point is plausible, they provide no evidence to support it. There is no evidence before the Court, such as statements made by RCCL personnel, suggesting that the deficiencies outlined above are really the product of inadequate documentation and that the various reports do not reflect the actual conditions of children in foster care. Moreover, it would be reasonable to expect that if this were the case, the PMU would flag it. The PMU seems careful about describing the methodology it uses, including its weaknesses. In the Master Casereading Survey it undertook as part of the 2012 Assessment, the PMU explains that the 389 casereads it analyzed “cannot be regarded as representative of the entire State or region as a whole. The statistics we receive are clearly of those operations that have risk, not a random sample . . . Simply put, we are studying the elements of what separates high risk from mediumhigh risk.” (PX 296 at DFPS004490533). If there was a widespread, general reason to doubt the PMU’s findings (and PMU reports were a major source that Chansuthus relied on), it is odd that the agency never mentions it. At most, Defendants have presented another way to consider one of the central questions of fact in this case—is RCCL failing to exercise effective oversight over licensed foster care placements?—and illustrated how it is still a live issue, one that should be 74 / 107 resolved at trial. Plaintiffs have shown how the policy or practice they identify is causallyrelated to the alleged harm, and all Defendants have offered against the weight of the evidence in the record is a possible reason to treat it with some skepticism. The Court concludes that Plaintiffs have established commonality with regards to this claim. Common questions for this claim include the following. Do DFPS’s policies regarding RCCL workload and regulation cause a greater likelihood of deficiencies by placement providers? Do these increased deficiencies constitute an unconstitutionally unreasonable risk of harm? Are such policies deliberately indifferent to the safety of class members; is the choice or implementation of the policies not based on any professional judgment? Resolving these questions will be dispositive of Plaintiffs’ first Fourteenth Amendment claim on behalf of the LFC Subclass. Plaintiffs’ second Fourteenth Amendment claim on behalf of this subclass is that DFPS maintains an insufficient array of placements. The array and distribution of placements constitutes a policy or practice of the State and flows from acts and policies of DFPS. Indeed, DFPS acknowledges that deficiencies in the array of placements are the result of the structure of the current foster care system and the way it procures services. (PX 150 at DFPS000630776). In DFPS’s plan for Foster Care Redesign it was noted that: “The geographic distribution of foster care services in Texas is inconsistent with the areas where services are needed. Children must move where services are offered rather than the services being offered where the children are located.” (PX 149 at DFPS000630358). The Adoption Review Committee arrived at a similar conclusion, reporting that children are too often moved outside of their home communities and this results in losing their educational continuity and important personal connections. (PX 323 at PLTF0019514). 75 / 107 The location of placements throughout Texas means that class members must be moved outside of their home communities and away from their siblings and support networks. Id.; see also (PX 322 at 11, 19). A 2010 report to the Texas legislature DFPS made similar statements: “Finding appropriate placements for children who come into foster care remains difficult because the needs of the children do not always match the number, type, and location of the placement options available.” (PX 324 at PLTF0027560). Gail Gonzalez also testified in depositions that 50% of the RTC beds were in the Houston area rather than dispersed throughout the State, indicating that many children with the acute needs that require this sort of specialized care would have to be relocated there. Gonzalez also stated that there are not enough providers throughout the State to supply a variety of services, especially without moving the children far from their home communities. Class Cert. Hr’g 1/23/13 (Gonzalez) at 69-70, 76. The shortcomings of Texas’ array of foster care placements tend to increase the number of placements children experience. As the Foster Care Redesign plan explains, even if children can be placed in their home communities, “there are few local service continuums that can meet the changing levels of need for children.” (PX 149 at DFPS000630358); see also (PX 148 at 2 (Foster Care Redesign powerpoint presentation)). So, as a child’s needs change, for instance, if the treatment at an RTC improves her condition, the placement must as well, which can involve yet another long-distance relocation. Gonzalez described the array of placements as the obstacle preventing DFPS from meeting its goal regarding number of placements. DFPS’s target was to keep 42% of the children who had been in its care for two or more years to only two placements, but only about half as many of those children (20.9%) had that few placements. Class Cert. Hr’g 1/23/13 (Gonzalez) at 72-73. As discussed previously, frequent placement moves can represent 76 / 107 a violation of class members’ Fourteenth Amendment rights in and of themselves. See Section VII.B.. As of February 29, 2012 almost one quarter of the children in the PMC were placed outside their home region and well over half were placed outside their home county. (PX 255); see also (DX 46 at DFPS005158136). The numbers vary widely: only 15.8% of PMC children from Region 6, which includes Houston, were placed outside their home region, whereas 54.6% of children from Region 9 were. Id.; see also (PX 66 (mapping the distribution of placements for children from Lubbock County according to level of care and type of placement)). An inadequate distribution and array can also cause separation of sibling groups. Among children in the PMC on March 31, 2011 only 53.6% of sibling groups were all placed together and 26% of sibling groups were entirely separated (i.e., no member of the sibling group was in a placement with a sibling). (PX 2 at APPX2078); see also (PX 191 at 1). There has been some improvement; in June, 2012 16.9% of sibling groups were entirely separated and 64.2% had all of the siblings placed together. (PX 311). The sibling relationship is important to children, and even more so for many in foster care. Since their needs were not being met by their parents, siblings often come to depend on each other. Separation of siblings in care adds to their emotional burdens and can cause “[t]rauma, anger and an extreme sense of loss.” And, “separating siblings may make it difficult for them to begin a healing process, form attachments, and develop a healthy self-image.” (DX 33 at 35-37); see also Aristotle P., 721 F.Supp. at 1006. Some of the named plaintiffs demonstrate these general problems. A.M. and her sister were placed in a home with the goal of eventual adoption in 2006. But, A.M. required more specialized care and was moved to an RTC. Since there were no appropriate placements in the area, she was relocated a considerable distance. Her sister was then adopted, but A.M. is only 77 / 107 able to see her rarely and has few contacts with her family. (PX 25 (filed under seal)). Z.H. had a similar experience when he was placed in a distant RTC. After this move he was no longer with or near his siblings or his mother. (PX 28 (filed under seal)). K.E. has been moved multiple times due to her placement needs and she has rarely been placed near her home. (PX 31 (filed under seal)). S.A. has resided in a series of placements in at least four of DFPS’s eleven regions. (D.E. 1 at 14). And, M.D. was placed in a foster home over 400 miles from her home community. Id. at 6. Plaintiffs allege that the inadequate placement distribution and array leads to two options, both of which put class members at an unreasonable risk of harm. The child can be placed in an inappropriate setting given her needs, such as a GRO when she requires a foster home,13 or viceversa. Or, the child can be placed in an appropriate setting, but one far away from her home community and family members. Both of these options can cause harm to class members. Intuitively, placing a child in an inappropriate setting is detrimental to her, and can carry significant risks, Class Cert. Hr’g 1/22/13 (Miller) at 241, and can increase the number of placement moves a child experiences. Placing a child far from her home community not only leads to the problems described above, it also exacerbates the harms described in relation to the General Class because it makes it more difficult for caseworkers and others responsible for the child’s well-being to visit and monitor her. (PX 150 at DFPS000630780). Defendants criticize the LFC Subclass claims as being based on the false premise that class members have a Fourteenth Amendment right to be free from, inter alia, inappropriate placements and frequent placement moves. (D.E. 163 at 62-63; D.E. 198 at 38 (re-urging previous arguments)). Class members do not have constitutional rights, per se, to these things. They do, however, have a right to be free form an unreasonable risk of harm. To the extent that 13 A variation of this concern is the gravamen of Plaintiffs’ claim on behalf of the Basic Care GRO Subclass. 78 / 107 any of the policies and practices Plaintiffs identify in these claims impair that right, Plaintiffs have made out a claim that they should be enjoined. Plaintiffs have presented evidence that Defendants do not exercise effective oversight over foster care placements, which subjects class members to risks of harm, allegedly unreasonable ones. Plaintiffs have presented evidence that the State’s array of placements leads to frequent placement moves, as well as evidence suggesting that frequent and long-distance placement moves cause psychological harm to class members. So, while class members do not have a per se Fourteenth Amendment right not to be removed a certain distance from their home communities, Plaintiffs have made out a more complex claim. They claim that the State’s placement array, as it currently stands, threatens foster care children with an unreasonable risk of harm. Alternatively, this can be phrased as the State has not implemented reasonable safeguards to prevent class members from deterioriating. The validity of Plaintiffs’ claim—this one and all of them in this case—depends on the connection between the policy or practice in question and the risk of harm it imposes on the relevant class or subclass. Plaintiffs have produced sufficient evidence, at this point in the litigation, to establish that the State’s placement array may cause harm to members of the LFC Subclass. They had made out a claim. It remains to be seen whether this risk rise to the level of being unconstitutional and whether Defendants displayed either deliberate indifference or a lack of professional judgment in setting up the placement array. As with Plaintiffs’ first claim on behalf of the Licensed Foster Care Subclass, the Court can identify common questions of law and fact that will resolve this claim. An important one will be whether DFPS policies and practices have, as Plaintiffs allege, caused Texas’ placement array to be deficient, and whether any possible deficiencies of the State’s array of placements 79 / 107 rise to the level of unconstitutional risk of harm. These questions are common ones for the class and will decide the validity of the claim. B. Typicality As these claims are based on the Defendants’ policies and practices that affect the LFC Subclass as a whole, the named plaintiffs satisfy typicality.14 The same policies affect the named plaintiffs, and in the same way, as they affect other subclass members. Further, it bears noting that children in the PMC frequently move from placement to placement. E.g., (DX 45 at 34; PX 276). So, even a member of the subclass who is presently in a safe, well-monitored placement near her home community might, with a significant probability, end up transferred to another placement. That transfer subjects her to the risks Plaintiffs allege: she may be forced to relocate far from her family, community, and social support system; she may be moved to an inappropriate placement; or she may find herself in an unsafe, inadequately monitored placement. All named representatives of the Licensed Foster Care Subclass depend on RCCL to conduct diligent oversight in order to ensure they are placed in safe homes and facilities. They all have claims typical of the class because no individual claims hinge on particular facts about their facility—instead they turn on the general oversight RCCL is able to provide. Similarly all named representatives, by virtue of being in the subclass, are affected by the placement array and as such have claims typical of the class as a whole. It may, of course, be true that individual named plaintiffs could have other, individual claims arising out of the oversight of their facility or their personal placements, but those are not the claims brought in this litigation. 14 D.I., due to changes in his status, is excluded from this analysis and will be considered in Section XI., infra. 80 / 107 C. Rule 23(b)(2) Plaintiffs ask for two injunctions. First, they ask the Court to require an assessment by professionals in order to determine what needs to be done regarding monitoring and enforcement of minimum standards for licensed foster care placements and for those recommendations to be implemented. Second, they ask that the Court require similar assessment and implementation with regards to the array of such placements. (D.E. 179 at 31). These assessments, rather than providing tailored relief for individual members of the subclass, would be intended to determine what practical steps need to be taken on a department-wide level to vindicate the constitutional rights of members of the subclass. These requested injunctions bear some superficial resemblance to relief sought previously by Plaintiffs that did not satisfy Rule 23(b)(2). In their First Complaint Plaintiffs asked for expert assessments on class members’ individual cases and placement needs. (D.E. 1 at 82-83). This is tantamount to individually-tailored relief, which cannot be granted in a class action certified under Rule 23(b)(2). MD, 675 F.3d 832, 846-47 (5th Cir. 2012). The expert assessments contemplated here, however, are intended to consider only the needs of the class as a whole. If Plaintiffs prevail, then the Court could identify a minimal threshold for an array of placements necessary to safeguard subclass members’ Fourteenth Amendment rights.15 Similarly, the Court could identify specific RCCL practices, such as allowing for phone followup inspections or the way the agency treats repeated violations, and order DFPS to modify them. Such equitable relief would not involve the Court in adjudications of individual class members’ needs or circumstances. The possibility of such relief satisfies Rule 23(b)(2). 15 Such a mandated array could require the placement array to be more sensitive to the population distribution or population trends in Texas. 81 / 107 IX. Foster Group Home Subclass Plaintiffs allege that placing PMC children in foster group homes violates their Fourteenth Amendment rights because such homes lack professional staff, nighttime awake supervision, and medical personnel on staff or on call. (D.E. 195 at 25, 27). Plaintiffs claim that foster group homes are in fact group facilities, like GROs, because of the number of children placed in them. Texas’ decision to hold them to different, less stringent standards than other group facilities therefore subjects members of the subclass to an unreasonable risk of harm. In essence, Plaintiffs claim that foster group homes, due to their structure and the standards they are held to, do not provide sufficient oversight or monitoring in order to keep subclass members reasonably safe. A. Commonality This claim is based on the institutional structure of foster group homes, so it is naturally a class one. Plaintiffs have identified a policy—the way foster group homes are structured—and allege that it exposes class members to an unreasonable risk of harm. This claim patently relates to a uniform policy: the existence, requirements and institutional features of foster group homes are explicit policies adopted by DFPS. See, e.g., (PX 50); (PX 390). Within the regulatory scheme they are little different than foster family homes. Texas defines them solely by reference to their size.16 Nearly all the licensing requirements are the same for these facilities and foster family homes. Screening is conducted in the same way, as is verification. (PX 50 at §§ 749.2445-749.2451, 749.2471). The physical environment is subject to the same regulations save, of course, that foster group homes will be required to have more 16 Foster family homes are verified, managed by a CPA, and have 6 or fewer children while foster group homes are verified, managed by a CPA, and have 7 to 12 children. (PX 50 at § 749.43(22-23)). 82 / 107 space in order to accommodate more children. Id. at §§ 749.3021-749.3041. Neither facility requires trained professionals on staff or awake nighttime supervision. The differences beyond capacity are few and minimal.17 Foster group homes can house up to 12 children. The caregiver to child ratio can be 8:1, and if the home is the primary residence of at least one caregiver they are permitted to exceed this ratio during waking hours for short periods and during nighttime hours so long as there is a safety plan for nighttime supervision. Id. at §749.2567. These features create a greater risk of harm for children placed in these homes since they will generally receive less supervision. Children in the foster care system abusing other children in the system is a major concern. As described above, many children coming into the PMC have suffered physical or sexual abuse and are traumatized. See e.g., (PX 322 at 11; DX 51 at DFPS005458044). These awful experiences make it more likely that the children will abuse or engage in inappropriate behaviors with other foster care children. (DX 51 at DFPS005458044); Class Cert. Hr’g 1/22/13 (Miller) at 245-46. For example, when named plaintiff D.I. was eight years old he was placed in a foster group home. Over the period of a month he was sexually abused at least three times by two teenage boys. The lone caregiver at the placement had no idea this was happening until one of the perpetrators approached him and informed him. (PX 26 at 1 RFP 2 RCCL 00930 (filed under seal)). There were only five boys in the home at the time, but the facility was—and continues to be—licensed at a capacity of 10 children. Id.; (PX 301 at 6). DFPS has recognized this risk. For example, on November 14, 2012 it proposed a number of recommendations intended to curb or better deal with “child to child inappropriate sexual acting out or physical 17 For example, foster group homes, in contrast to foster family homes, cannot provide regular day care. (PX 50 at § 749.2493). And, in contrast to some foster family homes, the required fire inspection at a foster group home may be conducted by the CPA only if it is unable to have a fire inspection done by a certified fire inspector or a local fire authority. Id. at §§ 749.2903-749.2905. 83 / 107 aggression.” (DX 51 at DFPS005458044 (DFPS Memorandum for the Commissioner)). Plaintiffs have not been able to present aggregate statistics regarding child on child abuse because Texas does not collect these statistics. (D.E. 195 at 27 (citing Class Cert. Hr’g 1/22/13 (Miller) at 234-35)). Although it appears to be a problem serious and widespread enough for DFPS to take action and revise policy. The way foster group homes are structured exacerbates these risks. These placements are authorized to house a relatively large number of children of varying ages and service levels. DFPS categorizes children according to their level of need, ranging from basic to moderate to specialized and finally to intense. (PX 312 at DFPS004766762-66). Based on the evidence before the Court, it does not appear that intense level children are placed in foster group homes. See (PX 199 at DFPS001141310 (e-mails from May 2009)). However, as of May 31, 2012 there are numerous foster group homes with children with both basic and specialized levels of needs. (PX 298). The difference between children at these service levels is considerable. A child at the basic service level is characterized by transient difficulties and occasional misbehavior; brief episodes of acting out in response to stress; and behavior that is minimally disturbing to others, is considered typical for the child’s age, and can be corrected. (PX 312 at DFPS004766762). A child at the specialized service level “has severe problems in one or more areas of functioning” and can have characteristics such as unpredictable anti-social acts; frequent or unpredictable physical aggression; major action of self-injury, including recent suicide attempts; and can present a significant risk to herself and to others. Id. at DFPS004766764. Similarly, foster group homes with an eight or ten year difference in the residents’ ages are common. (PX 298). Mixing children of different ages and service levels—especially when the differences are this stark—is risky. It increases the probability of one child abusing another, as illustrated by the case of D.I 84 / 107 and expert testimony. See Class Cert. Hr’g 1/22/13 (Ricker) at 132; id. (Miller) at 241; Section X.A., infra. By dint of having more children in the home, Plaintiffs allege that the caregivers are not able to provide the children with the supervision and attention they need, especially given their special needs. As Anna Ricker, the next friend for J.S., H.V., and P.O., put it: “it’s more difficult with more kids and these are difficult kids to start with.” Class Cert. Hr’g 1/22/13 (Ricker) at 131. In explaining DFPS’s policy of not placing children at the intense level in foster group homes, Ms. Gonzalez referred to this concern: “The theory behind this issue is related to the amount of supervision, attention and multiple demands that a child with intense level needs would place on the foster parents. In theory, the more children in a home the less supervision, attention and ability for the parents to meet the multiple needs of the child with these kind of needs.” (PX 199 at DFPS001141310). Further, the point is intuitive: the more children each caregiver is responsible for, the less time, attention, and supervision is allotted to each. Foster group homes, by allowing for a greater number of children without a corresponding increase in the number of caregivers, subject class members to less supervision and more risk than a foster family home. Congregate care facilities like GROs, of course, have an even greater number of children. But, these facilities contain a number of safeguards that foster group homes lack. In effect, Texas treats foster group homes as simply very large foster homes, and Plaintiffs argue that this practice creates an unconstitutionally unreasonable risk of harm. Texas foster group homes are unusual sorts of facilities in foster care systems. COA does not have best practice standards for them, because “we don’t believe that there are best practices to go along with bad practice.” (PX 7 at 135 (affirmation of Richard Klarberg, President and CEO of COA)). CWLA does recommend the use of group homes that house between 4 and 12 85 / 107 children as part of an array of residential services, but these are significantly different from the foster group homes used in Texas. (DX 57 at 35). These facilities have lower child to caregiver ratios generally, access to 24-hour professional services, nighttime awake supervision, and are designed for children who need more supervision than is available in a family home but do not need the intensive residential service that would be provided at a larger facility. Id. at 41-43. In the context of this case these would be something like a small GRO or RTC. Texas foster group homes are not used as an intermediate type of placement between institutional and family settings. Instead they are used as a way to place more children within a home without imposing any meaningfully different regulations on it. In order to succeed on the merits, Plaintiffs will need to prove that foster group homes subject class members to an unreasonable risk of harm. It is not enough that foster group homes entail less supervision, increasing the risks that class members face. The risks that members of this subclass are subjected to due to the structure of foster group homes must rise to the level of being unconstitutionally unreasonable. Put another way, foster group homes may be the least safe kinds of placements in the Texas foster care system, but being more dangerous than both ordinary foster family homes and GROs does not, in and of itself, make Defendants’ policy of having foster group homes a violation of class members’ Fourteenth Amendment rights. This forms the basis of the common questions of law and fact that will be dispositive of this claim: do foster group homes subject class members to less supervision and therefore an increased risk of harm, and is this risk an unconstitutionally unreasonable one? As Plaintiffs have identified a common practice or policy that is the cause of the alleged harm and there exist common questions of law and fact that will be dispositive of this claim, the Foster Group Home Subclass claim has satisfied Rule 23(a)’s commonality requirement. 86 / 107 B. Typicality Plaintiffs’ common claim is that the features of foster group homes subject members of this subclass to an unreasonable risk of harm. H.V., the sole named plaintiff for this class, was taken into custody by DFPS in 2006. Since then he has had eight different placements and has been moved between foster family homes, foster group homes, GROs, and placements with family. Since September 2011 he has resided in a foster group home. (PX 33 (filed under seal)). H.V. possesses a typical claim because he, like all members of the class, is in a placement that has the features Plaintiffs claim expose them to an unreasonable risk of harm. His foster group home, like all foster group homes, is subject to regulations almost identical to those imposed on foster family homes except that there is a larger capacity and a less stringent child to caregiver ratio. As such, his claim is typical of the subclass claim. Defendants object that H.V. cannot have a claim typical of class members because H.V.’s next friend, Ms. Ricker, has never objected to his placement in a foster group home in a pleading and cannot identify any specific harm H.V. has suffered in a foster group home. (D.E. 198 at 32). The Court fails to see how not objecting to the placement in a pleading shows that the claim is atypical—it is clear that Ms. Ricker believes that foster group homes expose residents to some risks. Class Cert. Hr’g 1/22/13 (Ricker) at 129; 132-33. Regardless of the specific foster group home H.V. is currently placed in, he is still exposed to the same alleged risks that other class members are; his claim arises from the same conduct and on the same legal theory. 87 / 107 C. Rule 23(b)(2) The subclass also satisfies the requirements of Rule 23(b)(2). Plaintiffs claim that they are all injured in the same way because Defendants’ policies regarding foster group homes, rather than particular circumstances within an individual group home, subject them to an unreasonable risk of harm. They request a permanent injunction halting placement in foster group homes until they are operated in compliance with “accepted professional standards.” (D.E. 179 at 31). In essence, Plaintiffs seek to eliminate the distinction between foster group homes and GROs so that the same standards govern both types of foster care placements. E.g., Class Cert. Hr’g 1/22/13 (Lowry) at 46, 48; id. (Miller) at 244. The Court could also order more targeted injunctions that modify some of the critical characteristics of foster group homes. The Court could determine that these placements require nighttime awake supervision, medical staff on call, and so on. In either case, there are a number of possible injunctions in accordance with Rule 65(d) that the Court can imagine. Injunctive relief would be appropriate, if Plaintiffs prevail, because the claim and injury concern general policies of Defendants and the Court could issue an injunction that specifies policy changes required of Defendants as to these facilities. X. Basic Care GRO Subclass Plaintiffs claim that the Fourteenth Amendment rights of the putative Basic Care GRO Subclass are violated because Texas uses GROs to house children who only need basic child care services. (D.E. 195 at 20). Non-emergency, basic child care services are defined as services “that meet a child’s basic need for shelter, nutrition, clothing, nurture, socialization and interpersonal skills, care for personal health and hygiene, supervision, education, and service 88 / 107 planning.” (D.E. 179 at 8 (citing 40 Tex. Admin. Code § 748.61)). GROs are more restrictive than other placements and Plaintiffs allege that they should not be used for non-emergency, nontherapeutic reasons. They argue that the policy of housing children who need only basic child care services in these facilities creates an unreasonable risk of harm because, by nature, these facilities are inappropriate and unsafe for these children. (D.E. 179 at 18). A. Commonality Plaintiffs’ identification of a systematic deficiency in regards to the Basic Care GRO Subclass is fairly straightforward. They allege that Texas has a policy of placing children who need only basic child care services in large group living facilities and that this violates their constitutional rights. To show a common policy or practice to that effect, Plaintiffs have provided evidence that 385 PMC children whose level of service requirements were listed as basic were placed in GROs as of May 31, 2012. (PX 299 (filed under seal)); (D.E. 160 at 60); see also Class Cert. Hr’g 1/23/13 (Butler) at 138-39 (stating that children with basic care service levels are placed in GROs and there is no special approval required to make such placements). Since Plaintiffs’ claim is that Texas unconstitutionally allows this class to exist, the showing of numerosity also shows that there is a policy at work that is established by the consistent practice of making such placements. See (D.E. 202 at 6) (“With respect to the Basic Care GRO Subclass … the structural deficiency is that Defendants make these placements at all”). There is no question that Defendants have a practice or policy of placing children with only basic needs in GROs. For example, there is evidence in the record describing numerous children at the basic of moderate service level who have been “lingering in RTCs” for over three months. (PX 173 (listing eight children, one of whom was in an RTC for five and a half months); PX 174 (listing 89 / 107 the total number of moderate level children in RTCs as of June 30, 2011 at 199 children); PX 175 (referring to 115 children at the basic or moderate level still in RTCs); PX 214). That is, these are not children whose conditions had recently improved and were simply waiting to find another placement; they had been in an RTC at this service level for an extended period of time. E.g., (PX 174 at DFPS000889283; PX 173 at DFPS000889234). According to communications between DFPS personnel, these placements are “not appropriate unless there were unique cases why children had to remain there. -Children [sic.] must be moved out also to make space for other children who really do need to be in RTC’s.” (PX 175); see also (PX 173 at DFPS000889233 (stating that it was “not ok” to have a service plan where a child at the moderate level was placed in an RTC till she graduates)). Plaintiffs point to two kinds of risks created by this practice. First, like foster group homes, GROs often contain children of varying ages and service levels. Almost by definition, members of this subclass will be placed alongside those with relatively intense service needs as many GROs are licensed to provide a wide range of services in the same setting. (PX 302). A list of children in these facilities as of May 31, 2012 reveals that some had as much as a 17 year gap between the oldest and youngest children placed there, and children with levels of care ranging from basic to specialized resided in the same facilities. (PX 299 (filed under seal)). As noted above, children at different ages and with different level of care categorizations have very different needs. And, mixing children in this way increases the risk of child on child abuse. As Dr. Miller put it: “if you’re mixing ages and you’re mixing children with conduct disorders or behavioral issues and children who are basic care children. You create a pretty risky situation.” Class Cert. Hr’g (Miller) 1/22/13 at 241. Ms. Ricker described the risks inherent in GROs because they place numerous children with special needs in the same environment: “when you 90 / 107 put a whole bunch of kids with problems together. If one of them craters, then it’s like a volcano. They all bubble over.” Id. (Ricker) at 122. Plaintiffs’ claim is that putting class members in that environment—when they, as the subclass is defined, have no therapeutic need to be there—subjects them to an unreasonable risk. Plaintiffs seem to be arguing that for children with more intense or specialized needs the risks of a GRO may be reasonable. The child will benefit from the treatment; there is some value to it that justifies the risks entailed. For children who only need basic care, however, this is not the case. While it is true that GROs have safeguards that foster group homes lack, the risk still exists, though it remains a matter for trial to determine its magnitude. GROs represent an increased risk as they house considerably more children—the highest capacity authorized for a GRO as of May 31, 2012 is 437, (PX 302), whereas foster group homes are capped at 12—and the difference in ages present in the same placement is larger. Furthermore, the increased supervision in GROs is far from perfect. As part of its risk analysis in the 2012 Assessment, the PMU determined that nearly 60% of all allegations or serious incidents at a GRO or RTC reported had a lack of supervision trend. (PX 296 at DFPS04490546). Second, Plaintiffs contend that because these facilities are designed to address serious therapeutic needs they pose a risk to class members’ psychological well-being as they effectively institutionalize these children for no good reason. (D.E. 195 at 23). Ms. Ricker testified that when he was about six years old J.S. was placed in a GRO where he lived in a cinderblock room with only a bed and some clothes. She related that “he said ‘would you please hug me because they can’t touch me here.’ And I was horrified.” Class Cert. Hr’g 1/22/13 (Ricker) at 110. Ms. Vasquez described the facility that L.H. and C.H. were placed in because DFPS could not find a family placement that would keep the sibling group together as devoid of personalization and “an 91 / 107 institution, a facility that just houses them … just a box that these kids are thrown in …” Class Cert. Hr’g 1/23/13 (Vasquez) at 168. It is not a stretch to believe these conditions can inflict emotional or psychological harm on class members. As detailed above, class members’ Fourteenth Amendment rights include a right to be free from an unreasonable risk of psychological and emotional, as well as physical, harm. E.g., K.H. through Murphy v. Morgan, 914 F.2d 846, 848 (7th Cir. 1990); R.G. v. Koller, 415 F.Supp. 2d 1129, 1156 (D. Haw. 2006); Marisol A. by Forbes v. Giuliani, 929 F.Supp. 662, 675 (S.D.N.Y. 1996) aff’d sub nom. Marisol A. v. Giuliani, 126 F.3d 372 (2d Cir. 1997); Aristotle P. v. Johnson, 721 F.Supp. 1002, 1010 (N.D. Ill. 1989). There is also some evidence that children who have been placed in GROs tend to experience more placements, and frequent placement moves can inflict psychological harm in violation of the class members’ Fourteenth Amendment rights. (PX 276);18 Section VII.B., supra. There seems to be a general agreement that these kinds of placements are disfavored, especially if they are long-term. Class Cert. Hr’g 1/23/13 (Butler) at 140 (stating that no caseworker would ever request a GRO over a foster family home or a foster group home); Class Cert. Hr’g 1/22/13 (Miller) at 242, 246; (DX 57 at 43-45 (CWLA standards recommending that a child’s stay in a residential treatment program institutional setting should not exceed the time it takes for a child to be able to move to a less restrictive environment)). To sum up, Plaintiffs claim that placements of this sort are not only inappropriate, but create an unreasonable risk of harm to class members’ physical and psychological well-being. Of course, as with all of Plaintiffs’ claims they still bear the burden of proving its merits. They have satisfied their burdens at class certification, though, by identifying a unified policy or practice and connecting that policy to the alleged harm. Common questions of law and fact that 18 This data relates to children in GROs (but not RTCs) generally; it does not separate out children in GROs who only need basic care. 92 / 107 will decide this claim include whether the risks Plaintiffs point to are sufficiently dire that they are unreasonable, constitutionally-speaking, and whether Defendants acted with the requisite deliberate indifference or failure to take professional standards into account. B. Typicality Typicality in this subclass follows naturally from commonality. C.H. and L.H. are classified as needing only basic child care services and are placed in a GRO. This is the basis of the claims for all members of this subclass. Particularities regarding their circumstances do not undermine the typicality of their claims because those particularities are not relevant to the very general constitutional claims Plaintiffs have chosen to assert here. Defendants argue, however, that named plaintiffs C.H. and L.H. fail to satisfy typicality because their next friend, Ms. Vasquez, asked to have them placed there so that they could be with their siblings. (D.E. 198, at 26-27). Under the circumstances, she believed placement in the GRO was the best option for them. But, as subsequent testimony made clear, Ms. Vasquez was not happy with their placement in the GRO; she just thought it was the best option available given the circumstances. Class Cert. Hr’g (Vasquez) 1/23/13 at 166-68. Believing that the current placement is the best of a set of very bad options does not entail that their claims are atypical. Since the claim asserted on behalf of the class is not based on specific features of their placements but instead simply on the policy of making such placements at all, the claims asserted by C.H. and L.H. are the same as those of the subclass.19 The Court concludes that C.H. and L.H. have claims typical of the Basic Care GRO Subclass. 19 Defendants also argue that typicality fails because Plaintiffs have not presented evidence related to the psychological harm that these plaintiffs have suffered, but as pointed out above, this mistakes the nature of the injury Plaintiffs are claiming. 93 / 107 C. Rule 23(b)(2) Plaintiffs claim that all subclass members are injured by being placed in these institutions and request that the Court issue an injunction halting placement of children only needing basic child care services in non-emergency GROs.20 (D.E. 179 at 31). The nature of this claim makes the Rule 23(b)(2) analysis straightforward: Plaintiffs object to a practice and if they prevail on the merits the Court could conceivably enjoin that practice. The Court could, for instance, determine that the Fourteenth Amendment requires that only PMC children with special therapeutic needs be placed in GROs. Since which facilities are GROs and which children are classified as only needing basic child care services are easily discernible by the definitions and categorizations DFPS currently uses, the Court could provide a clear injunction as to Defendants’ policies in regards to those facilities and those children. Defendants counter that Rule 23(b)(2) is not fulfilled because Plaintiffs have not shown classwide injury that violates the class members’ constitutional rights. (D.E. 198, at 51-52). That is a merits inquiry, though. The inquiry here is not whether Plaintiffs have earned an injunction that provides relief to the class as a whole. Defendants are right that they have not done so at this point. Here the Court only decides that this claim is of the sort that if Plaintiffs do prevail as a class, then appropriate equitable classwide relief can be fashioned. XI. Unverified Kinship Subclass Plaintiffs allege that the Fourteenth Amendment rights of the members of the putative Unverified Kinship Subclass are violated because they are placed in homes that have not gone through the verification process to ensure they meet the standards of other foster homes, their 20 An emergency placement is a short-term one utilized in exceptional circumstances, such as when a child must be unexpectedly removed from her current placement. 94 / 107 providers do not receive necessary training, and their homes receive inadequate monetary assistance. (D.E. 179 at 19). They claim that these policies place members of the putative subclass at an unreasonable risk of harm. A. Defendants’ Motion for Partial Summary Judgment Concurrent with their Post-Hearing Brief in Opposition to Class Certification, Defendants filed a Motion for Partial Summary Judgment as to the claims of D.I., S.R., M.R., and J.R. on the grounds that their individual claims are now moot.21 D.I. was adopted on October 24, 2012 and so is no longer in the PMC. As of February 7, 2013, S.R., J.R., and M.R. have had conservatorship transferred to relatives and are no longer in the PMC. Defendants assert that the R. children are the sole members of the Unverified Kinship Subclass and hence the claims of that subclass must be dismissed as moot as well. (D.E. 199). Plaintiffs oppose this motion, though they do not dispute the underlying facts. (D.E. 208). Concurrent with their response to Defendants’ motion Plaintiffs filed a Motion for Leave to File a Fourth Amended Complaint that adds two additional named plaintiffs, S.S. and A.R., as representatives of the putative Unverified Kinship Subclass. (D.E. 205). Plaintiffs have also filed a Motion for Leave to Reply in support of that motion and attached to that motion a corrected Fourth Amended Complaint. (D.E. 212). Since the Court finds that D.I., S.R., J.R., and M.R. cannot be adequate representatives because they are no longer members of the General Class or any subclass, Defendants’ Motion for Partial Summary Judgment is moot and does not need to be considered here. 21 D.I. has been offered as a representative of the General Class and the Licensed Foster Care Subclass. But since the issues regarding mootness are the same for D.I. and the R. children and the ramifications for this case concerning the motion relate mostly to the Unverified Kinship Subclass, D.I. is considered in this section. 95 / 107 B. Certification of the Unverified Kinship Subclass Plaintiffs bear the burden of demonstrating that they satisfy all of the elements of Rule 23, including showing that the named plaintiffs for a class or subclass can be adequate representatives of that class or subclass. Wal-Mart v. Dukes, 131 S.Ct. 2541, 2551 (2011). They have failed to carry this burden as to the putative Unverified Kinship Subclass and hence the Court finds it unnecessary to consider at this time the remaining requirements of Rule 23 for that subclass. In their Third Amended Complaint Plaintiffs offered four named plaintiffs to represent the Unverified Kinship Subclass: S.R., J.R., M.R., and P.O. (D.E. 179 at 29). In order to be an adequate representative of a class a plaintiff must show that she is willing and able “to take an active interest in and control the litigation to protect the interests of absentees.” Stirman v. Exxon Corp., 280 F.3d 554, 563 (5th Cir. 2002). The R. children are no longer in the subclass because they are no longer in the PMC. As they are no longer in the subclass their personal interests are no longer aligned with those of other members of that subclass. They no longer suffer the same alleged injuries and they would not benefit from the sought after injunctive relief. Their next friend, as such, cannot adequately pursue this litigation on their behalf. At the end of August 2011 there were 2,994 PMC children in unverified kinship homes. (PX 2 at APPX 2077). There are likely a commensurate number today. The subclass would be better represented by named plaintiffs who are still members of the subclass and with next friends who can pursue the litigation on behalf of those children for their particular benefit. In any case, exiting the subclass is a major development for these named plaintiffs in relation to this case, and Plaintiffs have failed to affirmatively show that the R. children can be adequate representatives given this significant change. Without such a showing, the Court cannot certify a class naming 96 / 107 them as representatives. Additionally, having exited the PMC the R. children cannot be adequate representatives of the General Class. For the same reasons D.I. cannot serve as an adequate representative of the General Class or the Licensed Foster Care Subclass. The remaining question is whether or not P.O. can serve as an adequate representative of the Unverified Kinship Subclass. At a minimum, to show that a named plaintiff can be an adequate representative it must be demonstrated that that named plaintiff is or at least has been a member of that subclass at some point. Plaintiffs have claimed that P.O. is a member of this subclass, but the record makes his status uncertain. At the Class Certification Hearing his next friend could only say that he was in a kinship placement and was not sure whether it was verified or not. Class Cert. Hr’g, 1/22/13 (Ricker) at 123-25. The case file for P.O. Plaintiffs have submitted to the Court last documents his placement over a year ago and does not have him in a kinship placement at all. (PX 33 (filed under seal)). Given P.O.’s uncertain membership in the putative subclass, Plaintiffs have not established that P.O. can be an adequate representative. Plaintiffs have failed to carry their burden to demonstrate adequacy of representation as to the Unverified Kinship Subclass, so at this time the Unverified Kinship Subclass cannot be certified. C. Plaintiffs’ Motion for Leave to File a Fourth Amended Complaint The Court next turns to Plaintiffs’ Motion for Leave to File a Fourth Amended Complaint. The Federal Rules of Civil Procedure provide that a party may amend its pleading once as a matter of course. Fed R. Civ. P. 15(a)(1). Beyond that “a party may amend its pleading only with the opposing party’s written consent or the court’s leave.” Fed. R. Civ. P. 97 / 107 15(a)(2). Defendants oppose Plaintiffs’ motion to file a Fourth Amended Complaint. (D.E. 205, 209). Motions for leave to amend are entrusted to the sound discretion of the district court. Quintanilla v. Tex. TV, 139 F.3d 494, 499 (5th Cir. 1998) (citing Wimm v. Jack Eckerd Corp., 3 F.3d 137, 139 (5th Cir. 1993)). Rule 15(a) favors granting motions for leave to amend and though such motions should not be automatically granted, a “district court must possess a ‘substantial reason’ to deny a request for leave to amend.” Jones v. Robinson Prop. Group, L.P., 427 F.3d 987, 994 (5th Cir. 2005) (quoting Lyn-Lea Travel Corp. v. Am. Airlines, 283 F.3d 282 (5th Cir. 2002)); see also Fed. R. Civ. P. 15(a)(2) (“The court should freely give leave when justice so requires.”). Courts may consider a variety of factors when determining whether “substantial reason” exists for denying a motion for leave to amend, including undue delay, bad faith or dilatory motive, repeated failure to cure a deficiency, undue prejudice to the opposing party, and the futility of the amendment. Foman v. Davis, 371 U.S. 178, 182 (1962); United States ex rel. Steury v. Cardinal Health, Inc., 625 F.3d 262, 270 (5th Cir. 2010); Wimm, 3 F.3d at 139; Dussouy v. Gulf Coast Inv. Corp., 660 F.2d 594, 598 (5th Cir. 1981). Plaintiffs have filed three amended complaints. In each successive amended complaint the claims and classes have remained constant. Plaintiffs’ changes have only been to add or alter named plaintiffs in order to ensure that there was adequate representation for the putative General Class and four subclasses. The same is the case with the Fourth Amended Complaint— Plaintiffs only seek to add two named plaintiffs to ensure that going forward there are adequate representatives for all of the putative classes. This litigation has gone on for over two years and the claims of the named plaintiffs are time-limited because they will all be adopted, have custody otherwise transferred, change 98 / 107 placements, or age out of the system. The Court has previously stated that it would permit Plaintiffs to re-plead multiple times and would let them “liberally amend.” Tr. of Tel. Conference, May 2, 2012 at 10, 14. The particular deficiency they seek to correct is new to develop. There has been no undue delay in correcting it, bad faith, or repeated failure to cure a deficiency by the Plaintiffs in this case. They filed their Motion for Leave to Amend concurrent with their response to Defendants’ Motion for Summary Judgment raising this deficiency. Defendants argue that amendment is futile and hence leave to amend should be denied. (D.E. 209 at 2). Futility of amendment applies when there is a defect the court has found in the complaint that cannot be cured by the proposed amendment and so allowing it would be a waste of judicial resources. See, e.g., Cent. Laborers’ Pension Fund v. Integrated Elec. Servs., 497 F.3d 546 (5th Cir. 2007) (affirming a district court ruling denying leave to amend when it was sought after the case was dismissed and the proposed amendment did not remedy the deficiency); Rosenzweig v. Azurix Corp., 332 F.3d 854 (5th Cir. 2003) (affirming a ruling to deny leave to amend sought after dismissal partially on the grounds of futility when the new facts and claim plaintiffs wished to add did not remedy their failure to meet the pleading standard); Emory v. Tex. State Bd. of Med. Exam’rs, 748 F.2d 1023 (5th Cir. 1984) (affirming an order denying a motion for leave to amend to add additional defendants for futility when the same defect applied to the additional defendants). The Fifth Circuit has interpreted futility in this context “to mean that the amended complaint would fail to state a claim upon which relief could be granted.” Stripling v. Jordan Prod. Co., 234 F.3d 863, 273 (5th Cir. 2000). Here, Defendants extend this interpretation and argue that the motion for class certification should be denied as to the claims of the Unverified Kinship Subclass for the reasons they have been arguing regarding 99 / 107 “commonality and cohesiveness” in their previous pleadings and that adding new named plaintiffs to represent that class does nothing to remedy those defects. Defendants are correct that the amendment does nothing to change the terrain of the motion for class certification regarding commonality. But that is not the purpose of the amendment. If the Court finds that commonality has not been satisfied then the subclass will not be certified in the future. But if the Court were to find that those requirements have been met for this subclass then Defendants’ argument has no force here. This is not a case where the court has determined there is a defect, the plaintiff seeks to amend to remedy the defect, and the court determines the amendment is futile and so denies the motion for leave to amend. The relevant defect is the fact that the R. children have quite recently changed custody and can no longer adequately represent the subclass, and it is that defect the amendment squarely addresses. Related to that defect, the amendment is not futile. Defendants also argue that the amendment should not be allowed because it contains “key allegations” as to D.I., S.R., M.R., and J.R. that are “simply and significantly false.” (D.E. 209 at 2-3). These allegations are statements, copied from earlier complaints, stating that each of the above is in the PMC of the State of Texas. This is no longer the case, as all admit. As Defendants point out, ordinarily the remedy for such a mistake would be for the plaintiff to withdraw and remove the errors or for the defendant to answer and deny the assertion. They claim, however, that this case is different because it will waste resources by necessitating a motion for summary judgment and will raise ethical issues. (D.E. 209 at 3 n.4). The Court fails to see either unpalatable consequence. Further, it is unclear which well-established rationale for denying a motion for leave to amend this would qualify as—Defendants gesture towards bad faith but “are reluctant to, and do 100 / 107 not yet assert” it. Id. at 3 n.5. Their sole support for this suggestion, Wimm, 3 F.3d at 137, is a case where leave to amend was denied when plaintiffs were aware of the basis for a particular claim and chose not to raise it until summary judgment was imminent on their preferred claim. Inferring bad faith from tactical decisions on which claims to make is quite different from the inference Defendants suggest the Court make—but are unwilling to make themselves—that a mistake in updating the complaint to reflect changed facts constitutes bad faith. In any case, the claims Defendants complain of pertain only to named plaintiffs who have been found not to be adequate representatives for the subclass. Moreover, Plaintiffs, made aware of the mistakes by the Defendants, have now corrected the mistakes and propose to file a corrected Fourth Amended Complaint. The Court grants Plaintiffs’ Motion for Leave to File Reply in Support of Motion for Leave to File Fourth Amended Complaint and considers the corrected version attached to their reply, (D.E. 212, Ex. 3), as the proposed Fourth Amended Complaint. A final reason to deny a motion for leave to amend is undue prejudice to the opposing party. “The touchstone of the inquiry under rule 15(a) is whether the proposed amendment would unfairly prejudice the defense by denying the defendants notice of the nature of the complaint.” Lowrey v. Tex. A & M Univ. Sys., 117 F.3d 242, 245 (5th Cir. 1997). Generally courts do not find undue prejudice when there is an adequate opportunity to address the amended complaint and complete discovery when needed. E.g. Williams v. Global Payments Check Servs., Inc., 2011 U.S. Dist. LEXIS 34568, at *3 (N.D. Tex. Mar. 30, 2011); MP Vista, Inc. v. Motiva Enters., 2010 WL 4366025, at *4 (E.D. La. Oct. 20, 2010). Contrarily, “[a] defendant is prejudiced if an added claim would require the defendant ‘to reopen discovery and prepare a defense for a claim different from the [one] … that was before the court.’” Smith v. EMC Corp., 101 / 107 393 F.3d 590, 596 (5th Cir. 2004) (quoting Duggins v. Steak ‘n Shake, Inc., 195 F.3d 828, 834 (6th Cir. 1999)); see also Little v. Liquid Air Corp., 952 F.2d 841, 846 (5th Cir. 1992) (affirming denial of a motion for leave to amend when the amendment “would have established an entirely new factual basis for the plaintiffs’ claims and, consequently, would have required that the parties reopen discovery and alter their trial strategies.”). Notably, Defendants do not argue that they are prejudiced by this amendment. It does not alter the claims and putative classes of the Plaintiffs and it does not change the substantive issues before the Court regarding class certification that have been extensively briefed. Indeed, Defendants have argued that the amendment is futile because the very same arguments about commonality apply. The Court has rejected this argument, but the general point is correct: the arguments in the case regarding the definition of the subclass, the claims asserted on its behalf, and most of the requirements of Rule 23 remain unchanged. The nature of the complaint has not changed, and in that sense Defendants are not prejudiced. This is true with regards to both the Fourth Amended Complaint and the corrected version Plaintiffs later submitted. What does change is the identity of the named plaintiffs representing the putative Unverified Kinship Subclass. Rule 23 requires, in addition to showing class commonality that has been the focus of the pleadings in this case, that Plaintiffs show that the claims of the representatives for each putative class or subclass are typical and that the named plaintiffs will be adequate representatives of that class or subclass. Courts must go beyond the pleadings to determine compliance with these requirements and Defendants should be afforded an adequate opportunity to challenge Plaintiffs’ meeting of them. Here it would prejudice Defendants—and would be premature for the Court—if Plaintiffs were allowed to file a Fourth Amended 102 / 107 Complaint with new named plaintiffs and immediately the Court were to rule on class certification, and in particular compliance with the requirements of Rule 23(a)(3)-(4). Nonetheless, here through no fault of Plaintiffs the status of several named plaintiffs has recently changed while they were diligently pursuing a motion for class certification. It would hence best serve the interests of justice to allow Plaintiffs to file their corrected Fourth Amended Complaint in order to include S.S. and A.R. but also to give Defendants a chance to conduct discovery regarding those additional named plaintiffs and the Unverified Kinship Subclass. Federal Rule of Civil Procedure 23(c)(1)(C) authorizes courts to alter and amend class certification orders before final judgment. The Supreme Court has described certification orders (both granting and denying) as “inherently tentative,” maintaining that “the judge remains free to modify it in light of subsequent developments in the litigation.” Gen. Tel. Co. of Sw. v. Falcon, 457 U.S. 147, 160 (1982); Coopers & Lybrand v. Livesay, 437 U.S. 463, 469 n.11 (1978). Other courts, including the Fifth Circuit, have made the stronger claim that district courts are obliged to reassess their class certification rulings throughout litigation. Richardson v. Byrd, 709 F.2d 1016, 1019 (5th Cir. 1983); Barnes v. Am. Tobacco Co., 161 F.3d 127, 140 (3d Cir. 1998) (citing Kuehner v. Heckler, 778 F.2d 152, 163 (3d Cir.1985)). Therefore, Plaintiffs will be permitted to file their corrected Fourth Amended Complaint and Defendants will be permitted three months discovery as to S.S. and A.R. and the Unverified Kinship Subclass. Plaintiff may then move to amend the certification order to add the Unverified Kinship Subclass and both parties will have an opportunity to brief the Court regarding whether that putative subclass satisfies the requirements of Rule 23. 103 / 107 XII. Appointment of Class Counsel Rule 23(g) requires that a court certifying a class must appoint class counsel. Rule 23(g)(1)(A) provides that the court must consider: i. the work counsel has done in identifying or investigating potential claims in the action; ii. counsel’s experience in handling class actions, other complex litigation, and the types of claims asserted in the action; iii. counsel’s knowledge of the applicable law; and iv. the resources that counsel will commit to representing the class. Fed. R.. Civ. P. 23(g)(1)(A). Rule 23(g)(4) requires that “[c]lass counsel must fairly and adequately represent the interests of the class.” Fed. R. Civ. P. 23(g)(4) The Court finds that counsel for the named plaintiffs fulfill these requirements. The attorneys of Children’s Rights engaged in a two year investigation of the Texas foster care system and a year of formal discovery in this case and have experience litigating similar cases in other jurisdictions. All proposed class counsel have extensive experience handling complex litigation and class actions. Counsel for named plaintiffs have demonstrated familiarity with the applicable law and dedication to this case. They have shown they will devote substantial resources to representing the class and pursuing this litigation. The Court concludes that counsel for named plaintiffs will fairly and adequately represent the interests of the class and appoints them as class counsel for the classes certified in this order. XIII. Conclusion IT IS HEREBY ORDERED: 1. Plaintiffs’ Motion for Class Certification is GRANTED in relation to Count I as to the General Class, the Licensed Foster Care Subclass, the Foster Group Home Subclass, and the 104 / 107 Basic Care GRO Subclass. Plaintiffs’ Motion for Class Certification is DENIED in relation to Count I as to the Unverified Kinship Subclass and in relation to Count II. 2. This action shall proceed as a class action with one General Class and three Subclasses: the Licensed Foster Care Subclass, the Foster Group Home Subclass, and the Basic Care GRO Subclass. 3. Membership of the certified General Class and each of the three certified subclasses is defined as follows: a. General Class: all children now, or in the future, in the Permanent Managing Conservatorship of the State of Texas; b. Licensed Foster Care Subclass: all members of the General Class who are now or will be in a licensed or verified foster care placement, excluding verified kinship placements; c. Foster Group Home Subclass: all members of the General Class who are now or will be in a foster group home; d. Basic Care GRO Subclass: all members of the General Class who are now or will be in a GRO and who are or will be receiving solely non-emergency, basic child care services. 4. The following named plaintiffs, by their next friends, are appointed as class representatives for the General Class: a. M.D., by her next friend, Sarah R. Stukenberg; b. Z.H., by his next friend, Carla B. Morrison; c. S.A., by her next friend, Javier E. Solis; d. A.M., by her next friend, Jennifer Talley; e. J.S., by his next friend, Anna J. Ricker; f. H.V., by his next friend, Anna J. Ricker; 105 / 107 g. P.O., by his next friend, Anna J. Ricker; h. K.E., by her next friend, John W. Cliff, Jr.; i. L.H., by her next friend Estela C. Vasquez; and j. C.H., by her next friend, Estela C. Vasquez. 5. The following named plaintiffs, by their next friends, are appointed as class representatives for the Licensed Foster Care Subclass: a. M.D., by her next friend, Sarah R. Stukenberg; b. Z.H., by his next friend, Carla B. Morrison; c. S.A., by her next friend, Javier E. Solis; d. A.M., by her next friend, Jennifer Talley; e. J.S., by his next friend, Anna J. Ricker; f. H.V., by his next friend, Anna J. Ricker; g. K.E., by her next friend, John W. Cliff, Jr.; h. L.H., by her next friend Estela C. Vasquez; and i. C.H., by her next friend, Estela C. Vasquez. 6. The following named plaintiff by his next friend, is appointed as class representative for the Foster Group Home Subclass: a. H.V., by his next friend, Anna J. Ricker. 7. The following named plaintiffs, by their next friends, are appointed as class representatives for the Basic Care GRO Subclass: a. L.H., by her next friend Estela C. Vasquez; and b. C.H., by her next friend, Estela C. Vasquez. 106 / 107 8. The following attorneys are appointed as co-counsel for the certified General Class and each of the three certified subclasses: a. attorneys from the firm Haynes and Boone, LLP; b. attorneys from the firm Yetter Coleman LLP; and c. attorneys from Children’s Rights, a non-profit organization with its office in New York. 9. Defendants’ Motion for Partial Summary Judgment as to D.I., S.R., M.R., and J.R. is DENIED as moot. 10. Plaintiffs’ Motion for Leave to File Reply in Support of Motion for Leave to File Fourth Amended Complaint is GRANTED. The Court considers the corrected Fourth Amended Complaint attached to that motion as the proposed Fourth Amended Complaint. 11. Plaintiffs’ Motion for Leave to File a Fourth Amended Complaint is GRANTED. Defendants are permitted three months to conduct discovery relating to the Unverified Kinship Subclass and the named plaintiffs S.S. and A.R. SIGNED and ORDERED this 27th day of August, 2013. ___________________________________ Janis Graham Jack Senior United States District Judge 107 / 107