M.D. v. Abbott

Full title: M.D.; bnf STUKENBERG, et al., Plaintiffs, v. GREG ABBOTT, et al.…

Court: UNITED STATES DISTRICT COURT SOUTHERN DISTRICT OF TEXAS CORPUS CHRISTI DIVISION

Date published: Dec 17, 2015

152 F. Supp. 3d 684 (S.D. Tex. 2015)

Opinion

CIVIL ACTION NO. 2:11-CV-84

12-17-2015

M.D.; bnf Stukenberg, et al., Plaintiffs, v. Greg Abbott, et al., Defendants.

Aaron Finch, Joshua Rosenthal, Adam Dembrow, Adriana Teresa Luciano, Christina Wilson, Elizabeth Pitman, Ira Lustbader, Jessica Polansky, Marcia Robinson Lowry, Melissa Cohen, Michael Kenneth Bartosz, Patrick Almonrode, Rachel Brodin Nili, Sarah T. Russo, Stephen Dixon, William Kapell, Children's Rights, New York, NY, R. Paul Yetter, Christopher D. Porter, Dori Kornfeld Goldman, Yetter Coleman LLP, Houston, TX, Barry F. McNeil, David Allen Dodds, Richard Thaddeus Behrens, Haynes and Boone LLP. Dallas, TX, for Plaintiffs. Andrew Bowman Stephens, Thomas A. Albright, Marc Rietvelt, Office of the Attorney General, James Patrick Sullivan, King Spalding LLP, Austin, TX, for Defendants.

Janis Graham Jack, Senior United States District Judge

Aaron Finch, Joshua Rosenthal, Adam Dembrow, Adriana Teresa Luciano, Christina Wilson, Elizabeth Pitman, Ira Lustbader, Jessica Polansky, Marcia Robinson Lowry, Melissa Cohen, Michael Kenneth Bartosz, Patrick Almonrode, Rachel Brodin Nili, Sarah T. Russo, Stephen Dixon, William Kapell, Children's Rights, New York, NY, R. Paul Yetter, Christopher D. Porter, Dori Kornfeld Goldman, Yetter Coleman LLP, Houston, TX, Barry F. McNeil, David Allen Dodds, Richard Thaddeus Behrens, Haynes and Boone LLP. Dallas, TX, for Plaintiffs.

Andrew Bowman Stephens, Thomas A. Albright, Marc Rietvelt, Office of the Attorney General, James Patrick Sullivan, King Spalding LLP, Austin, TX, for Defendants.

MEMORANDUM OPINION AND VERDICT OF THE COURT

Janis Graham Jack, Senior United States District Judge

Named Plaintiffs brought this class action under 42 U.S.C. § 1983 against officials of the State of Texas. Plaintiffs claim that Texas violates their Fourteenth Amendment substantive due process rights, including “the right to be reasonably safe from harm while in government custody and the right to receive the most appropriate care, treatment, and services” by how the State and its officials manage the Department of Family and Protective Services and the departments under its control. The Court has subject matter jurisdiction over this case under 28 U.S.C. § 1331. Because the issues to be decided are fact intensive, the Court's opinion is lengthy. TABLE OF CONTENTS

I. BACKGROUND ...689

A. Procedural History ...689

II. FINDINGS OF FACT AND LAW ...690

A. Overview of Texas Foster Care ...690

B. Plaintiffs' Claims ...693

C. Constitutional Rights ...695

D. Evidence Summary ...700

1. Reports ...700

2. National Standards ...701

3. Child and Family Service Reviews ...702

4. Witnesses ...703

a. Fact Witnesses ...703

b. Expert Witnesses ...709

5. Named Plaintiffs ...718

a. M.D ...718

b. D.I ...728

c. S.A ...731

d. A.M ...737

e. J.S ...741

f. H.V., J.V, and P.O ...745

g. Z.H ...758

h. K.E ...760

i. L.H. and C.H ...765

j. J.R., M.R., and S.R ...770

k. S.S ...771

l. A.R ...772

III. PRELIMINARY ISSUES ...773

IV. DISCUSSION ...776

A. General Class ...776

1. DFPS is Deliberately Indifferent Toward Caseload Levels ...790

2. DFPS Substantially Departs from Professional Judgment Toward CVS Caseworkers ...797

B. Licensed Foster Care Subclass ...798

1. Insufficient Oversight ...798

a. DFPS is Deliberately Indifferent Toward RCCL Operations ...804

b. DFPS Substantially Departs from Professional Judgment Toward RCCL Operations ...806

2. The State Maintains an Inadequate Placement Array ...808

a. DFPS is Deliberately Indifferent Toward its Placement Array ...813

b. DFPS Substantially Departs from Professional Judgment Toward its Placement Array ...816

C. Foster Group Home Subclass ...817

1. DFPS is Deliberately Indifferent Toward Foster Group Homes ...820

2. DFPS Substantially Departs from Professional Judgment Toward Foster Group Homes ...820

V. REMEDY ...822

VI. CONCLUSION ...828

VII. GLOSSARY ...828

I. BACKGROUND

A. Procedural History

Plaintiffs are minor children in the Permanent Management Conservatorship (“PMC”) of the Texas Department of Family and Protective Services (“DFPS”).¹ Plaintiffs filed suit through their next friends on March 29, 2011, seeking injunctive relief against Rick Perry, Governor of Texas; Thomas Suehs, Executive Commissioner of the Texas Health and Human Services Commission;² and Anne Heiligenstein, Commissioner of DFPS (collectively “Defendants”), in their official capacities.³ (D.E. 1). Shortly thereafter, Plaintiffs filed a Motion for Class Certification. The Court granted their motion, holding that the requirements of Fed. R. Civ. P. 23 had been met. (D.E. 49).

1 There is an acronym glossary at the end of the Opinion.

2 The Health and Human Services Commission oversees and administers the agencies in Texas's Health and Human Services system, which includes DFPS. See Tex. Gov't Code Ann. § 531.0055.

3 Greg Abbot has since replaced Perry as Governor of Texas, 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.

Defendants filed an interlocutory appeal of the class certification to the Fifth Circuit Court of Appeals pursuant to Fed. R. Civ. P. 23(f). (D.E. 63). While the appeal was pending, the Supreme Court decided Wal – Mart Stores, Inc. v. Dukes , 564 U.S. 338, 131 S.Ct. 2541, 180 L.Ed.2d 374 (2011). In light of Wal – Mart, the Fifth Circuit vacated the class certification order and remanded the case. M.D. ex rel. Stukenberg v. Perry (M.D. I ), 675 F.3d 832 (5th Cir.2012). Plaintiffs filed a second Motion for Class Certification in October 2012. (D.E. 160). After a three-day hearing in January 2013, the Court found that the requirements of Fed. R. Civ. P. 23(a), as explained in Wal – Mart, were satisfied. The Court certified a General Class and three subclasses on August 27, 2013. M.D. v. Perry (M.D. II ), 294 F.R.D. 7 (S.D.Tex.2013). The certified classes are 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; and

d. Basic Care General Residential Operation Subclass: all members of the General Class who are now or will be in a general residential operation and who are or will be receiving solely non-emergency, Basic childcare services.

Id. at 67. The Court denied certification of a fourth subclass for children in unverified kinship placements because it lacked adequate representation. Id. at 63. Defendants filed an untimely Petition for Permission to Appeal the Class Certification Order, which the Fifth Circuit dismissed on November 19, 2013. M.D. ex rel. Stukenberg v. Perry , 547 Fed.Appx. 543 (5th Cir.2013).

II. FINDINGS OF FACT AND LAW

Pursuant to Fed. R. Civ. P. 52(a), the Court makes the following findings of fact and conclusions of law. Any finding of fact that also constitutes a conclusion of law is adopted as a conclusion of law. Any conclusion of law that also constitutes a finding of fact is adopted as a finding of fact. All of the Court's findings of fact and conclusions of law are based upon a preponderance of the evidence.

A. Overview of Texas Foster Care

The Texas Department of Family and Protective Services is the agency responsible for protecting the State's children, elderly, and disabled. Two DFPS divisions are pertinent to this case: Child Protective Services (“CPS”) and Child Care Licensing (“CCL”). Both divisions work directly with families, children, and childcare providers to protect all of Texas's children, whether or not they are in foster care. (DX 31 at 6-7). CCL is made up of two divisions: the Residential Child Care Licensing (“RCCL”) division focuses on protecting children living in licensed foster care placements; the Performance Management Unit (“PMU”) provides quality assurance for all of DFPS. Id. at 163, 171. John J. Specia, Jr. (“Specia”) is the Commissioner and overall chief executive officer of DFPS. Id. at 4. He is DFPS's seventh Commissioner since 2004. An Assistant Commissioner heads each division: Lisa Black (“Black”) is Assistant Commissioner for CPS; Paul Morris (“Morris”) is Assistant Commissioner for CCL.⁴ Specia, Black, and Morris each acquired their position after this case began.

4 There is a chart of DFPS's officer hierarchy at the end of the Opinion. (See Attachment 1; PX 811).

When CPS determines that it is not safe for a child to live with her legal guardian, CPS petitions a court to remove the child and to obtain Temporary Managing Conservatorship (“TMC”). (DX 31 at 80). If TMC is granted, DFPS (through CPS) takes custody of the child, placing her in a temporary living arrangement with a certified caregiver or a family member. TMC lasts up to one year unless a court extends it another six months. Id. at 80. CPs's goal for each child is “permanency,” which is achieved when a child returns home after it is safe, moves in with a relative long-term, is adopted by a new family, or ages out of foster care at age 18.⁵ Id. at 73. Permanency is the “most important wellbeing issue” for foster children because it removes them from foster care and into stable environments where they better develop into successful adults. (D.E. 302 at 27, 38). As Commissioner Specia testified, “I want good foster care, but the answer is permanency.” (D.E. 331 at 48).

5 Extended foster care is available for youths over the age of 18 in limited circumstances. (DX 24 at 13).

If the child has not achieved permanency at the end of TMC, the child enters the State's Permanent Managing Conservatorship. (DX 31 at 80). There are approximately 17,000 children in TMC and 12,000 children in PMC at any given time. (Id. at 94; DX 119 at 72; DX 123). The change from TMC to PMC is significant. The act of designating children a “permanent” part of a foster care system is unique to Texas. (D.E. 299 at 60). Unlike TMC, PMC is considered a final order similar to reunification, adoption, or aging out. (DX 31 at 80).

In contrast with TMC children, PMC children do not have the same court deadlines or internal DFPS deadlines that their CVS caseworkers need to meet. (See D.E. 305 at 71). For example, Texas requires permanency review hearings once every four months for TMC children, but only once every six months for PMC children. Tex. Fam. Code Ann. §§ 263.305, 263.501. Likewise, Texas requires at least two permanency planning meetings and a status hearing for TMC children, but there is no such requirement for PMC children. See id. at § 263.001 et seq . Caseworkers also must review TMC children's service plans four times in the first year, while PMC children's service plans are only reviewed twice a year. (D.E. 311 at 12-13). CPS Regional Director Judy Bowman (“Bowman”) testified that a child's service plan is critically important and functions as a road map for children's path to permanency. Id. at 12. Also unlike TMC children, many PMC children do not have an attorney ad litem to set hearings and file pleadings with a court, or notify a court when the child needs assistance. (D.E. 299 at 60; D.E. 323 at 68). Most PMC children also do not have a Court Appointed Special Advocate (“CASA”), who are appointed by judges to watch over and advocate for foster children, even though a child's CASA “usually is the only person who truly knows the child and knows how the child is really doing.” (PX 1988 at 17).⁶ Thus, the State effectively deprives many PMC children of an individual advocate. Anna Ricker (“Ricker”), the attorney ad litem and next friend of J.S., H.V., and P.O., testified that her clients in PMC “get ignored more.” (D.E. 326 at 226). Karen Langsley (“Langsley”), another attorney ad litem who represents TMC and PMC children, testified that there is significantly less attention paid by caseworkers to children once they enter PMC. (D.E. 323 at 67-68). As seen in the record, PMC children tend to receive fewer visits from primary caseworkers, visits that are less meaningful and more rushed, and overall more cursory casework. As one report explained, “Though the State's responsibility for the child's life and well-being does not change—and arguably increases—the attention paid to the child's cases diminishes drastically. There is often a sense that the ‘clock stops ticking’ when the child enters Permanent Managing Conservatorship.” Id. at 15. Although all of the permanency options are available to PMC children, the State often “just maintain[s] them in foster care until they age [ ] out.” (D.E. 323 at 89-92).

6 CASA is a national volunteer organization with state and local member programs. “CASA volunteers are screened and highly trained and then appointed by judges to represent and advocate for a child's best interests in the child protection system. CASA volunteers are each assigned to help one child or set of siblings at a time, so they can focus on giving that child or sibling group the individualized advocacy and attention they need.” Texas CASA, http://texascasa.org/ (last visited December 11, 2015). CASA volunteers “make sure [foster children] don't get lost in the overburdened legal and social service system or languish in inappropriate group or foster homes. Volunteers stay with each case until it is closed and the child is placed in a safe, permanent home.” National CASA, www.casaforchildren.org/site/c.mtJSJ7MPIsE/b.5301303/k.6FB1/About_Us__CASA_ for_Children.htm (last visited December 11, 2015).

DFPS has four service levels for children in its care, depending on a child's physical and psychological needs: Basic, Moderate, Specialized, and Intense. (D.E. 328 at 91). The higher the service level, the more the State pays the foster care facility. (D.E. 322 at 90-92). Placements must be licensed to provide for children at specific service levels. Therefore, a child's service level dictates which placements are available to her. See id.

PMC children are placed in a variety of residential settings. Approximately 90% of these placements are managed by private child-placing agencies (“CPAs”) that contract with the State. (DX 119 at 72). DFPS directly manages the remaining 10%. In both cases, RCCL is ultimately responsible for inspecting, investigating, and licensing. (DX 31 at 163-64).

Foster family homes are traditional foster homes that contain one to six children. (DX 109 at 1273). The State verifies these facilities and provides training and financial support to the caregivers. The verification process involves screening and inspecting the home to ensure it meets the requirements for the type of care it will provide. Foster group homes—facilities unique to Texas—contain 7 to 12 children but are otherwise almost identical to foster family homes. (See id. ; DX 119 at 22). They are regulated and receive financial support, but, unlike other larger facilities, are not required to have awake-night supervision. Foster family and foster group homes can be “therapeutic,” meaning that the caregivers receive additional training to look after children with higher levels of care. Facilities that contain 13 or more children are called general residential operations (“GROs”). (DX 109 at 1273). They are subject to extensive regulations and receive financial support. GROs appear to have no capacity limit. One GRO in North Texas is authorized to house 437 children.⁷ Residential treatment centers (“RTCs”) are a type of GRO that provide therapeutic treatment “for children with serious emotional disturbances or mental health issues.” (PX 1864 at 101; see also DX 109 at 1273). Most of Texas's RTCs are in 3 of the State's 11 DFPS Regions, with a disproportionate number around Houston, and none south of San Antonio. (D.E. 327 at 203).⁸ GROs and RTCs are also called “congregate care facilities.” Foster family homes are the least restrictive, most family-like placement. Foster group homes, GROs, and RTCs are progressively more restrictive.

7 See Texas Department of Family and Protective Services, Cal Farley's Boys Ranch , https://www.dfps.state.tx.us/Child_Care/Search_Texas_Child_Care/ppFacilityDetails.asp?ptype=RC&fid=94127 (last visited November 3, 2015).

8 There is a map of DFPS's Regions at the end of the Opinion. (See Attachment 2; PX 806)

Texas also uses kinship placements,⁹ in which a child is placed with a relative or someone with a longstanding and significant relationship with the child or the child's family. Kinship placements, whether verified or unverified, must be approved by DFPS and have a home assessment. (DX 109 at 430, 1173). 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 unverified, they are eligible for limited monetary assistance and the caregivers are not required to complete the training provided for foster parents.

9 Although the Court did not certify the Unverified Kinship Subclass, these children are still in the General Class.

Texas's foster care system has a checkered history. In 2009, the Texas Legislature and Governor Perry formed the Texas Adoption Review Committee “to take a hard look at the Texas foster care system.” (PX 1964 at 2). The Committee conducted a ten-month review, which included testimony from DFPS employees, foster care advocates, policy analysts, foster and adoptive parents, CPAs, and experts from ten areas of DFPS. Id. at 8. The Committee reviewed articles, reports, and publications, including one by the Child Welfare League of America (“CWLA”), one by Specia (not yet DFPS Commissioner), and a report by Texas Appleseed that had previously been submitted to the Texas Supreme Court Commission for Children, Youth, and Families. Id. at 25–28. After drafting its recommendations, the 2009 Committee unearthed a report from 1996, produced by a similarly charged committee that was formed by Governor George W. Bush. Id. at 2, 7. The 2009 Committee found that 11 of its 14 general recommendations were made in 1996, leading it to conclude that “many of the same problems identified in 1996” had not been fixed. Id. at 2, 7–12. The Committee released its final report in December 2010. At the outset the report notes that, despite the good intentions and hard work of the people involved in Texas's child welfare system, “there is increasing evidence to show that our foster care system is sometimes doing more harm to our children than good.” Id. at 2.

B. Plaintiffs' Claims

Plaintiffs' claims have changed substantially over the course of this litigation. In their Original Complaint, Plaintiffs alleged a variety of problems with Texas's management of foster care and claimed that the State's mismanagement violated the rights of children in the State's PMC. (See D.E. 1). The Fifth Circuit vacated the first Class Certification Order, doubting whether Plaintiffs could “even advance a due process claim based on a bare finding that Texas has 'organized or managed' DFPS improperly.” M.D. I, 675 F.3d at 841 n. 3. In response, Plaintiffs narrowed and re-characterized their allegations, and proposed four subclasses. The Court certified one General Class and three subclasses. M.D. II , 294 F.R.D. 7 ; supra Section I.A. For each certified class, Plaintiffs allege that DFPS's “policies and practices result in structural deficiencies” that place the class members “at an unacceptable risk of harm” in violation of the Fourteenth Amendment. (D.E. 215 at 2, 17, 18, 19). The Court understands Plaintiffs' argument as saying that each policy and practice does not, on its own, have to result in a constitutional violation. Although some policies and practices may reach this threshold, Plaintiffs also argue that the State's combined policies and practices for each certified class create structural deficiencies, which violate class members' Fourteenth Amendment right to be free from an unreasonable risk of harm while in State custody. See Alberti v. Klevenhagen , 790 F.2d 1220, 1224 (5th Cir.1986) (“In determining the constitutional question, we need not separately weigh each of the challenged institutional practices and conditions, for we instead look to ‘the totality of conditions.’ ”) (citation omitted). Plaintiffs' specific claims break down as follows:

On behalf of the General Class:

(1) DFPS does not employ enough primary conservatorship caseworkers, causing those workers to have excessive caseloads, which prevents them from properly fulfilling their required duties, protecting the children in the State's custody.

(2) Caseworkers' excessive caseloads increase caseworker turnover, further worsening caseloads and initiating a vicious cycle.

(3) By not having the time and resources to adequately monitor the children assigned to them, caseworkers cannot ensure that Plaintiffs are free from an unreasonable risk of harm while in the State's care.

(4) These practices and policies, separately and combined, cause an unreasonable risk of harm to all General Class members.

(5) The State's practices and policies substantially depart from professional judgment.

(6) The State is aware of these risks, yet refuses to assess or address them.

On behalf of the Licensed Foster Care Subclass:

(1) The State does not exercise sufficient oversight of the facilities in which foster children are placed by failing to properly investigate and inspect those facilities and hold them accountable for licensing violations.

(2) Defendants do not properly track incidents of child-on-child abuse.

(3) The State maintains an insufficient number, geographic distribution, and array of placements, such that foster children cannot be placed in homes or facilities appropriate for their needs, in their home communities.

(4) These practices and policies, separately and combined, cause an unreasonable risk of harm to all Licensed Foster Care Subclass members.

(5) The State's practices and policies substantially depart from professional judgment.

(6) The State is aware of these risks, yet refuses to assess or address them.

On behalf of the Foster Group Home Subclass:

(1) The State's foster group homes are deficient:

(a) Group home caregivers are not properly trained or qualified;

(b) Group homes lack sufficient professional staff;

(c) Group homes do not require 24-hour awake-night supervision like other congregate care facilities; and

(d) The State does not restrict the placement of unrelated children of different genders, ages, and service levels in the same group home.

(2) These practices and policies, separately and combined, cause an unreasonable risk of harm to all Foster Group Home Subclass members.

(3) The State's practices and policies substantially depart from professional judgment.

(4) The State is aware of these risks, yet refuses to assess or address them.

On behalf of the Basic Care GRO Subclass:

(1) The State places children who need only Basic-level childcare services in GROs (and sometimes RTCs) due to a lack of appropriate placements.

(2) Placing Basic-level children in GROs causes neglect, maltreatment, and institutionalizes otherwise healthy children.

(3) This practice causes an unreasonable risk of harm to all Basic Care GRO Subclass members.

(4) The State's practice substantially departs from professional judgment.

(5) The State is aware of this risk, yet refuses to assess or address it.

C. Constitutional Rights

The State generally does not have an obligation to protect citizens from private harm. DeShaney v. Winnebago Dep't of Soc. Servs. , 489 U.S. 189, 195–96, 109 S.Ct. 998, 103 L.Ed.2d 249 (1989). However, “in certain limited circumstances the Constitution imposes upon the State affirmative duties of care and protection.” Id. at 198–200, 109 S.Ct. 998. These duties arise when the State takes a person into custody, thereby limiting that person's freedom to act on her own behalf. Id. at 200, 109 S.Ct. 998. Custody, in other words, creates a “special relationship” between the State and that person, which triggers a constitutional duty to provide basic needs. Id.

The State's affirmative duty of care and protection for people in custody was first identified for prisoners. Estelle v. Gamble , 429 U.S. 97, 97 S.Ct. 285, 50 L.Ed.2d 251 (1976), held that the Eighth Amendment's prohibition of cruel and unusual punishment obliged the State to provide medical care for prisoners. By depriving inmates of the ability to care for their own medical needs, the State assumed that affirmative duty. Id. at 103–04, 97 S.Ct. 285. These rights were extended into the Fourteenth Amendment in the context of patients involuntarily committed to state mental institutions. In Youngberg v. Romeo , 457 U.S. 307, 102 S.Ct. 2452, 73 L.Ed.2d 28 (1982), the Supreme Court reasoned, “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, 102 S.Ct. 2452. The Supreme Court held that patients in these settings possessed the right to “food, shelter, clothing, medical care,” and safe living conditions, which the State must provide. Id. Likewise, pretrial detainees who are injured while in police custody have the constitutional right to medical care. City of Revere v. Mass. Gen. Hosp. , 463 U.S. 239, 244, 103 S.Ct. 2979, 77 L.Ed.2d 605 (1983). Although the Eighth Amendment does not apply to pretrial detainees because they have not been convicted of a crime, the Fourteenth Amendment provides the affirmative duty of care. See Youngberg , 457 U.S. at 315–16, 102 S.Ct. 2452. These cases stand for the straightforward proposition that when the State restrains an “individual's freedom to act on his own behalf—through incarceration, institutionalization, or other similar restraint of personal liberty,” the State assumes an affirmative duty to provide basic needs. DeShaney , 489 U.S. at 200, 109 S.Ct. 998.

State custody of a child creates a “special relationship” that triggers substantive due process protections. The Fifth Circuit first recognized this special relationship, and the State's corresponding duty to provide “constitutionally adequate care,” in Griffith v. Johnston , 899 F.2d 1427, 1439 (5th Cir.1990). More recently and more explicitly, the Fifth Circuit held that, under the Fourteenth Amendment, the State owes its foster children “personal security and reasonably safe living conditions.” Hernandez v. Tex. Dep't of Protective & Regulatory Servs. , 380 F.3d 872, 880 (5th Cir.2004) ; 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). Put another way, foster children have the right to be free from an unreasonable risk of harm.

The Fifth Circuit is not alone. With near unanimity, the other circuits have found that states owe a duty of care to their foster children. See e.g., Tamas v. Dep't of Social & Health Servs. , 630 F.3d 833, 846–47 (9th Cir.2010) ; James ex rel. James v. Friend , 458 F.3d 726, 730 (8th Cir.2006) ; Nicini v. Morra , 212 F.3d 798, 808 (3d Cir.2000) ; Lintz v. Skipski , 25 F.3d 304, 305 (6th Cir.1994) ; Yvonne L. ex rel. Lewis v. N.M Dep't of Human Servs. , 959 F.2d 883, 892 (10th Cir.1992) ; K.H. ex rel. Murphy v. Morgan , 914 F.2d 846, 848–49 (7th Cir.1990) ; Taylor ex rel. Walker v. Ledbetter , 818 F.2d 791, 795 (11th Cir.1987) ; Doe v. N.Y. Dep't of Soc. Servs. , 649 F.2d 134, 144–45 (2d Cir.1981) ; Connor B. ex rel. Vigurs v. Patrick (Connor B. I ), 771 F.Supp.2d 142, 160 (D.Mass.2011) (“Courts have found, with apparent unanimity, that [a special] relationship exists in the foster-care context.”); see also DeShaney , 489 U.S. at 201 n. 9, 109 S.Ct. 998. Most circuits consider this right “clearly established” for qualified immunity purposes. See Tamas , 630 F.3d at 846–47 (surveying the circuit courts and finding that, by 1996, foster children had a “clearly established” liberty interest in safe foster care placements).

A foster child's right to be free from an unreasonable risk of harm “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) ; see also LaShawn A. by Moore v. Kelly , 990 F.2d 1319 (D.C.Cir.1993). Moreover, foster children “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) ; accord K.H. ex rel. Murphy , 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.”). Harm, both psychological and physical, can be inflicted in a variety of ways, including neglect, physical abuse, sexual abuse, and psychological maltreatment. “[T]he Constitution requires the responsible state officials to take steps to prevent children in state institutions from deteriorating physically or psychologically.” K.H. ex rel. Murphy , 914 F.2d at 851 (citing Youngberg , 457 U.S. 307, 102 S.Ct. 2452 ).

Foster children need not wait until actual harm occurs before obtaining relief. This has been made most explicit in the prison context, where the right is 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). Unreasonable threats of injury to prisoners, such as poor fire safety conditions, are themselves violations of a prisoner's constitutional rights. Id. at 784 ; see also Battle v. Anderson , 564 F.2d 388, 395 (10th Cir.1977). As such, prisoners can obtain relief from unreasonable conditions even when no fire has occurred. The risk of harm is the legal injury. In the foster care context, a structural deficiency that puts foster children at an unreasonable risk of harm is the legal injury. See Hernandez , 380 F.3d at 881 (explaining that “the risk of severe physical abuse to a foster child's bodily integrity” is the legal injury).

Defendants acknowledge foster children's constitutional right to personal security and reasonably safe living conditions. Defendants argue, however, that foster children do not possess an “unlimited” right “to be free from an unreasonable risk of harm,” (D.E. 277 at 29; D.E. 359 at 47), and that the Court should “resist the temptation to augment the substantive reach of the Fourteenth Amendment.” (D.E. 163 at 22 (quoting Griffith , 899 F.2d at 1435 )). The Court disagrees. There is no difference between harm in general and harm to personal security and reasonably safe living conditions. All harms affect either a foster child's person or environment, and the right to be free from an unreasonable risk of these harms is unlimited. The Court holds that foster children have a Fourteenth Amendment substantive due process right to be free from an unreasonable risk of harm caused by the State.

Plaintiffs also claim that foster children have a Fourteenth Amendment right to the “most appropriate care, treatment, and services.” (D.E. 215 at 19). Plaintiffs have aimed too high, at least by substantive due process standards. Foster children have the right to “minimally adequate” care, treatment, and services such that it prevents an unreasonable risk of harm. Youngberg , 457 U.S. at 321, 102 S.Ct. 2452 ; Hernandez , 380 F.3d at 880. They do not have a constitutional right to the most appropriate care, treatment, and services.

To succeed on their substantive due process claim, Plaintiffs must show that: (1) the State has acted in a way that breached the duty of care it owes to foster children in its custody, namely the duty to keep them free from an unreasonable risk of harm; and (2) the State action that did so rises to the requisite level of culpability. In making the latter determination in challenges to executive action, the State will be liable if its alleged action “shocks the conscience.” Cnty. of Sacramento v. Lewis , 523 U.S. 833, 846, 118 S.Ct. 1708, 140 L.Ed.2d 1043 (1998) ; Rochin v. California , 342 U.S. 165, 172, 72 S.Ct. 205, 96 L.Ed. 183 (1952) ; Hernandez , 380 F.3d at 880. Determining the level of culpability that “shocks the conscience” in the foster care context, however, has proven elusive. Courts have articulated culpability standards to further define the test, two of which—deliberate indifference and substantial departure from professional judgment—the parties dispute. (D.E. 195 at 48-49).

In Estelle, the Supreme Court held that “deliberate indifference” to a prisoner's serious medical needs states a cause of action under the Eighth Amendment. 429 U.S. at 106, 97 S.Ct. 285. Deliberate indifference is determined by a subjective standard of recklessness: “the official must be both aware of facts from which the inference could be drawn that a substantial risk of serious harm exists, and he must also draw the inference.” Smith v. Brenoettsy , 158 F.3d 908, 912 (5th Cir.1998) (quoting Farmer v. Brennan , 511 U.S. 825, 837, 114 S.Ct. 1970, 128 L.Ed.2d 811 (1994) ). The State must have consciously disregarded a known and excessive risk to the victim's health and safety. Farmer , 511 U.S. at 837, 114 S.Ct. 1970.

Six years later, the Supreme Court refused to apply the deliberate indifference standard when considering the rights of involuntarily committed patients at mental facilities. Youngberg , 457 U.S. at 321–22, 102 S.Ct. 2452. Under the Fourteenth Amendment, as opposed to the Eighth Amendment at issue in Estelle, the rights of persons involuntarily committed to mental facilities are violated if the State substantially departs from professional judgment. Id. at 323, 102 S.Ct. 2452. According to the professional judgment standard, 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.” Id. The main difference between the two standards is that the professional judgment standard does not require the State to know that the person in its custody will be harmed. Youngberg justified the less demanding standard by stating, “Persons who have been involuntarily committed are entitled to more considerate treatment and conditions of confinement than criminals whose conditions of confinement are designed to punish.” Id. at 321–22, 102 S.Ct. 2452 (comparing to Estelle , 429 U.S. at 104, 97 S.Ct. 285 ).

The Supreme Court discussed, but did not decide, the contours of foster children's substantive due process rights in DeShaney. The DeShaney Court explained that it is the deprivation of a person's liberty, not what motivates the deprivation, that triggers the Constitution's protection. 489 U.S. at 200, 109 S.Ct. 998. This seems to signal a uniform liability standard for any person in state custody, but DeShaney never reached this question. Because the child in DeShaney was harmed in the custody of private actors, the Court found there was no special relationship between the child and the State, and therefore did not decide between the deliberate indifference and professional judgment standard.

The Supreme Court's most recent comment in this area comes from County of Sacramento v. Lewis . 523 U.S. 833, 118 S.Ct. 1708. Lewis made clear that both the deliberate indifference and professional judgment standards remain good law, and are simply more precise articulations of the “shocks the conscience” standard. Id. at 850, 118 S.Ct. 1708. According to Lewis, “Rules of due process are not ... subject to mechanical application in unfamiliar territory.” Thus, “Deliberate indifference that shocks in one environment may not be so patently egregious in another.” Id. Youngberg “can be categorized on much the same terms.” Id. at 852 n. 12, 118 S.Ct. 1708.

In the Fifth Circuit, a foster care plaintiff must prove deliberate indifference to have a substantive due process cause of action. The Fifth Circuit has emphasized that the deliberate indifference test is a “significantly high burden for plaintiffs to overcome.” Hernandez , 380 F.3d at 882 (citing Doe v. Dallas Indep. Sch. Dist. , 153 F.3d 211, 218 (5th Cir.1998) ). The test demands “a degree of culpability beyond mere negligence or even gross negligence; it must amount to an intentional choice, not merely an unintentional oversight.” Hall v. Smith , 497 Fed.Appx. 366, 377 (5th Cir.2012). Liability based on deliberate indifference is inappropriate if an official can demonstrate “that [she] did not know of the underlying facts indicating a sufficiently substantial danger and that [she was] therefore unaware of a danger, or that [she] knew the underlying facts but believed (albeit unsoundly) that the risk to which the facts gave rise was insubstantial or nonexistent.” Farmer , 511 U.S. at 844, 114 S.Ct. 1970. Moreover, an official who is aware of a substantial risk of serious harm is not deliberately indifferent if she “responded reasonably to the risk, even if the harm ultimately was not averted.” Id.

Yet the “high burden” is not insurmountable. To prove deliberate indifference, “it is enough that the state official acted or failed to act despite his knowledge of a substantial risk of serious harm.” Hernandez , 380 F.3d at 881 (citing Farmer , 511 U.S. at 842, 114 S.Ct. 1970 ). “[A]n obvious showing that the state ... exhibited a conscious disregard for known severe physical abuses in a state-licensed foster home by itself sufficiently demonstrates deliberate indifference to a child's right to personal security.” Hernandez , 380 F.3d at 881. Actual knowledge of a substantial risk of harm can be inferred if “the risk of harm is obvious.” Id. (citing Hope v. Pelzer , 536 U.S. 730, 738, 122 S.Ct. 2508, 153 L.Ed.2d 666 (2002) ).

Most circuits apply the deliberate indifference standard to foster care cases. See, e.g. , Tamas , 630 F.3d at 844 (9th Cir.2010) ; James , 458 F.3d at 730 (8th Cir.2006) ; J.H. ex. rel Higgin v. Johnson , 346 F.3d 788, 792 (7th Cir.2003) ; Nicini , 212 F.3d at 808 (3d Cir.2000) ; Lintz , 25 F.3d at 305 (6th Cir.1994) ; Taylor v. Ledbetter , 818 F.2d at 795 (11th Cir.1987) ; Doe , 649 F.2d at 144-45 (2d Cir.1981). In this case, Defendants argue that deliberate indifference applies.

Plaintiffs, however, point out that some circuits, relying on Youngberg, use the professional judgment standard in foster care cases. After all, foster children bear a closer resemblance to the patient in Youngberg than the prisoner in Estelle : their rights derive from the Fourteenth Amendment rather than the Eighth, and they are not in the State's custody to be punished. The Tenth Circuit, for example, recently used the professional judgment standard in a foster care case. Schwartz v. Booker , 702 F.3d 573, 580 (10th Cir.2012) ; see also Yvonne L. , 959 F.2d at 893–94. Although the First Circuit has not decided the issue, two of its district courts recently used the professional judgment standard. See Cassie M. ex. rel. Irons v. Chafee , 16 F.Supp.3d 33, 43–44 (D.R.I.2014), vacated on other grounds sub nom. Danny B. ex rel. Elliott v. Raimondo , 784 F.3d 825 (1st Cir.2015) ; Connor B. I , 771 F.Supp.2d at 163.

Plaintiffs also note that courts have decided which standard to apply based on whether the remedy requested is monetary damages or injunctive relief. When injunctive relief is sought, as it is here, some courts use the professional judgment standard. See, e.g. , LaShawn A. v. Dixon , 762 F.Supp. 959, 996 n. 29 (D.D.C.1991) ; Cassie M. , 16 F.Supp.3d at 47 (D.R.I.2014). Deliberate indifference applies for monetary damages, the reasoning goes, because monetary damages may have a chilling effect on state policymakers. LaShawn A. , 762 F.Supp. at 996 n. 29. Courts are less concerned with a chilling effect when Plaintiffs seek only injunctive relief, and therefore apply the less stringent professional judgment standard. Id.

All that said, when deciding whether state action shocks the conscience in the foster care context, cases often come out the same regardless of which standard is applied. See, e.g. , Yvonne L. , 959 F.2d at 894 (“As applied to a foster care setting we doubt there is much difference in the two standards.”); Connor B. I , 771 F.Supp.2d at 162 n. 4 (“It is far from obvious, however, that the professional judgment standard creates an appreciably lower hurdle for plaintiffs in this foster care case ... in this context the decision seems to matter little.”); LaShawn A. , 762 F.Supp. at 996 n. 30 (“The Court notes that although it has applied the professional judgment standard to this case, the result would have been the same had it used the deliberate indifference standard.”).

Again, the main difference between the standards is that the professional judgment standard does not require the State to have actual knowledge that foster children will be harmed. This difference, however, is dulled by Hernandez's statement that actual knowledge can be inferred when “the risk of harm is obvious.” 380 F.3d at 881 (citing Hope , 536 U.S. at 738, 122 S.Ct. 2508 ). Yet, in light of the uncertainty, the Court analyzed each claim under both standards. The Court holds that, judged by either standard, Texas's conduct shocks the conscience.

D. Evidence Summary

The Court carefully reviewed the Named Plaintiffs' case files, the testimony of 28 fact witnesses and 12 expert witnesses (as well as their expert reports, with a few exceptions), more than 400 exhibits (totaling over 390,000 pages), the parties' briefs, and the relevant caselaw. For clarity, the Court outlined the more frequently cited reports and national standards, introduced the witnesses, and summarized the Named Plaintiffs' narratives.

1. Reports

Casey Family Programs . Casey Family Programs is the nation's largest operating foundation focused on safely reducing the need for foster care. Its mission is to “provide and improve—and ultimately prevent the need for—foster care.”¹⁰ DFPS has contracted with Casey Family Programs on multiple occasions to provide internal reviews and assessments, as well as staff training and professional development. (See, e.g. , DX 61).

10 Casey Family Programs, Who We Are, http://www.casey.org/about/ (last visited December 15, 2015).

Texas Adoption Review Committee . The Texas Adoption Review Committee was created by the Texas Legislature and Governor Perry in 2009 “to take a hard look at the Texas foster care system and to uncover barriers to adoption that exist for Texas' most vulnerable children.” (PX 1964 at 2). After a ten-month exhaustive review, the Committee released a final report in December 2010 providing recommendations for how DFPS can improve. Prior to releasing its report, the Committee found a report from 1996, produced by a similarly charged committee that was formed by Governor Bush. The 2009 Committee found that 11 of its 14 general recommendations were made in 1996, leading it to conclude that “many of the same problems identified in 1996” had not been fixed. (Id. at 2, 7–12; see also supra pp. 10–11).

Texas Appleseed . Texas Appleseed is a nonprofit organization whose “mission is to promote justice for all Texans by using the volunteer skills of lawyers and other professionals to find practical solutions to broad-based problems facing the most vulnerable—including the State's foster children.” (PX 1988 at 4). The Court reviewed Texas Appleseed's 2007 and 2010 reports on Texas foster care. (PX 1966; PX 1988). The 2010 report was commissioned by the Supreme Court of Texas Permanent Judicial Commission for Children, Youth and Families, of which Specia and Defendant's expert Dr. Jane Burstain (“Burstain”) were members before joining DFPS. (PX 1988 at 5).

Texas Comptroller Report, Forgotten Children . In April 2004, Texas Comptroller Carol Keeton Strayhorn published a report titled: Forgotten Children, A Special Report on the Texas Foster Care System . The 306-page report detailed a number of shortcomings in Texas's foster care system. Although Forgotten Children is not itself in the record, it was frequently cited and relied on by reports that are in the record. (See, e.g. , PX 1966 at 5-6).

Texas Sunset Advisory Commission . The Texas Sunset Advisory Commission is a 12-member group created by the Texas Legislature to report on the conditions in Texas foster care and find ways to eliminate waste, duplication, and inefficiency. DFPS was mandated to provide unfettered access to the Sunset Commission, including allowing the Sunset Commission to survey all 11,000 DFPS employees, participate in CPS investigations, tour foster care facilities, meet with former foster children, and attend internal DFPS meetings. (PX 1861 at 131). The Sunset Commission issued its report in May 2014.

The Stephen Group . The Stephen Group is a highly regarded national business and government consulting firm. The Group has performed multiple child welfare consulting jobs across the country. Specia contracted with The Stephen Group in February 2014 to conduct an operational review of DFPS and CPS to figure out why so many problems remain despite “all the money and [effort] that have been put into the agency since 2006.” (D.E. 300 at 25; D.E. 331 at 45-46; PX 1993 at 9). Specia voluntarily provided The Stephen Group the same unfettered access that he was required to provide the Sunset Commission.

2. National Standards

Child Welfare League of America . The CWLA is “the nation's oldest and largest membership-based child welfare organization.” (PX 2114 at 3). The CWLA's professional standards are universally known and respected in the child welfare community. The Texas Legislature mentions the CWLA's standards in its guidelines for DFPS.

Council on Accreditation (“COA”) . The COA is another respected national nonprofit organization that recommends professional standards for state child welfare systems. The COA's standards are intended to guide foster care practices. States that meet COA standards are accredited. Only six states, not including Texas, have received COA accreditation.

Plaintiffs frequently refer to the CWLA and COA standards to show that DFPS substantially departs from professional judgment. DFPS argues that a deviation from these standards of excellence does not prove a “substantial departure” from professional judgment. DFPS is correct, in part.

A failure to meet CWLA and COA standards is not a per se constitutional violation. Professional standards, however, can be evidence for or against a constitutional violation. A “significant deviation” from professional standards could itself “constitute a substantial departure from accepted professional judgment that, under certain circumstances, may give rise to a substantive due process claim ....” Cassie M. , 16 F.Supp.3d at 48. A substantial departure from all relevant professional standards can also indicate that a professional was subjectively aware of, and opted to disregard, a substantial risk of serious harm, i.e., that the professional was deliberately indifferent. It is beyond question that DFPS is aware of CWLA standards. The Texas Legislature refers to the CWLA in discussing recommended caseloads for conservatorship workers: “Professional caseload standards ... are established or are recommended for establishment for employees of health and human services agencies by management studies conducted for health and human services agencies or by an authority or association, including the Child Welfare League of America, the National Eligibility Workers Association, the National Association of Social Workers, and associations of state health and human services agencies.” Tex. Gov't Code Ann. § 531.001(5). Furthermore, as of 2010, Texas (along with 37 other states) was a member of the CWLA. Although Texas did not need to recognize or adopt CWLA standards to become a member, Texas did have to pay dues, which gave it access to resources published or provided by the CWLA. (D.E. 302 at 47; D.E. 325 at 97).

Many courts have allowed experts to draw on CWLA and COA standards when analyzing foster care systems. Courts generally find that while neither standard imposes legal obligations on child welfare agencies, both are “reflective of the bar to which child welfare agencies are generally expected to measure up.” Connor B. ex rel. Vigurs v. Patrick (Connor B. II) , 985 F.Supp.2d 129, 136, 138, 151 (D.Mass.2013) ; see also LaShawn A. , 762 F.Supp. at 964, 966 (considering CWLA standards a relevant professional standard for assessing the District of Columbia child welfare agency); Kenny A. v. Perdue , 2004 WL 5503780, at *12 (N.D.Ga. Dec. 13, 2004) (finding that experts adequately relied on “accepted professional standards,” including those from CWLA and COA). Courts have found these standards relevant because they are “national in scope” and widely followed by child welfare agencies. Doe ex rel. G.S. v. Johnson , 52 F.3d 1448, 1454, 1462 (7th Cir.1995).

The Court finds that CWLA and COA standards are relevant and admissible to help the Court determine whether DFPS's challenged policies and practices constitute substantial departures from professional judgment. The Court also finds Texas's knowledge of these standards relevant to whether the State is deliberately indifferent toward its polices and practices.

3. Child and Family Service Reviews

Texas relies heavily on its performance on the Child and Family Service Review (“Review”). The U.S. Department of Health and Human Services implemented the Review in 2001 to determine which states' child and family welfare systems met the requirements of Titles IV-B and IV-E of the Social Security Act. (DX 143 at 1). The current Review (the third since inception) assesses states' performance on seven indicators: (1) Maltreatment in Foster Care; (2) Recurrence of Maltreatment; (3) Permanency in 12 Months for Children Entering Foster Care; (4) Permanency in 12 Months for Children in Foster Care 12 to 23 Months; (5) Permanency in 12 Months for Children in Foster Care 24 Months or More; (6) Re-Entry to Foster Care in 12 Months; and (7) Placement Stability. Id. The first two indicators relate to children's safety; the last five relate to children's permanency. On Round 3 of the Review, Texas met the federal standard on six of seven indicators, falling short on number 5, Permanency in 12 Months for Children in Foster Care for 24 Months or Longer. (DX 147 at 1-18). The State argues that its compliance with these standards “contradicts” Plaintiffs' claims of conscience-shocking deliberate indifference and evinces the “exercise, not total abdication, of professional judgment.” (D.E. 277, attachment 8, at 5). Defendants' arguments are inapposite.

Plaintiffs' claims only relate to children in Texas's PMC. Of the seven indicators, numbers 3 and 6 do not incorporate any data for PMC children, focusing exclusively on children who have been in foster care less than 12 months. Numbers 1, 2, 4, and 7 use data for both PMC and TMC children, muddling the ability to draw meaningful conclusions about only PMC children. As Defendant's expert Burstain testified, “To really understand what's happening, you have to look at sub-populations” and cannot rely on overall averages. (D.E. 330 at 191). The only indicator that exclusively accounts for Texas's PMC children is the one that Texas failed. (D.E. 277, attachment 8, at 5; DX 147 at 1, 9-10). Yet even if Texas had met the national standard for Permanency in 12 Months for Children in Foster Care for 24 Months or Longer, that does not decide this case. Whether Texas meets the federal permanency standard as it relates to the Social Security Act—which it does not—does not answer whether Texas's PMC foster children are placed at an unreasonable risk of harm due to either deliberate indifference or a substantial departure from professional judgment.

Additionally, Texas's Review performance is based on data it submitted. (DX 143 at 1). As discussed infra , DFPS's investigations of foster child maltreatment are woefully deficient and often inaccurate, making the State's self-reported data unreliable. See Section IV.B.1. Even when DFPS discovered deficiencies in a large subset of its investigations, it did not update its submissions to the Review. Furthermore, DFPS does not even track certain abuse, such as child-on-child abuse, which is included in the Review. (DX 143 at 2). Texas's Review performance is therefore provides little, if any, reliable evidence for this case.

4. Witnesses

a. Fact Witnesses

The Court heard from 28 fact witnesses, including six next friends and attorney ad litems of Named Plaintiffs, a leader of a nonprofit that works with former foster children, five former foster children, two former DFPS caseworkers, and 14 current DFPS officers.

Next Friends and Attorney Ad Litems. Karen Langsley is the next friend and attorney ad litem for former Named Plaintiff D.P. Langsley is a solo practitioner who primarily practices in child welfare and family law and regularly represents PMC children and parents in DFPS cases. She has been licensed in Texas since 2004 and currently practices in Austin and the surrounding counties. She was one of the first Texas lawyers certified as a Child Welfare Law Specialist by the American Bar Association, through the National Association of Counsel for Children. Langsley also sits on the State Bar Committee on Child Abuse and Neglect and annually teaches a day-long continuing legal education course on child welfare. That course is part of a four-day conference on Advanced Family Law, which is the largest seminar of its type in the country. (D.E. 323 at 60-67). Anna Ricker is the next friend and attorney ad litem for J.S., H.V., and P.O. Ricker has served as an attorney ad litem for 17 years, representing around 150 children in Texas's PMC. She currently has between 10 and 12 PMC children on her caseload. Ricker also represented DFPS for four years as an Assistant District Attorney. She currently practices in West Texas, primarily in Levelland, Littlefield, and the surrounding counties. She often interacts with her clients' conservatorship caseworkers. (D.E. 326 at 197-98; D.E. 187 at 89). Javier Solis (“Solis”) is the next friend for S.A. Solis previously worked in the Cameron County District Attorney's Office prosecuting CPS cases. Since June 2001, Solis has served as an attorney ad litem for over 900 children in DFPS's custody. He currently practices in Cameron, Hidalgo, and Willacy Counties, representing 55 to 60 children in any given month. In his capacity as an attorney ad litem , Solis interacts with conservatorship caseworkers on a daily basis. (D.E. 323 at 6-7). Sarah Stukenberg (“Stukenberg”) is the next friend and attorney ad litem for M.D. She has practiced family law for over seven years in Texas, mainly representing foster children in Texas's TMC and PMC, although she sometimes represents parents in CPS cases as well. (D.E. 324 at 211-12). Jennifer Talley (“Talley”) is the next friend for A.M. She has worked in child welfare for over 22 years. Talley worked for one year in a psychiatric hospital, primarily with foster youth, and worked for CPS for 12 years. At CPS, Talley first worked as a conservatorship caseworker with both TMC and PMC children, and then for seven years as a Preparation for Adult Living coordinator for Region 8. She has volunteered at a camp for older youths in foster care for over 24 years. Later, she worked for six and a half years with youths aging out of care through a private foundation and another two years with a CPA, focusing on adolescent adoptions. Her work with the Preparation for Adult Living program focused on ensuring that foster children aging out of care received independent living skills and other necessary post-foster care guidance and support. Through most of her time as PAL coordinator Talley was the sole full-time Region 8 PAL staff member, with only a half-time administrative assistant. The majority of the 800 children she worked with in this program were in Texas's PMC. (D.E. 323 at 84-88). Estella Vasquez (“Vasquez”) is the next friend and attorney ad litem for L.H. and C.H. She has practiced family law and criminal law since 2008. Vasquez regularly serves as an attorney ad litem in Cameron County for children and parents in CPS cases. She represents over 100 children per year in such cases, and has represented over 600 children in Texas's TMC and PMC as an attorney ad litem . (D.E. 327 at 187-95). Vasquez is also the elected city commissioner for the City of Brownsville.

Nonprofit Leader . Sandra Carpenter (“Carpenter”) runs Angel Reach, a nonprofit in Conroe, Texas that helps children who age out of foster care. Carpenter initially started Angel Reach to help children and their families in kinship placements. Although Angel Reach still works with kinship placement families, as of 2011 it switched its focus to transitional living programs for youths aging out of foster care, due to the large number of homeless former foster youths living in the Conroe area. Angel Reach provides these aged-out youths with independent living skills and tools. Id. at 11. Angel Reach houses up to 29 former foster youths and provides nonresidential services such as transportation, mentoring, education, and job placement for up to 20 others. Id. Since 2011, Carpenter has worked with approximately 180 former foster youths. Id. at 13. She was a foster parent for 65 children over 16 years. Id. at 6. (D.E. 307 at 4-13)

Former Foster Children . Jordan Arce (“Arce”), Crystal Bentley (“Bentley”), Darryl Jackson (“Jackson”), Patricia Virgil (“Virgil”), and Kristopher Sharp (“Sharp”) are former foster youths who aged out of care. Arce, 19 years old at trial, entered foster care at age 14 and aged out at 18. (D.E. 324 at 48). Arce had six placements, including a children's shelter, a foster home, a GRO, and kinship placements. Id. at 49, 51–52. His level of care was Basic throughout his time in the State's custody. Id. at 53–54; see id. at 28–29. Today he is a student at Texas Tech University. Id. at 49. Bentley entered foster care at age 2 and aged out at 18. She was 23 years old at trial. Id. at 61. Bentley is a client of Angel Reach. Id. at 61–62. Bentley is currently a sophomore at Lone Star College and a client of Angel Reach. Id. at 61–62. Jackson, 18 years old at trial, entered Texas foster care at age 12 and aged out at 18. Id. at 183. While in Texas foster care, Jackson had between 35 and 40 placements, including foster family homes, foster group homes, and RTCs. Id. at 184, 187–88. Jackson is a client of Angel Reach. Id. at 183. When he aged out of care, Jackson was referred to Angel Reach by his caseworker, but otherwise received no support or preparation for independent living. Id. at 185. Virgil, 25 years old at trial, entered foster care at age 12 and aged out at 19. Id. at 198. She was placed in several foster family homes, foster group homes, and emergency shelters. Id. at 198, 201, 207. She too received no preparation or support when she aged out of care. Id. at 208–09. Sharp, 24 years old at trial, entered care at age 10 and aged out shortly before turning 18. (D.E. 325 at 163-64). He had around 25 placements, including foster group homes, GROs, and RTCs. Id. at 164–65. Sharp is currently studying social work at the University of Houston Downtown and working with advocacy groups that serve former foster youths who have aged out. Id. at 170, 176. He similarly received no preparation or support before he aged out of the State's care. Id. at 178–80.

Former DFPS Caseworkers. Beth Miller (“B. Miller”) worked for four and a half years as a CPS conservatorship caseworker, first in San Antonio and then in Abilene. (D.E. 323 at 32). Katrina Voelkel worked for two years as a CPS conservatorship caseworker in the PMC unit in Lubbock. (D.E. 324 at 8-9).

Current DFPS Officers . Colleen McCall (“McCall”) is the Director of Field Operations at DFPS. (D.E. 305 at 5). She has been at CPS for several decades but has been in her current position since 2005. Id. at 9. McCall's responsibilities include coordinating with and overseeing CPS regional directors and bringing the concerns of CPS field staff to DFPS leadership. Id. at 10. Caseworkers ultimately report to McCall. Id. She is also tasked with ensuring that primary conservatorship caseworkers have manageable caseloads. Id. at 14, 17–18.

McCall reports to the Assistant Commissioner of CPS, Lisa Black . (See Attachment 1; D.E. 300 at 5; D.E. 305 at 8-9). Black is CPs's top executive and reports directly to Specia. (D.E. 300 at 5). Black has been at CPS for approximately 28 years, but has been in her current position only since February 1, 2014. Id. at 6. Black spent six years as the regional director of Region 3 (during which time she reported to McCall). (D.E. 300 at 6; D.E. 305 at 9). Black oversees CPs's policies, practices, and overall operation, and has ultimate responsibility for conservatorship caseworkers and supervisors. Id. at 6–7. The Court was troubled by Black's testimony. At trial, Black admitted she had not read multiple high-level reports about problems at CPS, including the 2010 Texas Adoption Review Committee report that discussed CPS caseworkers' high caseloads and turnover rates.¹¹ (See, e.g. , id. at 33–35, 102; D.E. 323 at 141-42). She also could not testify about the rate of abuse and neglect in foster care placements, an area for which she is responsible. (D.E. 300 at 45-48). Black made no effort to acquire this information before trial despite being questioned about it at her deposition two months prior. Id. at 47–48. Black is also involved in Foster Care Redesign, the State's effort to improve its placement array. Id. at 80. As of trial, the State had data on only one of the two contracts it had entered under Redesign. Black did not know basic information about this contract. Id. at 84–85.

11 That report was issued while Black was a Regional Director and presumably should have been concerned about problems with CPS caseloads and turnover, at least as they affected caseworkers and supervisors in her Region.

Paul Morris , who also reports directly to Specia, is the Assistant Commissioner of CCL, the top executive of that division. (D.E. 301 at 4-5). The Director of RCCL, Darla Jean Shaw (“Shaw”), and the Division Administrator of PMU, Leslie Reed, report to Morris.¹² (D.E. 301 at 7-8, 13; D.E. 323 at 160-61; PX 811 at 1, 4). Morris has an undergraduate degree in Accounting from the University of Texas at Austin and is a Certified Public Accountant. (D.E. 301 at 6). He first joined DFPS in 2008 in the Internal Audit division. Id. Morris was appointed Assistant Commissioner for CCL in July 2013. Id. at 5. The Court was also troubled by Morris's testimony. He had no experience in child welfare before joining DFPS in 2008. (Id. at 6; D.E. 323 at 159). He was hired because of his regulation experience with the U.S. Coastguard. Id. Morris generally defers to Shaw because he “didn't come into Licensing with a great deal of [relevant] background.” (D.E. 323 at 152-53). Like Black, Morris had not reviewed crucial reports, including several produced by PMU. (See, e.g., D.E. 301 at 69-71, 94). Morris was often unaware of important information relevant to his division, repeatedly stating that he relied on Shaw to know that information. See, e.g. , id. at 62–63, 89–90, 93.

12 While Morris oversees all of the units within CCL, RCCL and PMU are the only ones relevant to this case.

It is clear to the Court that a main reason DFPS has not improved in the face of decades' of reports outlining deficiencies and recommending solutions is that there is no institutional memory. Black and Morris never read critical reports about the departments they are charged with overseeing. It is no surprise then that Texas commissions full-scale audits of its foster care system every five to ten years, always reaching the same conclusions, but never producing improvement.

Darla Jean Shaw is the Director of RCCL and reports to Morris. (D.E. 329 at 5). She has held that position since 2010, but has been at DFPS since 1995 (in CPS and RCCL) as an investigator, caseworker, inspector, supervisor, and district manager. (DX 258). RCCL licenses, investigates, and monitors Texas's residential foster care facilities. (D.E. 329 at 5).

Christina Rawls-Martin (“Rawls-Martin”) was chair of the DFPS Council at the time of trial. (D.E. 308 at 5). The DFPS Council's job is to “assist the Commissioner in developing rules and policies” for DFPS and “to study and make recommendations to the Commissioner concerning the management and operations” of DFPS. Id. at 10, 13. Rawls-Martin has been on the Council since 2009 and was named chair by Governor Perry in 2013. Id. at 7–8. Although Rawls-Martin's term appears to have expired on February 1, 2015, she is still listed as chair on the Council website.¹³ Rawls-Martin is also vice-chair and founding member of the Hidalgo and Starr County Children's Advocacy Center. (D.E. 308 at 11).

13 See DFPS Council, https://www.dfps.state.tx.us/About_DFPS/Public_Meetings/Council/ (last visited December 10, 2015).

Judy Bowman is the Regional Director for CPS Regions 4 and 5, which encompass 38 counties in East Texas. (D.E. 327 at 7, 11). Bowman is the highest-level DFPS employee in Regions 4 and 5 and reports to McCall. Bowman has been at DFPS for more than 39 years, beginning as a caseworker and later being promoted to supervisor, program director, program administrator, and district director before becoming Regional Director in 2005. Id. at 7–10. As Regional Director, she oversees the decision-making programs for investigations, family-based safety services, conservatorship, kinship, and family groups in her regions. Id. at 10. Bowman manages approximately 750 employees. Id. at 12–15. There are approximately 160 caseworkers in Region 4 and 95 caseworkers in Region 5, caring for approximately 1950 and 1000 children, respectively. Id. at 14–15, 33–34. Approximately 35% of these children are in PMC. Id. at 33.

Camille Gilliam (“Gilliam”) is the Regional Director for CPS Regions 1 and 9, which encompass 71 counties in West and Northwest Texas. Id. at 56, 59–60. Like Bowman, Gilliam reports to McCall. She has been at DFPS for nearly 34 years, also starting as a caseworker. Gilliam was promoted to supervisor, program director, program administrator, and in 2005 to Regional Director of Region 1. Id. at 56–58. She has been Regional Director of Regions 1 and 9 for the last year, but at different times had also been the acting Regional Director for Region 10 and acting Regional Director for Regions 1, 2, and 9. Id. at 59, 61. There are 114 caseworkers in Region 1 and 75 caseworkers in Region 9 caring for approximately 1600 and 1150 children, respectively. Id. at 63, 67.

George Cannata is the Deputy Assistant Commissioner for CPS. Id. at 93. He reports to Black. He has been at DFPS for 18 years, working primarily in investigations and family-based safety services, starting as an investigative caseworker before being promoted to supervisor, program specialist, program manager, program director, program administrator, and eventually Deputy Assistant Commissioner in August 2014. Id. at 93–97. His current responsibilities include heading CPS “Transformation.” (Id. at 141; D.E. 313 at 6). Transformation is CPs's response to the 2014 evaluations by The Stephen Group and Sunset Commission. (D.E. 327 at 141-43).

Jenny Hinson (“Hinson”) is the Division Administrator for Permanency at CPS. (D.E. 314 at 4). She supervises a team of seven subject matter experts and with them is responsible for developing and administering policies and programs for children in DFPS conservatorship. (See id. at 221; DX 259 at 1). She has been in her current position since 2010, but has been at DFPS since 1998. Hinson has worked as a statewide intake specialist, caseworker, supervisor, program director, program administrator, and program specialist. (D.E. 327 at 222-23; DX 259 at 1-2). While Hinson focuses on CPS policies that relate to permanence, she seems to know little about how her policies actually affect permanency. Id. at 6–9. For example, she oversees extended foster care for children who age out, but does not know how many children benefit from that program, or the effectiveness of that program. Id. at 8–9. Similarly, while she said that independent living classes are offered to all foster children age 16 and up, Hinson does not know how many children attend those classes. She believes that number is fewer than 50 out of the 1300-1400 children who age out annually. (Id. at 21–22; see also DX 24 at 14 (showing that 1410 youths aged out in 2011); DX 119 at 221 (showing that 1328 youths aged out in 2013)).

Gail Gonzalez (“Gonzalez”) is the Director of Placement at CPS and reports to Black. (D.E. 328 at 88, 124; DX 257 at 3). She also oversees the Foster and Adoptive Home Development and Interstate Compact for Placement of Children programs. (D.E. 328 at 88, 100-101; DX 257 at 3). She has been in that position since 2009 but has worked at DFPS (mostly in CPS) since 1988 as a specialist, supervisor, placement team lead, and unit manager. (DX 257 at 1-3). In 2009 Gonzalez was appointed to the Texas Adoption Review Committee discussed supra , serving as the Committee's vice-chair. (PX 1964 at 2, 4; supra pp. 10–11). In her current position, Gonzalez oversees CPs's placement function, which, at a high level, entails “ensuring that [CPS has] some level of expertise when ... matching children to the best placement options.” (D.E. 328 at 104). Her responsibilities also include overseeing policy and practice relating to services for children in State-regulated facilities as well as policy and practice relating to CPAs (for both the 122 private providers that DFPS contracts with and each CPS region when those regions act as CPAs). Id. at 88–89. The Foster and Adoptive Home Development program is responsible for recruiting, verifying, training, and monitoring foster and adoptive homes and families. Id. at 94–96. Gonzalez additionally works with special projects and initiatives, liaises with central and regional leadership, legislators, and other stakeholders, and coordinates and leads implementation of legislative changes. (DX 257 at 3).

Kaysie Reinhardt (“Reinhardt”) is the Director of Foster Care Redesign, Texas's attempt to improve its placement array. She reports to Black. (DX 260 at 1; PX 811 at 1). She has held that position since 2013, but has been at DFPS since 2000 as a conservatorship caseworker, centralized placement coordinator, lead investigator, supervisor, program specialist (for multiple different programs), and project manager. (Id. at 1–3; D.E. 328 at 178). Before joining DFPS she was a substance abuse counselor and a residential counselor at an RTC. (DX 260 at 3-4). In her current position she oversees and manages the implementation of Foster Care Redesign. (D.E. 328 at 178). She is also the head of the Public Private Partnership, the body guiding Foster Care Redesign. Id. at 181.

Frianita Wilson (“F. Wilson”) is the Director of Purchased Client Services for CPS and reports to Black. (D.E. 329 at 53, 56-57; DX 256). She has held that position since 2013. She monitors the private contracted services for residential foster care placements (meaning services provide by CPAs, GROs, RTCs, etc.). Id. F. Wilson had previously worked for DFPS in 1996-2003 in contract management and monitoring before spending 10 years with HHSC in its contract compliance department. (D.E. 329 at 55; DX 256). The DFPS Contract Oversight and Support division sets the policies and procedures that Purchased Client Services follows in maintaining and managing contracts. (D.E. 329 at 57). Additionally, when Purchased Client Services staff conduct risk assessments (as part of their monitoring of CPAs, GROs, etc.), they submit the results of those assessments to Contract Oversight and Support. Id. at 59.

Christine Maldonado (“Maldonado”) is the Contract Oversight and Support Director at DFPS. (D.E. 329 at 89). She reports to DFPS's Chief Operating Officer. Id. at 92. She establishes policies, procedures, and training for contract staff, and conducts quality assurance of contract management. (D.E. 329 at 90). Additionally, when receiving risk assessments from Purchased Client Services Staff, Contract Oversight and Support uses the results to adjust its statewide contract monitoring plan. Id. at 59–61. Maldonado has held this position since 2009 and has been at DFPS for over 18 years (primarily in Contract Oversight and Support), as an investigative caseworker, contract manager, policy and technical assistance specialist, and quality assurance team lead. (DX 261).

John Specia, Jr. is the Commissioner of DFPS. He has held that position since December 2012. (D.E. 331 at 21). Specia's career in child welfare and public service is extensive and commendable. (DX 276). Specia began to practice in child welfare in 1978. Id. at 23. Specia practiced as a child welfare attorney for seven and a half years in Region 8 (for the Texas Department of Human Resources) before becoming an associate judge for the juvenile court in Bexar County. Id. at 23–24. He later served as a judge for the 225th District Court in San Antonio, with special jurisdiction in the area of child abuse and neglect, for 18 and a half years. Id. at 24. He eventually helped establish the Texas Supreme Court's Children's Commissions as well as various Child Protection Courts throughout Texas. (DX 276 at 1). Further, he helped establish the Bexar County Children's Court, which provides specialized services for children. Id. Specia was previously vice-chair of the Texas Supreme Court's Permanent Judicial Commission for Children, Youth, and Families and chair of the Supreme Court Task Force on Foster Care. Id. Over the years he has received multiple awards, most recently the Texas Appleseed Award in 2014 for his impact on improving the lives of Texas's most vulnerable children. Id. at 2. As Commissioner, Specia oversees approximately 12,000 employees, including 8000 in CPS. (D.E. 331 at 31).

b. Expert Witnesses

Federal Rule of Evidence 702 requires district courts to act as gatekeepers, excluding expert testimony that is not relevant or reliable. Fed. R. Evid. 702 ; see also Kumho Tire Co., Ltd. v. Carmichael , 526 U.S. 137, 152, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999) ; Daubert v. Merrell Dow Pharm., Inc. , 509 U.S. 579, 597, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). A trial court may consider several non-exclusive factors when evaluating the reliability of an expert's opinion, including: “(1) whether a theory or technique can be tested; (2) whether the theory or technique has been subjected to peer review and publication; (3) the known or potential rate of error; (4) the existence and maintenance of standards and controls; and (5) general acceptance of the theory in the scientific or expert community.” United States v. Valencia , 600 F.3d 389, 424 (5th Cir.2010) (citing Daubert, 509 U.S. at 593–95, 113 S.Ct. 2786 ). The list of Daubert factors “neither necessarily nor exclusively applies to all experts or in every case.” Kumho Tire , 526 U.S. at 141, 119 S.Ct. 1167. District courts have “wide latitude in determining the admissibility of expert testimony, and the discretion of the trial judge will not be disturbed on appeal unless manifestly erroneous.” United States v. Cooks , 589 F.3d 173, 179 (5th Cir.2009) (quoting Watkins v. Telsmith, Inc. , 121 F.3d 984, 988 (5th Cir.1997) ). The safeguards outlined in Daubert are less essential in a bench trial. Gibbs v. Gibbs , 210 F.3d 491, 500 (5th Cir.2000). Regardless of how a court evaluates an expert's reliability, the object of the gatekeeping responsibility is “to make certain that an expert, whether basing testimony upon professional studies or personal experience, employs in the courtroom the same level of intellectual rigor that characterizes the practice of an expert in the relevant field.” Kumho Tire , 526 U.S. at 152, 119 S.Ct. 1167 ; Valencia , 600 F.3d at 424.

According to these principles, the Court allowed all proffered experts to testify during trial over both parties' objections. The Court let both parties cross-examine the experts and challenge their opinions during trial rather than conduct separate time-consuming reliability hearings. The Court similarly admitted all proffered experts' reports without first determining their relevance or reliability, as those reports supported the experts' testimony. Defendants objected to the admission of the expert reports as hearsay. When reviewing the reports after trial, the Court only considered the facts and data that “experts in the particular field would reasonably rely on ... in forming an opinion on the subject.” Fed. R. Evid. 703 ; see also Daubert , 509 U.S. at 595, 113 S.Ct. 2786. The Court disregarded all other parts as inadmissible hearsay. (See, e.g. , D.E. 325 at 13-14). Even though all experts were qualified by education, knowledge, or experience, several experts' opinions were not credible or unreliable. The Court disregarded the reports and testimony of Drs. Dennis Zeller and Edwin Basham in their entirety and accorded reduced weight to the reports and testimony of others.

Dr. Dennis Zeller (“Zeller”) is Plaintiffs' expert on child welfare performance measurements. (See D.E. 322 at 129-33, 141). Zeller has a B.A. in Philosophy from the University of Oklahoma, a Ph.D. in Philosophy from the University of Texas at Austin, and a master's in Social Work from the University of Texas at Austin. Zeller is currently the president of Hornby Zeller Associates, Inc., a management consulting firm specializing in child welfare, juvenile justice, mental health, and public assistance. He and his firm have consulted for 29 states in the area of child welfare. Those consulting projects include analyzing whether child welfare agencies are meeting their expected outcomes and compliance goals, performing quality assurance and case reviews, and conducting workload studies to determine the number of caseworkers that a given agency needs. His firm regularly produces reports in connection with these projects, which include findings as well as recommendations for improvement. For example, when he and his firm audited the performance of the Oklahoma Department of Human Services (in 2008-2009), focusing specifically on child welfare, they produced a report with 26 recommendations, 25 of which the legislature passed into law. Prior to starting his consulting firm in 1988, Zeller was a program research specialist, supervisor, and later director of the Policy Analysis Unit at the New York State Division of Family and Children's Services. In New York, he developed or oversaw the development of regulations for child protective services, proposed legislation for preventive services, and standards for eligibility for preventive services. Before that, between 1977 and 1980, he was a statistician and supervisor of the Survey and Analysis Unit at the Texas Department of Human Services.

Although Zeller, based on his experience and education, was accepted without objection as an expert on child welfare performance measurement, the Court finds Zeller's conclusions unreliable. First, Zeller did not support several of his opinions. His connections between the policies and practices of DFPS and the outcomes for Texas foster children were arrived at with only a cursory review of the actual outcomes for children in Texas's PMC. See, e.g. , id. at 183. For example, Zeller concluded that Texas's caseworker workloads affect PMC children's outcomes. His finding, however, was based entirely on a review of other states' foster care systems. He never analyzed this data for Texas. Id. at 172. He made the same mistake when asserting a connection between the frequency of parental visits in foster care and PMC children's outcomes. Id. at 181. Second, Zeller failed to provide context for his conclusions. He testified that 75% of children in PMC who were separated from their siblings had no sibling visitation over a given three-month period, but Zeller did not know how many children comprised that 75% figure or how many sibling groups were separated in PMC during that three-month period. Id. at 184–90. Lacking context, the Court cannot determine how often siblings are separated. Zeller repeated these mistakes throughout trial. See id. at 193–96, 199–203, 206–11. The Court therefore cannot verify the reliability of his testimony and report. The Court disregards Zeller's report and testimony in their entirety and any part of testimony by other experts that relied on Zeller's data or report.

Dr. William Lee Carter (“Carter”) is Plaintiffs' child psychology expert. He has been a licensed psychologist for over 31 years. (D.E. 326 at 69-70). Carter received his B.A. in Psychology as well as his M.S. and Ed.D. in Counseling Psychology from Baylor University. (PX 2015 at 38 (filed under seal)). Over his career, he has worked directly for school districts, hospitals, and clinics, working almost exclusively with children. See id. at 39–40. Carter also spent five years as the Director of Psychology at an RTC. Id. at 40. He has published eight books relating to the field of child psychology and welfare. Id. Carter has performed around 7500 psychological evaluations of children and youths, two thousand of which were of children involved with CPS, and provided counseling services for another one thousand children in Texas psychiatric hospitals and RTCs. Carter currently consults for CPS, several CPAs, a juvenile detention center, a special education co-op, an advocacy center medical advisory board, and an RTC. Id. at 38–39. Carter has conducted many expert psychological examinations for DFPS and for both plaintiffs and defendants in criminal and civil matters. Id. at 2. In the last five years, he has consulted or testified as an expert in over 200 cases, most often on behalf of the State of Texas in child sexual abuse prosecutions. (See id. at 158–70; D.E. 326 at 71).

Carter conducted comprehensive psychological evaluations of some of the Named Plaintiffs in 2012 and conducted follow-up evaluations in 2014. (PX 2015 at 3 (filed under seal)). His evaluations included a clinical interview with each child, an observation of child-caregiver interactions, a personal history questionnaire, an intelligence test, a behavioral assessment, and a risk evaluation. Id. He also reviewed the children's DFPS case files and, when possible, interviewed their treatment providers and caregivers. Id. Carter paid particular “attention to the progression of each Named Plaintiff's psychological well-being over the course of time spent” in foster care. Id. Carter found that the experiences of the Named Plaintiffs while in DFPS custody, including the psychological harm that they suffered, were typical for the Texas foster children he evaluated and counseled over the years. (D.E. 326 at 132-33, 194). The Court finds Carter's reports and testimony relevant, reliable, and admissible.

Daryl Chansuthus (“Chansuthus”) is Plaintiffs' expert on continuous quality improvement and appropriate staffing levels for oversight functions in child welfare systems. (D.E. 324 at 88). She has, among other degrees, a master's in Social Work from the University of Tennessee at Knoxville. (D.E. 324 at 74). She began her career in social work before joining the Tennessee Department of Children's Services (“DCS”) (Tennessee's CPS equivalent) in 2002 as a compliance monitor. In 2004, Chansuthus was promoted to director of the Quality Assurance and Continuous Quality Improvement division. Id. at 80. In 2006, she was promoted to the newly created executive director position and was given responsibilities over the accreditation, evaluation and monitoring, policy and planning, program accountability review, and licensing divisions. Id. at 81. Part of her oversight of the accreditation division included helping Tennessee become accredited by the COA. Id. at 82. In both positions she reported directly to the Commissioner of DCS. Id. She also helped ensure Tennessee's compliance with the settlement agreement arising from Brian A. by Brooks v. Sundquist , 149 F.Supp.2d 941 (M.D.Tenn.2000). (PX 2022 at 3). Further, when Chansuthus was executive director, she helped design and implement a continuous quality improvement structure for DCS, which determines how well the agency performs its required tasks, and whether those tasks lead to desired outcomes. (D.E. 324 at 79-80).

In 2007, Chansuthus left Tennessee's CPS equivalent and became the executive director of the Tennessee Center for Child Welfare, a consortium of university social work departments that aided Tennessee's CPS. Id. at 77, 83. In that role, she worked primarily with DCS staff to identify practice strengths and weaknesses and to provide training and support to address the areas of identified weakness. (D.E. 324 at 77, 83). In 2013, she became the executive director of the Wo/Men's Resource and Rape Assistance Program, where she continues to work with children and families who have experienced domestic or sexual violence. Id. at 78. She has also worked, since 2007, at the University of Tennessee, first as a clinical assistant professor and then as an adjunct professor, in the areas of social work and child welfare. (PX 2022 at 2-3). The Court finds Chansuthus's report and testimony relevant, reliable, and admissible.

Dr. Viola Miller (“Dr. Miller”) is Plaintiffs' child welfare systems expert. (D.E. 302 at 6-7). Plaintiffs asked Dr. Miller primarily to review how caseworker caseloads and placement arrays affect PMC children. Id. at 15. Her extensive experience and background in child welfare was unrivaled by any other witness, expert or otherwise, testifying at trial. Dr. Miller's career in child welfare spans more than 40 years. (Id. at 7, 12; PX 2037 at 64-66). From 1995 to 2003 she served as Secretary of Kentucky's equivalent to Health and Human Services (Specia's position), overseeing approximately 10,000 employees and a budget of $950 million. (D.E. 302 at 9-10; PX 2037 at 64). Dr. Miller was then Commissioner of Tennessee's DCS from 2003 to 2011, overseeing approximately 5000 employees and a budget of $640 million. (PX 2037 at 64). She significantly improved the child welfare systems in both states. (D.E. 302 at 6-7). Since retiring in 2011, Dr. Miller has consulted for child welfare agencies in Massachusetts, Illinois, and Oklahoma, and volunteers on the boards of child welfare organizations. Id. at 12–14. She has authored 24 publications about social work, child welfare, and child development. (PX 2037 at 66-68). She has also over the years presented at countless workshops and conference in those fields. See id. at 69–73. At least one other court recently relied on Dr. Miller's expert testimony on child welfare systems and caseworkers caseloads in a class action similar to the present case. D.G. ex rel. Strickland v. Yarbrough , 278 F.R.D. 635 (N.D.Okla.2011) (denying defendants' motion to decertify the class after full discover). The Court finds the majority of Dr. Miller's report and testimony relevant, reliable, and admissible. The Court disregarded any references in Dr. Miller's report to Zeller's data or report and any opinions or conclusions offered by Dr. Miller that were based on Zeller's analysis.

Dr. Caroline Long Burry (“Burry”) is Plaintiffs' social work expert. (D.E. 325 at 12). She has a B.A. in Sociology from Furman University, a master's in Social Work from the University of Georgia (with a specialization in families and children), and a Ph.D. in Social Work from the University of South Carolina (her dissertation was based on creating and evaluating a training program for foster parents). (PX 2003 at 113 (filed under seal)). Burry is currently an associate professor of Social Work at the University of Maryland in Baltimore and chair of the Families and Children specialization within that department. (D.E. 325 at 15). She is licensed at the highest level for a social worker in both South Carolina and Maryland. Id. at 18. She has been a social worker in Georgia and South Carolina and a therapeutic foster parent in South Carolina. (PX 2003 at 118-19). Between 1992 and 2000 she founded and then co-directed the Adoption Center of South Carolina, which provided adoption-related services to children and families. Id. at 118. She has extensive experience teaching, mentoring, researching, and consulting in social work and child welfare. (Id. at 18–19; PX 2003 at 113-21 (filed under seal)). She has also given countless presentations and authored or co-authored many articles, book chapters, and training manuals in the fields of social work and child welfare. (See PX 2003 at 121-34 (filed under seal)). For this case, she reviewed the case files of three Named Plaintiffs (S.A., H.V., and A.M.), evaluating each child's casework. (D.E. 325 at 19-21). The Court finds Burry's report and testimony relevant, reliable, and admissible.

Mary Dee Richter (“Richter”) is Plaintiffs' foster group home expert. (D.E. 326 at 5; PX 2014 (filed under seal)). Richter has a B.S. in Education from Northwest Missouri State University and an M.S. in Education from the College of Saint Rose. (PX 2014 at 45 (filed under seal)). She is currently an independent consultant for juvenile justice and child welfare programs. Id. She began her career as a speech pathologist, working with children in various settings. Id. Richter then spent 15 years working with Father Flanagan's Boys' Town, an agency operating group homes for foster children in Nebraska, as both an on-site staff member and a manager. (D.E. 326 at 7-9). She started as a group home parent/family teacher, caring (along with her husband) for nine teenage boys in a group home, and later became the assistant community director for Boys' Town, supervising nine additional group homes. Id. at 8. She then became the site director for a pilot Boys' Town branch in Florida, operating first in Tallahassee and later also in Orlando. Id. at 8–9. Richter then worked for five years at the Florida Department of Children and Families (“DCF”) (Florida's DFPS) as the chief of Permanency Planning. Id. at 9–10. She left the Florida DCF for ten years to be the executive director of the Florida Network of Youth and Family Services, which was a consortium of private CPAs that served approximately 20,000 children each year, before returning to DCF as assistant director for Family Safety. Id. at 11. For the present case, Richter evaluated Texas's use of foster group homes by considering CWLA and COA standards, Texas state law and DFPS policy, data from DFPS on foster group homes between 2012-14, and seventy investigation files for incidents at Texas foster group homes in 2012. (PX 2014 at 7 (filed under seal)). The Court finds Richter's report and testimony relevant, reliable, and admissible.

Laurie Bensky (“Bensky”) is Plaintiffs' expert statistician. Bensky has a B.A. in Psychology and a master's in Social Work. (D.E. 325 at 187). She is currently a senior policy analyst at Children's Rights. Id. She previously worked as a foster care caseworker and supervisor. Id. Before joining Children's Rights she worked at the New York City Administration for Children's Services (essentially New York City's CPS), in the Quality Assurance department. Id. at 188. Her testimony laid the foundation for three Plaintiffs' exhibits, which reflected PMC caseworker caseloads, and the number of children 12 and younger in DFPS custody for more than 18 months (i.e. PMC) that reside in group homes and institutions. Id. The Court finds Bensky's report and testimony relevant, reliable, and admissible.

Dr. Alan Salzberg (“Salzberg”) is Defendants' expert who was called to review and rebut Zeller's report and testimony. (DX 252). Salzberg is currently the director at Summit Consulting, an economics and statistical consulting firm. Id. at 57. While Salzberg has extensive background in statistical consulting, unlike Zeller, he does not have any background or experience in child welfare. (See DX 240; D.E. 328 at 61). Salzberg did not perform any statistical analyses of his own and did not offer any independent opinions or conclusions. (See D.E. 328 at 70-71, 80). Because the Court disregarded Zeller's findings, Salzberg's testimony was not relevant. (D.E. 328 at 60-64). Dr. Edwin Basham (“Basham”) is Defendants' expert psychologist. Basham has a B.S. in Psychology as well as an M.S. in Marriage and Family Counseling from Kansas State University, and an Ed.D. in Counseling Psychology from the University of Northern Colorado. (DX 242 at 2). For the past 25 years he has worked as a contract psychologist for DFPS. Id. at 1. He has also worked as a contract psychologist for a children's temporary shelter, two juvenile probation departments, and a psychiatric hospital. Id. Before that he was a staff psychologist at an RTC, two psychiatric hospitals, and an outpatient clinic. Id. He has performed over 4,000 psychological evaluations of adults and children over his career. Id. at 3. He has also appeared in court as an expert witness in over 200 cases, the vast majority on behalf of DFPS. Id. at 3–4.

Basham was offered to rebut Carter's and Burry's conclusions that the Named Plaintiffs suffered harm in the State's care. (D.E. 330 at 52). Basham performed forensic evaluations of the Named Plaintiffs in order to compare his evaluations to Carter's. Id. at 67. The Court disregards Basham's testimony because of his misconduct and self-contradiction.

In attempting to refute Carter's conclusion that A.M. suffered harm in the State's care, Basham shared Carter's findings with A.M. and asked her whether she agreed. (DX 266 at 7-8 (filed under seal)). Basham's conduct is shocking. Many of Carter's conclusions were critical of A.M.'s ability to function in society. For example, Carter found that A.M. “does not possess [the] basic tools ... to maintain friendships” and that “she is drastically unprepared for future marriage or family life.” See id. Basham parroted those findings to A.M. nonetheless. Moreover, A.M. likely believed that her meetings with Carter were confidential. For a fragile youth such as A.M., who already struggles to trust other people, Basham's conduct is devastating. (D.E. 326 at 137-38). The kind of damage that Basham may have caused to A.M. is precisely what the Court sought to avoid when allowing these psychological evaluations, and the type of inappropriate conduct that the Court previously warned all parties against.

Basham also contradicted himself during a phone conference with the Court on November 4, 2014, describing his interview with A.M. differently than he did in his report. (D.E. 330 at 81-82, 108-111). In Basham's declaration, he purported to reconstruct from his notes his conversation with A.M., directly contradicting the plain language of his written report. (See D.E. 274-2). Basham attempted to explain that he actually read to A.M. statements that Carter attributed to her, which Carter had allegedly noted in his report. However, several of the statements that Basham read to A.M. were not attributed to her by Carter, as direct quotes or otherwise, and were instead taken directly from Carter's actual clinical impressions and diagnostic conclusions. (See PX 2015 at 152-56 (filed under seal)). Basham also impeached himself at trial, describing his interview with A.M. differently than he had in either his report or during the telephone conference. (See D.E. 330 at 81-82, 108-111). He further admitted that, even though his report said A.M. disagreed with one of Carter's statements, A.M was actually “ambivalent” about the statement. Id. at 81. Basham showed an overall indifference to A.M.'s well-being, professional standards, and the truth.

Even without these transgressions, Basham's findings are unreliable. At his deposition in 2012, Basham denied seeing evidence in the Named Plaintiffs' case files of psychological or emotional harm, even though the files for A.M., D.I., and S.A. each contained reports of sexual abuse. Id. at 86–87, 90–93. When pressed on this finding at trial, Basham first claimed that he misunderstood the question during his deposition. He then said that he never reviewed his deposition to submit corrections. Later, however, after being corrected by Defendants' counsel, Basham admitted that he did review the deposition and submitted corrections, but just not to that statement. Id. at 86–90. Basham also said that D.I. showed “substantial improvement” since entering foster care. (DX 262 at 21 (filed under seal)). In reality, D.I. suffered three episodes of rape at the age of eight that were “initiated by adolescent boys in the home over a period of two weeks.” Id. at 56. D.I. is now “very disturbed,” “heavily sexualized,” and “a high risk for sexually harming children.” (D.E. 326 at 190-91). The Court completely disregards Basham's reports and testimony for this case, and is all too aware that Basham has long provided testimony on behalf of the State of Texas.

Linda McNall (“McNall”) is Defendants' expert on social work. McNall has a B.A. from Adams State College and a master's in Social Work from the University of Michigan. (DX 244 at 1). For most of her long career in child welfare she has worked primarily for the New Mexico Children, Youth & Families Department (“CYFD”) (essentially New Mexico's DFPS). (See id. at 1–3; D.E. 329 at 123-26). Unlike Richter, McNall has no experience working directly in, or managing, group homes. (See DX 244). Before joining CYFD she worked with both children and adults as a social worker and adoption consultant. (DX 244 at 2-3). She joined CYFD in 1989 as an adoption supervisor and was later promoted to county office manager, policy and procedures bureau chief, adoption program manager, and finally regional manager. Id. at 1–2. McNall retired from CYFD in 2011 and is currently an independent consultant. (D.E. 329 at 127). McNall was offered to rebut Richter's report and testimony. (D.E. 329 at 124). She wrote two reports: the first analyzed 49 investigations conducted in foster group homes; the second analyzed 57 others. Id. at 129. She did not perform any independent study of Texas's foster group homes. (D.E. 319 at 4). The Court finds McNall's reports and testimony unreliable.

First, McNall impeached herself. She admitted at trial that unrelated children living in a home are exposed to different risks than related children living in a home; she said the opposite in her report. Id. at 24–25. Second, McNall's reports say that in the 106 investigations she reviewed, she saw no “indication that any of the incidents resulted in traumatic or long lasting harm to any of the children involved, either as victims or alleged perpetrators.” (DX 253 at 11 (filed under seal)). In one investigation reviewed by McNall, DFPS confirmed that in a ten-child foster group home, a 14-year-old male foster child sexually molested a four-year-old female foster child, causing her vagina to bleed. (Id. ; D.E. 319 at 21-22). The boy had previously inappropriately touched an 18-year-old female foster child and a 13-year-old male foster child in that same home, both of whom had mental retardation. Id. at 22–23. DFPS confirmed negligent supervision because the foster parents were on vacation when the boy molested the four-year-old. Id. It is unclear why McNall opined that there was no traumatic or long-lasting harm in any of the reviewed cases, in part because she has no expertise in that area and was not retained to offer opinions on that topic. See id. at 18. Even more troubling, McNall admitted at trial that the above incident could have caused long-term harm, and that she made her statement “probably ... without thinking it through completely.” Id. at 23–24. McNall also admitted at trial, despite stating to the contrary in her report, that multiple unrelated children living in one home are exposed to different risks than a similar number of related children living in a single home. See id. at 24–25. The Court disregards her reports and testimony.

Patricia Wilson (“P. Wilson”) is Defendants' expert on child welfare systems, and was retained to rebut the reports and testimony of Zeller, Dr. Miller, and Richter. (DX 250 at 4; D.E. 329 at 138). She has a B.A. in Education and a master's in Social Work, both from the University of Kentucky. (DX 238 at 3). She is currently a human services consultant, with an emphasis on child welfare. Id. at 1. P. Wilson's career in child welfare began in 1973 in Kentucky, first as a front line caseworker, and later as a county office supervisor, working primarily with foster children and parents. (D.E. 329 at 136-37). In 1992, she became a policy and budget analyst and then a budget director in Kentucky's Cabinet for Families and Children (Kentucky's HHSC). Id. at 135, 137. During this time, Dr. Miller was Secretary of Kentucky's Cabinet for Families and Children, and P. Wilson was on Dr. Miller's staff. See id. at 135. Wilson also worked for two and a half years as a consultant and later as a regional director for the CWLA. Id. at 137. She then worked for a year at a child caring/placing organization in Tennessee before returning to Kentucky in 2007 to work as an executive advisor in its Department of Community Based Services (essentially Kentucky's DFPS). (Id. at 135, 137–38; DX 250 at 4). Between 2008 and 2011, she served as the commissioner of that agency. (D.E. 329 at 138). In that role, she was the number two person in Kentucky's Health and Social Services System. (D.E. 320 at 8). P. Wilson stated at trial that, despite disagreeing with Dr. Miller's expert opinion, she respected Dr. Miller's professional judgment. (D.E. 329 at 141-42).

Parts of P. Wilson's report provided little value to the Court. For instance, she extensively analyzed DFPS based on the Child and Family Service Review seven indicators. (DX 250 at 9-30). As discussed supra , Texas's Review performance provides little relevant evidence to this case. Supra pp. 27–28.

P. Wilson's credibility was also undermined by the revelation that under her leadership, the Kentucky child welfare agency withheld records relating to Amy Dye, a nine-year-old girl who died in an adoptive home. (See D.E. 320 at 11, 14, 17-19). The agency had apparently approved the adoptive placement despite earlier reports of child abuse in the home. See id. at 14. The agency also failed to properly investigate, and did not substantiate, multiple reports from Amy's school that she was being abused. Id. Following her death, the agency performed an investigation (possibly based on an earlier report of abuse and neglect) and six months later found only that Amy's adoptive parents were guilty of being “inattentive caregiver[s]” and using “unusual discipline.” Id. at 55–56. The agency took no further action in relation to children still in that home. Id. at 56. Although the circumstances of Amy's death are troubling, it is not clear how much blame P. Wilson is to blame, even though she was Commissioner of Kentucky's foster care system at that time. The Court does fault P. Wilson for her agency's refusal to release to the public the records relating to Amy's death, a decision which was at least in part P. Wilson's. (D.E. 320 at 17). At trial, P. Wilson defended withholding the records even though it violated Kentucky law. (D.E. 320 at 17); see also Courier – Journal, Inc. v. Cabinet for Health & Family Servs. , 2011 WL 2173921, at *1 (E.D.Ky. June 1, 2011). Kentucky newspapers had to take the child welfare agency to court to force the release of Amy's records. (D.E. 320 at 17). The Kentucky court ordered the agency to release the records over P. Wilson's objections, fined the agency, ordered it to pay the newspaper's attorneys' fees, and admonished the agency for turning a “blind eye” to the child abuse that was reported prior to Amy's death. Id. at 17–18. The agency continued to withhold the records until the Governor of Kentucky ordered their release. Id. at 18. P. Wilson then testified about transparency in child welfare investigations before a joint committee of the Kentucky Legislature, where she said that, in Amy's case, Kentucky's child welfare system had “not failed to operate according to the laws and rules of the State.” Id. at 18–23. Wilson further undermined her credibility when she said at the present trial that foster care systems should not track child-on-child abuse, which is a critical aspect of foster child welfare. Id. at 22–24; see also infra Section IV.B.1. For these reasons, the Court accords lesser weight to P. Wilson's report and testimony.

Dr. Jane Burstain is Defendants' expert on child welfare policy who primarily rebutted the expert opinions of Zeller and Dr. Miller. (DX 251 at 2). Burstain joined DFPS after this case was filed, and is currently the Director of Systems Improvement for CPS. Id. at 2–3. Burstain has a B.S. in Finance from the University of Virginia, a J.D. from the UC Hastings College of Law, and a Ph.D. in Policy Analysis from the Pardee RAND Graduate School. (DX 267 at 1). She began her career in child welfare in 1999 as an attorney at the Children's Law Center in Los Angeles, California, where she represented children in child welfare proceedings and advocated on their behalves for various social, mental, and educational services. Id. She left that position in 2004 to pursue her doctorate. Id. After her doctorate, in 2008, she moved to Texas and joined the Center for Public Policy Priorities as a policy analyst, where she, inter alia , researched and wrote briefing papers and reports relating to policy and budget issues at DFPS. Id. In 2012, she assumed her current position with DFPS. Id. She has authored or co-authored at least 16 articles, reports, and working papers in the field of child welfare. Id. at 5. In her current position, Burstain considers herself and her ten staff akin to “ ‘internal consultants.’ ” (D.E. 330 at 144). They analyze how CPS is operating, identify areas for improvement, create improvement plans, and track how well those plans are followed and whether improvement occurs. Id. While she and her staff are not subject matter experts, they take information from DFPS's subject matter experts and work to turn that information into accessible department policy. See id. at 145. She reports to Cannata and previously reported to Black. Id.

Before Burstain joined CPS, she published several reports that were highly critical of DFPS. At trial, she backtracked and glossed over many of her published comments. (See, e.g. , D.E. 310 at 48-59; PX 1871-77). In her current position, Burstain has improperly manipulated DFPS statistics and data. For example, to suppress primary caseworker caseload numbers, Burstain included not only full-time primary caseworkers in her calculation, but also part-time primary caseworkers, secondary workers (who do not have the responsibilities of primary caseworkers), and fictional workers “created out of all the overtime,” which “are not actually even people.” (D.E. 310 at 65-69). On another occasion, Burstain e-mailed DFPS employees to suggest that four foster child deaths be re-characterized to negate the possibility of caregiver negligence. (PX 1711). In that e-mail, Burstain highlighted three cases in which she thought “reasonable minds might disagree about whether the caregivers could have prevented the deaths.” Id. at 1. One of these cases involved a child with a diagnosed history of Schizophrenia and Bipolar Disorder who hung herself in an RTC. Id. at 2. The child had recently been hospitalized for two weeks for suicidal ideations and self-harming behaviors. The child's level of care required RTC staff to check on her every 15 minutes during the night, but that night the caregiver did not perform the routine checks. Id. Another case involved a three-month-old infant with multiple medical problems, including obstructive sleep apnea and a smaller jaw that affected her breathing. Id. at 3. That child's foster mother placed her on her side with a foreign object (a “vibrating pillow”) in front of her and did not check on her for 30-75 minutes. Id. The foster parent “had been trained on safe sleep practices including the dangers of ... placing an infant in a face-down or side sleeping position and the dangers of placing foreign objects under or around an infant when that infant is sleeping.” Id. Besides the dubious implications of Burstain's request, it is unclear why, in her exclusively policy-related role at CPS, and with no background in licensing or investigations, Burstain sent that e-mail in the first place. (D.E. 310 at 27-28, 39-41). Specia said he had “no idea” why Burstain inserted herself into this area. (D.E. 299 at 24). The Court finds that Burstain was attempting to manipulate the State's child fatality statistics. In light of these incidents, the Court assigns lesser weight to Burstain's expert report and testimony.

5. Named Plaintiffs

The Court does not base any system-wide findings solely on the Named Plaintiffs' experiences in foster care. Their experiences, however, paint a similar picture: children often enter foster care at the Basic service level, are assigned a carousel of overburdened caseworkers, suffer abuse and neglect that is rarely confirmed or treated, are shuttled between placements—often inappropriate for their needs—throughout the State, are migrated through schools at a rate that makes academic achievement impossible, are medicated with psychotropic drugs, and then age out of foster care at the Intense service level, damaged, institutionalized, and unable to succeed as adults. Numerous witnesses testified that the Named Plaintiffs' experiences are typical for children in Texas's PMC. (See, e.g. , D.E. 326 at 132-33, 216, 225; D.E. 323 at 18; D.E. 324 at 215-17; D.E. 327 at 189-92). The Named Plaintiffs' narratives are the lens through which the Court viewed Plaintiffs' systemic and class-wide evidence. The Court has written out these lengthy narratives in part to establish a record, but also because these children have for too long been forgotten.¹⁴ Their stories deserve to be told.

14 (See PX 1966) (discussing Carole Keeton Strayhorn, Forgotten Children, A Special Report on the Texas Foster Care System (2004)).

a. M.D.

M.D. is a class representative for the General Class and the Licensed Foster Care Subclass. At the time of trial, M.D. was 17 years old. She was still in the State's PMC, but her whereabouts were unknown. She was presumed to be living “on the street somewhere,” in the Houston area, over 200 miles away from her home community of Corpus Christi, Texas. (D.E. 324 at 230). While in DFPS's custody, M.D. changed placements 19 times. Id. at 1 RFP CPS 175008-28, DFPS 009007116-17. On four separate occasions while in the State's care, M.D. was arrested and taken to juvenile detention facilities. Id. at DFPS # 28544, # 28574-75, # 30210. She was placed in psychiatric hospitals 11 times and went to emergency rooms twice for suicidal ideations and cutting. Id. at DFPS # 28399-428, DFPS 009007117-18, 009011721-819. Many of M.D.'s placements moved her far from her siblings, who were not in foster care, and her home community. She had 16 different primary and secondary caseworkers. M.D. attended nine schools over her seven years in State care. Id. at DFPS # 28382, 28394, 28397; DFPS 009007119-20, 009008456-60. During fourth grade alone, she attended five different schools. Id. at DFPS0098458-60.

M.D. and her siblings were raised by their maternal grandparents until they died in 2005. Id. at DFPS # 29816. The children then lived with their mother who had a history of drug use with resulting jail and prison sentences. Their mother also had Bipolar Disorder. M.D.'s father lived in Mexico and was never involved in her life. In 2005, M.D.'s mother tested positive for illegal drugs. Id. at DFPS # 29815-17. When CPS determined that the mother used marijuana and cocaine in the home, CPS found Reason to Believe (“RTB”) neglectful supervision. The mother was sentenced to seven years imprisonment by the State of Texas for testing positive for drugs in violation of a prior probation. Id. at # 29934. In February 2006, a maternal aunt took custody of the children to prevent them from entering foster care. Id. at DFPS # 29815-17.

M.D. had significant behavior problems while in her aunt's care. M.D. was violent toward her aunt, oppositional, and threatened to report fictitious claims of abuse to CPS. Id. at DFPS009014656. In March 2007, at the age of ten M.D. was admitted to a psychiatric hospital to determine if she qualified for long-term psychiatric care and to help her adjust to her medication regimen. Id. at DFPS # 29751. While there, M.D. claimed that her much older cousin, who did not live in the aunt's home, had been sexually abusing her for several years. Id. at DFPS # 28220-29. The hospital reported this outcry to CPS. RCCL interviewed M.D. and found RTB that she was sexually abused. Id. at DFPS # 28277, # 28283, # 229847. M.D.'s case files do not indicate that any action was taken as a consequence.

After M.D. was released from the hospital to her aunt, she started acting out at school, became defiant at home, and ran away. Id. at DFPS # 282230-35. She was found by police and while being transported back to her aunt's, M.D. assaulted the officer, was arrested, and taken to a juvenile detention center. M.D.'s maternal aunt refused to take M.D. back in March 2007, and formally relinquished custody of M.D. to CPS in August 2007. Id. at DFPS # 28231, # 29829-38.

Over the next five months, M.D. was placed in an emergency shelter, a psychiatric hospital, and two different foster homes. She was removed from each foster home into hospitals for self-harming behaviors. Id. at DFPS009011721-36. In November 2007, after her third hospitalization, M.D. was placed in an RTC where she remained for over two years. Id. at DFPS009011733. M.D.'s behavior problems persisted at the RTC. In March 2009, she assaulted a police officer and a teacher and was taken to a juvenile detention center for a night. Id. at DFPS009012378. In August 2009, she was placed in psychiatric care. Id. at DFPS009011739.

In April 2010, M.D. was again hospitalized, this time for “intensifying behavioral issues,” which coincided with a medication change and the incarceration of her brother. Id. at 1 RFP CPS 143178. In May 2010, the RTC gave CPS a 30-day discharge notice because it considered itself unequipped to handle her. Id. at DFPS009015051. M.D. was placed in another hospital, and then moved to a different RTC's Intensive Psychiatric Transition Program.¹⁵ Id. at DFPS009011755-59. While M.D. was in the Program, in July 2010, CPS terminated her mother's parental rights. Id. at DFPS009007915. Soon after M.D. learned of the termination, she became depressed and distanced herself from her peers, in contrast with her previously social, outgoing demeanor. Id. at DFPS009019597, 009019892, 0090132396.

15 An Intensive Psychiatric Transition Program is a 60-day program in an RTC designed to help children transition into a less restrictive environment.

In August 2010, CPS moved M.D. to another RTC. Id. at DFPS009011763. A month later, M.D. had two major incidents. She first attempted to pull pieces of the baseboard off the wall and hurt herself. Id. at DFPS009015702. Second, she attacked an elderly teacher and instigated what the RTC staff called a “riot.” Id. at DFPS009015730. In October, M.D.'s level of care was upgraded from Specialized to Intense. On October 31, 2010, she ran away from the RTC. When she returned the next day, she reported being sexually assaulted and using alcohol and marijuana while away from the RTC. (Id. at 1 RFP CPS 089285, DFPS # 901163, # 9011998, # 9012334; see also PX 2015 at 31-32 (filed under seal)). She was taken to Denton Regional Hospital for a rape kit. (DX 120 at DFPS009015798-99 (filed under seal)). M.D.'s case files do not indicate the results.

Once back at the RTC, M.D. alleged that a staff member physically abused her during a restraint. Id. at 1 RFP 1 RCCL 00611. A nurse at the RTC confirmed that M.D. had a bruise on her back following the incident. Id. M.D. also alleged that the same staff member sexually abused her and her roommate. Id. at 1 RFP 1 RCCL 00594-618. M.D. did not make the sexual abuse outcry until a year later, by which time the RTC had closed after being put on “Corrective Action/Evaluation” for improper “Medication Storage, Destruction, and Medication Records, Serious Incident Documentation ..., as well as de-escalation, and Post Discussions with children in care.” Id. at 1 RFP 1 RCCL 00597. The specific staff member had been fired for an unrelated event. RCCL Ruled Out (“R/O”) M.D.'s sexual abuse allegation “based on a preponderance of the information gathered” because “no outcries were made by the alleged victims” for a year. Id. at 1 RFP 1 RCCL 00604-05.

After another Intensive Psychiatric Transition Program, M.D. was moved to the Hector Garza RTC. Id. at DFPS009011834-36. In April 2011, she and her roommate accused another roommate of sexually abusing them at night between staff room checks. The allegations were Ruled Out because “staff conducted room checks every 15 minutes,” and “none of the children had any history of sexually acting out or were under any special supervisory precautions.” Id. at DFPS009013763-64. M.D.'s behavior deteriorated over the next two months. Id. at DFPS009012238. She had to be physically restrained multiple times, sustaining injuries during two of those restraints. M.D. told her caseworker that the RTC staff purposely provoked her, hoped to see her fail, and were excessively rough during the restraints. Id. at DFPS009012239.

During M.D.'s stay at Hector Garza, she and multiple other residents reported that one staff member often sexually harassed them. Id. at DFPS009014015-22. On one occasion in June 2011, M.D. saw the staff member place his hand underneath her roommate's bed covers while the roommate was in bed. Id. at DFPS009014015. Three days after an investigation was initiated, the staff member stopped coming to work. Id. at DFPS009014016. After he was fired for not appearing at work, the sexual harassment investigation was terminated without any findings. Id. at DFPS009014016-22. M.D. also accused a different staff member of physical abuse, but RCCL Ruled Out any physical abuse based on lack of “preponderance of the evidence.” Id. at DFPS009014023. The RCCL investigator based his finding in part on M.D.'s lack of bruising at the time of her interview and the Hector Garza staff saying that M.D. “would do anything to get moved out of the facility.” Id.

In August 2011, M.D. was arrested for assaulting a teacher. Id. at DFPS009012249. M.D. told the police that she preferred “to remain [in jail] indefinitely” rather than return to Hector Garza. Id. at DFPS009012252. She told the police that the Hector Garza staff used improper restraints, including placing their hands on residents' faces and throwing them to the ground. Id. at DFPS009093781. M.D. claimed her arms were bruised repeatedly from the restraints. She also said the staff ignored residents fighting amongst themselves. M.D. said she felt unsafe at Hector Garza. Id. As a result, Stukenberg, her next friend and attorney ad litem , scheduled an emergency hearing to change her placement. Id. at DFPS009014022. At the September 8 emergency hearing, the judge agreed that M.D. should not return to Hector Garza. Id. When M.D. was released from juvenile detention on September 12, she was sent to an RTC where she had previously lived for two months in 2010, and placed in its Intensive Psychiatric Transition Program. Id. at DFPS009011783.

On September 20, after only one week at that RTC, M.D. accused a staff member of sexually assaulting and raping her. Id. at DFPS009014041-42. A psychiatrist who interviewed M.D. the following week believed that she was telling the truth. Id. at DFPS009014132. M.D. also experienced extended vaginal bleeding for two weeks following the alleged incident, “even though her period had ended approximately three days prior.” Id. at DFPS009014133, 1 RFP 1 RCCL 00201. The day of the alleged incident, the staff member in question left work early, purportedly due to issues with his blood sugar. Id. at DFPS009014141. M.D.'s caseworker was aware that “sexual abuse [was] not new at the facility.” Id. at DFPS009014177. The police stressed that M.D. should have a Sexual Assault Nurse Examiner test “immediately” to aid a possible criminal investigation. Id. at DFPS009014151. Yet by the time a doctor saw M.D. on September 23, it was too late to administer the exam. Id. at DFPS009014149. The doctor performed a vaginal exam instead, and “did not observe any lacerations or tearing that would indicate a violent crime.” Id. The doctor observed vaginal bleeding but “concluded that to be a result of vaginitis.” Id. It appears that a rape kit was done, but the results were never revealed to M.D., her caseworker, or ad litem . Id. at 1 RFP 1 RCCL 00022, 00421-23. On September 26, the RCCL investigator called the RTC to retrieve M.D.'s clothes and linens for testing, but those items had been washed. Id. at DFPS009014147. The RCCL investigator did not find a “preponderance of evidence that abuse occurred,” but made an Unable to Determine (“UTD”) disposition because he could not definitively rule out the allegations. Id. at DFPS009014154. The staff member was offered his job back following the investigation, on the condition that he never work in the female unit, which he declined. Id. at DFPS009014137.

During the investigation, M.D. complained to Stukenberg and her primary caseworker that the RTC staff was “covering for” the staff member she accused of rape, and had begun to “retaliate” against her. Id. at DFPS009014177, 009014185. In particular, one female staff verbally harassed M.D., ordering M.D. to stop “spreading rumors” and “stop telling everyone about what happen [ed].” Id. at DFPS009014323-24. Beyond verbal abuse, that staff member allegedly physically abused M.D. during a restraint. Id. at DFPS009014177, 009014185. According to the intake report, the staff member's “inappropriate physical treatment” of M.D. and “threats caus[ed] emotional injury that results in observable impairment to [M.D.'s] psychological functioning.” Id. at DFPS009014177. RCCL Ruled Out these allegations because there was not a “preponderance of the evidence that abuse occurred.” Id. at DFPS009014338.

Besides being mistreated by the RTC staff, M.D. was subjected to inappropriate living conditions. When Stukenberg visited M.D. in October 2011, she observed that M.D. shared a trailer with nine other girls, sharing a single shower and toilet even though regulations prohibit more than eight girls to one shower or toilet. Id. at DFPS009014185. The trailer also smelled of mildew and urine. Id. Stukenberg observed several other violations, such as only one staff was assigned to the trailer, and there was a noticeable lack of training among all RTC staff. Id. Stukenberg also noticed multiple self-inflicted cuts on M.D.'s arms and legs—including a particularly deep one on her wrist—reportedly due to the stress caused by the rape and consequent retaliation. Id. M.D. had not received medical attention for these lacerations. Id. at DFPS009014177. Stukenberg reported to M.D.'s caseworker that M.D. was not adequately supervised and was denied medical care. Stukenberg told the caseworker, “If [the RTC] had been a parent the child would have been removed from the home immediately.” Id. at DFPS009014186.

RCCL investigated these allegations and determined that the RTC was deficient in its showers and toilets. Id. at DFPS009014300-01. RCCL ordered Technical Assistance for the RTC to meet legal requirements, but Ruled Out the allegations of neglectful supervision and medical neglect. Id. at DFPS009014304-05, 009014317. That disposition was based largely on the investigator not noticing a urine smell when visiting the trailer three days after the report. Id. at DFPS009014317. When the investigator interviewed M.D., she recanted her allegations of retaliation, neglect, and physical and verbal abuse against the staff. Id. at DFPS009014323. It is impossible to determine from the record whether M.D. recanted because the allegations were untrue, or because she wished to remain at the RTC, or believed recanting would stop the retaliation. During the investigation, M.D. informed Stukenberg that she liked this current RTC much more than Hector Garza and preferred to remain there. Id. at DFPS009014324-25.

Nonetheless, M.D. attempted to run away from that RTC in November 2011. She reported to her ad litem that she and another girl were punished for the runaway attempt by being forced to stand barefoot on milk crates for several hours and was told that her food portion would be reduced. Id. at 1 RFP CPS 000005. After that, M.D. had more self-harm incidents. Her case files, however, are relatively free of disciplinary incidents in 2012. See, e.g., id. at DFPS009012204. By February 2013, she had made sufficient progress at the RTC and was placed in a foster home (although her level of care remained Intense). Id. at DFPS009011791.

M.D. lasted less than a week at the foster home. She assaulted a ten-year old girl in the home for telling the foster parents that M.D. had a consensual sexual relationship with an adult male. Id. at DFPS009014351. M.D. was then placed in a psychiatric hospital for ten days. Id. at DFPS009011795. Upon release, she moved to a different foster home where she tried to run away daily. She also resumed harming herself, id. at DFPS009012094, which resulted in two trips to the emergency room and another psychiatric hospital in March 2013. Id. at DFPS009011803. Nonetheless, M.D.'s behavior had apparently improved enough that DFPS reduced her level of care from Intense to Specialized for the first time in two and a half years. Id. at 1 RFP CPS 142883.

After this hospitalization, the foster home would not accept M.D. back, and her caseworker needed to find her a new placement. She could not return to the RTC where she had been less than two months earlier and had done well because it only took Intense-level children. See id. at DFPS009011787. Instead, she was moved to the San Antonio State Hospital. Id. at DFPS009011803, 009011807. While there, M.D. reportedly engaged in two incidents of inappropriate sexual behavior with two female residents. DFPS investigated each allegation for inadequate supervision by the staff. The first allegation was deemed “not credible” and ruled “Unconfirmed,” the equivalent of Ruled Out. Id. at 1 RFP CPS 146463. The investigator confirmed the second allegation, but did not find a “preponderance of evidence” that the caretakers were guilty of inadequate supervision, resulting in an “Inconclusive” disposition, the equivalent of Unable to Determine. Id. at 1 RFP CPS 146304.

M.D. was discharged from the San Antonio State Hospital on April 17, 2013. She was sent back to the Hector Garza RTC despite her negative experiences there and her repeated pleas, as well as a previous court order, to not return. See, e.g., id. at DFPS009012252. At trial, Stukenberg testified that other RTCs did not want to take M.D. because she had previously made outcries while in care, which RTCs view as a liability. (D.E. 324 at 225). During her second stay at Hector Garza, M.D. admitted to abusing alcohol and marijuana. (DX 120 at 1 RFP CPS 177011 (filed under seal)). She was again subjected to inappropriate restraints. On May 18, a staff member allegedly verbally provoked M.D. while taking her back to her room and then pushed her into a wall. Id. at DFPS009012113. When M.D. resisted, the staff member restrained M.D. by placing her face-down on the ground and pressing her elbow on M.D.'s neck, making it difficult to breathe. Id. M.D. was bruised on her neck and had a six-inch scratch from the staff member's ring, which M.D.'s “I See You” worker¹⁶ observed during her visit two days later. Id. Shortly after the incident, an RTC staff member told M.D.'s primary caseworker that “no matter what M.D. did there is a proper way to restrain, just like in jail/prison guards have to act a certain way.” Id. at DFPS009012112. M.D.'s 33,000 pages of case files do not mention an investigation of this improper restraint, or of the staff member being cited for it. For the second time in two years, M.D. told her caseworker that she did not feel safe at Hector Garza. Id. at DFPS009012131. Her behavior noticeably deteriorated. Id.

16 I See You workers, as well as Centralized Placement Unit (“CPU”) workers, are secondary workforces that assist primary caseworkers. CPU workers do not interact with foster children; they only help find placements. I See You workers only visit foster children once a month when primary caseworkers are too busy or too far away. They often just show up at the child's placement, ask “five questions and leave.” See infra pp. 171-76.

On September 13, 2013, a court ordered CPS to return M.D. to her mother, who had been released from prison seven months before. Id. at DFPS009011998. In order to take custody of her daughter, M.D.'s mother needed to find appropriate housing. M.D.'s mother needed copies of M.D.'s birth certificate and social security card to complete a housing application. Id. at DFPS009012001-02. M.D.'s mother repeatedly contacted CPS to obtain those documents, but a new caseworker explained that she needed to speak to her supervisor. The Department eventually refused to give copies to M.D.'s mother, without explanation. Id. at DFPS009012001-10. Internal emails suggest that CPS did not want to return M.D. to her mother because the caseworkers did not believe that the mother could handle M.D. and that the mother was a “stressor” to M.D. Id. at DFPS009011626, 009012005. On September 24, 2013, Stukenberg requested an emergency hearing to get M.D. out of foster care. Id. at DFPS009011998, 009012002. M.D. learned that CPS was not cooperating with her mother, preventing M.D. from leaving Hector Garza. Id. On September 25, when her I See You worker visited, M.D. “appeared tired and was not her usual animated self,” and had reportedly attacked a peer and injured her own arm earlier that day. Id. On September 26, M.D. was taken off all of her psychotropic medications because she refused to take them and instead gave them to the other residents. Id. at DFPS009013482, 009012009.

On October 13, Hector Garza placed M.D. on close supervision because of her refusal to take medication. Id. at 1 RFP CPS 177042. On October 17, M.D. met her new primary caseworker for the first time and immediately asked about her old caseworker. Id. at DFPS009012007-10. M.D.'s previous caseworker had been one of the few constants in her life for the previous four and a half years. Although M.D.'s prior caseworker had not left CPS (she was promoted), no one notified M.D. of the turnover. M.D. told her new caseworker that she wished to live “anywhere but in this placement.” Id. On October 21, Hector Garza put in a 14-day discharge notice for M.D. On October 25, M.D.'s mother obtained an apartment, although CPS still believed that reunification was “not to [M.D.'s] benefit” based on M.D.'s observed behaviors following visits with her mother. Id. at DFPS009012013. Stukenberg continued to push for reunification, which disconcerted DFPS. See id. at DFPS009011998, 009012002. It is not clear from the record why CPS ignored a court order to reunite M.D. with her mother.

On October 27, M.D., upset because she had been unable to contact her mother that day, destroyed furniture in her room and threatened another resident. Id. at DFPS009012012. Staff also found a razorblade in her eyeglasses case. Id. On October 29, CPS decided to send her to a Houston RTC. Id. at DFPS009011815. She was moved on November 5. Id. at DFPS0091011819. During the move, M.D. showed her caseworker markings on her feet. Id. at DFPS009012013. She told her caseworker that she had not received any medical treatment. Id. She was not given any of her prescribed psychotropic medications at the new RTC because she had refused to take them at Hector Garza. Id. at DFPS009011680.

The Houston RTC was not told about M.D.'s history of running away and self-harming, or that she had been off her medication for 30 days. (D.E. 324 at 227). She ran away from the RTC within a week. Id. at 226-27. Soon after, the Houston Police Department found M.D. in a park and returned her to the RTC. Id. at 227. M.D. said that she had been “selling her body to men for money.” Id. Within two days of returning to the RTC, M.D. told Stukenberg that she could not “take it” anymore and that she wanted to go back home to Corpus Christi. Id. at 227-28. Stukenberg informed M.D. that she would do what she could and would request another emergency hearing, but upon hanging up the phone M.D. ran away again. Id. at 228. During M.D.'s short stay in Houston, she received medical and sexually transmitted disease (STD) tests. She tested positive for hypokalemia and pelvic inflammatory disease. (DX 120 at DFPS009012017). The latter condition is often caused by—and therefore an indication of—an STD. Id. at 1 RFP CPS 177109. M.D. was also tested for gonorrhea and chlamydia. The hospital never released the results of those exams to her caseworker or ad litem . Id. at DFPS009012017. As of January 2014, the caseworker was still trying to obtain those test results. Id. at DFPS009013507.

M.D. was found just before Thanksgiving in 2013. She was taken to a Houston CPS emergency shelter, where she told the staff that she was “high on crack” and continuously said that she wanted to go back to a man who was “someone that makes money off of her by having her sleep with other boys/men.” (Id. at # 1 RFP CPS 175293 (filed under seal); D.E. 324 at 228). CPS staff allowed M.D. to leave within 15 minutes of her arrival at the shelter. (DX 120 at 1 RFP CPS 175293 (filed under seal)). Only after that incident did M.D.'s case files state, “If she is found or returns she will need a secure facility to prevent further run away's [sic].” Id. at 1 RFP CPS 175295. The Houston Police Department handled M.D.'s case as a sex trafficking case. Id. at 1 RFP CPS 175308. In the summer of 2014, a national missing children organization informed DFPS that M.D. may have been arrested by the police and that they might have her address. (D.E. 324 at 229). M.D.'s caseworker cautioned Stukenberg not to get her “hopes up” because DFPS was not sure it was M.D., but upon seeing a picture, Stukenberg immediately recognized M.D. Id. M.D.'s caseworker apparently did not “even know what her client look[ed] like” because neither the caseworker nor anyone else in her unit had ever met M.D. Id.

Despite having a possible address for M.D., the State declined to retrieve her. DFPS refused to pick up M.D. without police assistance. In an appalling Catch-22, the Houston police department would not pick M.D. up until DFPS found a placement for her, and DFPS could not find a placement until M.D. was in State custody. Id. at 230. Stukenberg urged M.D.'s caseworker to go to her. Id. After approximately two weeks, CPS went to find M.D., but the person who answered the door said that M.D. was no longer there. Id. M.D. has not been seen nor heard from since. Id. On January 9, 2015, M.D. turned 18 and aged out of foster care.

When M.D. entered State PMC in January 2008, she was already in an RTC. (DX 120 at DFPS # 31586 (filed under seal)). From that date until she ran away in November 2013, M.D. spent almost all of her time in RTCs. See, e.g ., id. at DFPS # 28393-97, 28403, 28419-34. For the seven years she was in PMC, M.D. spent only 14 days in foster family homes. DFPS009007116-17, 009011791, 009011799. Defendants' and Plaintiffs' experts agreed that sexually abused children should live in single-child placements. (See D.E. 328 at 161; D.E. 329 at 161; D.E. 303 at 8). Yet neither of M.D.'s foster homes in 2013 were single-child placements, even though she had been sexually abused.

When the State assumed custody of M.D. in April 2007, her level of care was Basic. Id. at DFPS009007116. Within four months her level of care rose to Specialized. Three months later, her level of care rose to Intense. Id. Except for three months (August-October 2010, March 2013), M.D.'s level of care remained Intense over her eight years in DFPS custody. See id. at DFPS009007116-17, 009020068, 1 RFP CPS 142883.

Coinciding with M.D.'s rising levels of care, M.D. was steadily diagnosed with psychological disorders and prescribed a litany of medication. In November 2007, after seven months in foster care, M.D. was diagnosed with Bipolar Disorder and Post-Traumatic Stress Disorder (“PTSD”). Id. at DFPS009017082. In April 2008, M.D. was diagnosed with Attention Deficient Hyperactive Disorder (“ADHD”) and Oppositional Defiant Disorder. Id. at DFPS009017394. In 2010, Impulse Control Disorder and Borderline Personality Disorder were added to the list. Id. at DFPS009014552. In 2011, Schizophrenia, Mood Disorder, and Conduct Disorder. Id. at DFPS009014040. In 2013, Major Depressive Disorder, Mathematics Disorder, and Disorder of Written Expression. Id. at 1 RFP CPS 142880.

M.D. was prescribed a pharmacy of psychotropic medications to treat these diagnoses, including Adderall, Intuniv, Clonidine, Concerta, Vyvanse, Tenex, and Focalin¹⁷ for ADHD; Depakote and Divalproex¹⁸ for Bipolar Disorder ; Fluoxetine (or Prozac ), Celexa, and Trazodone¹⁹ for Depression; Abilify, Risperdal, and Seroquel²⁰ (all anti-psychotics) for Schizophrenia, Depression, and Bipolar Disorder ; and Trileptal²¹ for mood disorders. She often took five or six medications at a time. (PX 2015 at 34, 49 (filed under seal)). According to Carter, M.D.'s problems are likely not solvable by medication. In his view, the State over-relied on psychotropic drugs and overmedicated M.D. Id. at 49.

17 All of these drugs are used to treat ADHD or ADD. Adderall is also used to treat narcolepsy ; Clonidine and Tenex are also used to treat high blood pressure. Possible side effects of Adderall include bladder pain, bloody or cloudy urine, fast irregular heartbeat, stunted growth, trouble breathing, aggression, seizures, numbness, muscle twitches, changes in vision, unexplained muscle pain, stomach pain, weight loss, mood changes, nervousness, fast heart rate, headache, dizziness, insomnia, and dry mouth. http://www.drugs.com/adderall.html (last viewed on July 27, 2015). Side effects of Intuniv include mild dizziness or drowsiness. http://www.drugs.com/intuniv.html (last viewed on July 27, 2015). Side effects of Clonidine include urination problems, confusion, hallucinations, feeling tired and irritable, mood changes and sleep problems. http://www.drugs.com/clonidine.html (last viewed on July 27, 2015). Side effects of Concerta include new or worsening symptoms such as mood swings, aggression, hostility, or changes in personality or behavior, panic, delusion, extreme fear, hallucinations, unusual behavior, motor tics, irritability, and insomnia. http://www.drugs.com/concerta.html (last viewed July 27, 2015). Side effects of Vyvanse include confusion, paranoia, unusual thoughts or behavior and hallucinations. http://www.drugs.com/vyvanse.html (last viewed on July 27, 2015). Side effects of Tenex include chest pains, lightheadedness, dizziness, drowsiness, constipation, weakness, blurred vision, and confusion. http://www.drugs.com/tenex.html (last viewed on July 27, 2015). Side effects of Focalin XR include headache, upset stomach, trouble sleeping, anxiety, and nervousness. The label contains a warning to tell a doctor about any mental problems a child may have, or about a family history of suicide, bipolar illness, depression, new or worse behavior or hostility and or new psychotic symptoms or new manic symptoms. http://www.drugs.com/focalin.html (last viewed July 27, 2015). Most of these medications should not be given to children under the age of six. Adderall, Focalin XR, Concerta, and Vyvanse are all central nervous system stimulants that can be abused or lead to dependence.

18 Depakote and Divalproex are used to treat seizure disorders and bipolar depression. Common side effects include liver and pancreas problems, stomach pain, nausea, vomiting, headaches, drowsiness, dizziness, problems with balance, blurred vision, dark urine, jaundice, mood changes, depression, anxiety, panic attacks, insomnia, irritability, hyperactivity, tiredness, confusion, delusions, nose bleeding, easy bruising, purple or red spots on skin, seizures, inflammation, and bleeding gums. http://www.drugs.com/depakote.html (last viewed on July 27, 2015). Side effects of Divalproex include stomach pain, nausea, vomiting, headaches, dizziness, and blurred vision. http://www.drugs.com/divalproex-sodium.html (last viewed July 27, 2015).

19 Fluoxetine, Celexa, and Trazodone are all antidepressants. Fluoxetine is only supposed to be used after two other medications have failed to treat depression. Common side effects include insomnia, strange dreams, headache, dizziness, vision changes, tremors, anxiety, nervousness, upset stomach, nausea, vomiting, diarrhea, dry mouth, flu symptoms, and restlessness. http://www.drugs.com/fluoxetine.html (last viewed July 27, 2015). Common side effects of Celexa include drowsiness, dizziness, weakness, anxiety, shakiness, insomnia, vision changes, nausea, loss of appetite, diarrhea, constipation, dry mouth, cold symptoms, sweating, and weight changes. http://www.drugs.com/celexa.html (last viewed July 27, 2015). Side effects of Trazodone include drowsiness, dizziness, vision changes, lightheadedness, and dry mouth. http://www.drugs.com/trazodone.html (last viewed July 27, 2015).

20 All three drugs are used to treat schizophrenia in children (generally over the age of 13) and adults. Abilify is also used to treat bipolar disorder and major depressive disorder in adults and to treat irritability, symptoms of aggression, mood swings, and temper tantrums related to autism in children who are at least six years old. Side effects include severe agitation, distress, or restlessness. http://www.drugs.com/abilify.html (last viewed on July 27, 2015). Risperdal is also used to treat symptoms of bipolar disorder in adults and symptoms of irritability in autistic children. Possible side effects include aggressive behavior, agitation, anxiety, restlessness, depressed mood, inability to concentrate, increase in amount of urine, constipation, diarrhea, and trouble sleeping. http://www.drugs.com/risperdal.html (last viewed July 27, 2015). Side effects of Seroquel include dizziness, drowsiness, and loss of energy. http://www.drugs.com/seroquel.html (last viewed July 27, 2015).

21 Trileptal is an anticonvulsant, or antiepileptic drug that works by decreasing nerve impulses that cause seizures. Side effects include dizziness and drowsiness. People with depression and mood disorders are advised to tell their doctors before taking Trileptal. http://www.drugs.com/trileptal.html (last viewed on July 1, 2015).

M.D.'s numerous placement changes caused absences and disruption in her schooling. M.D.'s academic performance declined, mostly due to her disruptive classroom behaviors, but also because she missed school time with every placement change and psychiatric hospitalization. Carter evaluated M.D. in October 2012, when she was 15 years old. He found that her overall academic abilities were on the third to fifth grade level. M.D. told Carter that the very thought of attending a normal school made her nervous because virtually all of her education to that point had been within RTCs. (PX 2015 at 28 (filed under seal)). M.D. passed the ninth and tenth grade in special education classes. (DX 120 at DFPS009007110, 009007119 (filed under seal)). She was placed in 11th grade in August 2013 at a charter school at her RTC, but refused to attend. She was moved to a public high school when her placement changed in November 2013, but only attended one day before she ran away. Id. at DFPS009007120. The school was still looking at her records to determine which grade level she would be placed. M.D. later told her caseworker that one reason she ran away was her discouragement at being placed in the ninth grade again. Id. at DFPS009007110.

A review of M.D.'s stay in PMC exemplifies many of the problems that Named Plaintiffs alleged about Texas foster care. On multiple occasions M.D. alleged physical, sexual, and verbal abuse. Many allegations received cursory reviews, with RCCL terminating investigations without gathering important facts. Other allegations were not investigated at all. M.D.'s living conditions often violated DFPS regulations. Because of Texas's inadequate placement array, M.D. returned to the Hector Garza RTC in 2013. An untimely caseworker turnover coincided with her return to Hector Garza. In addition, DFPS failed to reunite M.D. with her family despite a court order to do so, which contributed to her running away. When M.D. was found, miscommunications between her ad litem , her new caseworker, her shelter home, and the Houston Police Department allowed her to run away for good. Unfortunately, as Carter testified and the Court so finds, M.D.'s experience is “typical ... of the entire foster care system in the State of Texas,” especially in the PMC. (See D.E. 326 at 132-33, 194).

b. D.I.

D.I. is a class representative for the General Class and the Licensed Foster Care Subclass. At the time of trial, D.I. was 12 years old. (D.E. 215 at 3). He became involved with DFPS in 2007 at age six when DFPS received reports of neglect by his mother, a drug addict. D.I. ran away from home in 2008, prompting DFPS to institute a safety plan, though he remained 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 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 behavior. (PX 185 at 4236 (filed under seal)). In April 2009, D.I. received a psychological evaluation. Id. at 2202. The psychologist found that D.I. was having “emotional and behavioral problems, both, at a level of mild clinical significance” and that he displayed “highly significant aggressive behaviors, attention problems, conduct problems and social problems and significant problems related to rule-breaking behavior, ADHD symptoms, oppositional defiant disorder symptoms and mildly significant affective/depressive symptoms and anxiety problems.” Id. at 2208. D.I. denied experiencing suicidal or homicidal ideations, though, and did not appear to experience auditory or visual hallucinations. Id. at 2207.

D.I. entered the State's PMC on March 8, 2010 at a Basic-level of care. Id. at 1968, 2122. In June 2010, he was placed by the Bair Foundation, a CPA, in a foster group home. Id. at 417. In his first month there, two older boys in the home, aged 16 and 17, sexually abused him on at least three occasions involving oral and anal sex. He was eight years old. Id. at 28-29. Each incident of abuse occurred at night in the boys' bedroom, while the sole caregiver was asleep on an air mattress in the downstairs living room. Id. at 56. The caregiver in the foster group home was unaware of this sexual abuse until one of the older boys confessed. Id. at 48. The 16-year-old who initiated the abuse had entered foster care due to being sexually abused by his father. Id. at 54. In 2007, he had sexually abused a younger child with special needs in a different foster home. Id. The CPA's case files contained information about the 16-year-old's history of being sexually abused but made no mention of him abusing others. Id. at 3, 54.

During the investigation into this incident, the 16-year-old's primary caseworker told the RCCL investigator that the CPA “should have been” aware of the 2007 incident and that “this incident was similar to the previous incident.” Id. at 54. The caretaker informed the investigator that he knew the 16-year-old had been abused by his father and that he had concerns that D.I. and the 16-year-old would “play together so much, since the two children have a significant age difference.” Id. at 50, 54. Further, CPS notes state that the 17-year-old, “is a homosexual and has regular HIV tests” and that therefore “new testing will be done on all three boys.” Id. at 18. There is no mention in D.I.'s case files of HIV testing.

D.I.'s caseworker told the RCCL investigator that while D.I. never made a sexual abuse outcry, he had previously made a joke at school of a sexual nature, which caused her to be concerned that D.I. may have been sexually abused. Id. at 55. According to D.I.'s caseworker, after spilling ketchup at school D.I. said, “this is what happens to my butt.” Id. The caseworker she spoke to D.I. about that statement, and told her he was only joking. Id. at 55. D.I.'s case provides an example of DFPS mixing ages and not providing 24/7 care.

Following the investigation, RCCL confirmed that the sexual encounters occurred, but Ruled Out all allegations of neglectful supervision and sexual abuse. Id. at 58. The investigator noted several factors that led to a Ruled Out disposition as to neglectful supervision, including: “The sexual encounters between the children most likely occurred late at night when the children were believed to be asleep”; the caregiver “was likely asleep at the time of the sexual encounters”; the caregiver “reported that for the month of June he slept on an air mattress in the downstairs living room so that he could provide increased supervision for a child who was a runaway risk”; “None of the children in the home were known by The Bair Foundation to have a history of sexually acting out with one another or of sexually abusing other children”; and the caregiver “did not have any information that would have led him to believe that the children were engaging in sexual activity.” Id. The investigator also Ruled Out the sexual abuse allegations, despite confirming that the incidents indeed occurred, because there was “not a preponderance of the evidence that abuse occurred.” Id. The RCCL supervisor agreed with the investigator's disposition decisions, commenting: “[Neither] Foster parent nor CPA had knowledge that either of the children involved had [history] of [sexual] acting out, even though [illegible] reported that [16-year-old] had an incident of [sexual] acting out in a previous [placement].” Id. at 5. This appears to be a set pattern of DFPS's investigations of abuse in care. First, DFPS refuses to document or recognize child-on-child abuse. (See PX 1026 at 5; D.E. 300 at 47-48; D.E. 303 at 6-8). Second, even when a child is abused in foster care, the standards for finding that abuse occurred are impossibly difficult. It appears that, unless witnessed by an adult who admits seeing the abuse and thereafter refuses to do anything about it, the preponderance of evidence standards cannot be met. Children are interviewed and discounted. Adults are similarly discounted. The disposition of the sexual abuse of D.I. as Ruled Out is based on DFPS determining that child-on-child abuse findings would result in labeling one child a perpetrator. (PX 185 at 58 (filed under seal)). Certainly each child's caseworker should have been aware of the sexualization of these three boys and, as both Defendants' and Plaintiffs' agree, immediately placed them in single-child placements. (Id ; see also D.E. 328 at 161; D.E. 329 at 161; D.E. 303 at 8). Instead, the only action recommended in the investigation report was “[r]outine monitoring.” This Dickensian approach to child abuse keeps the confirmed reports of abuse in care findings artificially low.

The Bair Foundation instituted an identical safety plan for the 16 and 17-year-old, requiring the caregiver to: have each of the youths within hearing and seeing distance at all times during waking hours; check on each of the youths at least 3-4 times each night; and prohibit the youths from sharing a room with other children in the home. Id. at 104. Six months after the incident was reported, the same RCCL investigator completed his investigation of the Bair Foundation, finding no minimum standard deficiencies. Id. at 66-67. Specifically, the Bair Foundation was found to be in compliance with the following standards: “Adhere to the child's rights to be free of abuse, neglect, and exploitation as defined in Texas Family Code 261.401 ”; “The caregiver is responsible for being aware of and accountable for each child's on-going activity”; and “The caregiver is responsible for ensuring each child's safety and well-being including auditory and/or visual awareness of the child.” Id. at 67. Neither the Bair Foundation nor the foster group home was penalized or admonished in connection with this incident.²² DFPS instituted no internal changes after this incident.

22 The Bair Foundation had its license placed on probation the following year for unrelated reasons. (PX 978 at 1).

After that incident, D.I. was removed to an emergency respite placement and then to another foster home—still not a single child placement. Id. at 109. D.I. received a psychosocial assessment on September 8, 2010, which inexplicably stated that he was removed from his previous placement “due to problems getting along with the older foster kids in the home.” Id. at 2501. The assessment made no mention of D.I.'s sexual abuse history. See id. at 2501-02. Two weeks later, D.I. was hospitalized for a week “due to behaviors and suicidal ideation.” Id. at 2489, 2492. On D.I.'s Common Application for Placement of Children in Residential Care form,²³ filled out by his primary caseworker in November 2010, more than four months after he was abused, marks his sexual abuse history as “unknown.” Id. at 384. On that same form, the caseworker noted, “Recently child was involved with a problem at a previous foster home that included participating in sex with other children in the home.” Id. at 382.

23 Foster children's Common Applications contain detailed information about the children and are provided by DFPS to a CPA or residential facility to allow the CPA or facility to determine whether they can meet the child's needs. Before placing a child with Moderate, Specialized, or Intense service needs, a CPS caseworker must provide the child's full Common Application to the facility. (PX 50 at 80). When placing a child with only Basic needs, or in case of an emergency placement, the caseworker need only provide an abridged version of the Common Application at the time of the placement, but must provide the full form within 30 days of the placement. Id. at 81. Although a facility may complete the Common Application on its own, it is ultimately the CPS caseworker's responsibility to ensure its completeness and accuracy. See id.

D.I. received another psychological evaluation in December 2010. Id. at 2482-91. That evaluation was significantly worse than his 2009 evaluation. D.I. had an elevated score on the Anxiety Scale (94th percentile) and the Depression Scale (96th percentile). Id. The psychologist noted that the clinically elevated anxiety score “may reflect the presence of an anxiety disorder” and may “reflect posttraumatic fears associated with previous victimization.” Id. During the evaluation, D.I. endorsed that “ ‘lots of times' he wants to hurt himself and that ... ‘lots of times' he wants to kill himself.” Id. Further, the evaluation states under the heading “Notable Item Responses,” that D.I. responded “true” to many troubling statements, including:

(8) I have thought about killing myself , ... (12) I hear voices that others don't hear , ... (22) I keep thinking about something terrible that happened to me, ... (37) I feel very, very sad, (38) People say I see things that aren't there , ... (41) There are voices in my head that tell me to do silly or bad things , ... (46) Things in my life go from bad to worse, ... (52) I have tried to hurt myself , ... and (54) I am afraid of some of the ideas I have in my head .

Id. at 2487.

Once again, there is no indication that D.I., who was repeatedly sexually abused while in foster care, and became sexualized as a result, was ever placed in a single-child home. In fact, the department did everything it could to cover up that D.I. was sexually abused, from ruling out sexual abuse allegations, to referring to his rape in his case files as “problems getting along with the older foster kids,” “sexual behavior,” “a sexual act with some boys,” and “participating in sex with other children in the home.” Instead of placing D.I. in a single-child home, CPS placed him next in a home with another child who had a sexual abuse history. Id. at 193. That child accused D.I. of attempting to have sex with her although she later admitted to her therapist that she had tried to initiate sexual contact with D.I. but was not able to do so. Id. Neither child was removed from that placement.

On October 24, 2012, D.I. was adopted. Id. at 2883. D.I. spent less time in PMC than any of the other Named Plaintiffs. Over his three years in care, D.I. had two primary caseworkers, although he did have the same caseworker during his entire time in PMC. See id. at 659-74. He also only changed schools once during that time. Id. at 1983-84. In early 2012, D.I. was in the third grade and passing all of his classes, although he had behavior problems and often disrupted class or tried to distract other students. See id. at 2058, 2096, 2100, 2111. Since being adopted, D.I. has engaged the five-year-old grandchild of his adoptive mother in sexually inappropriate conduct on more than one occasion, mostly recently in March and August 2014. (D.E. 326 at 113; DX 262 at 57 (filed under seal)). D.I. was placed in the Behavioral Hospital of Bellaire from November 18 to December 9, 2014, for “aggression.” (PX 2171 at 2-3 (filed under seal)). According to the Bellaire discharge report, dated December 9, 2014, D.I. was diagnosed as a sexual abuse victim and perpetrator. Id. at 2. Carter found that D.I. is still affected by the injury that he suffered while in Texas's PMC. (D.E. 326 at 190-91). Carter testified that D.I. “continues to be a very disturbed boy,” is “heavily sexualized,” and “a high risk for sexually harming children.” Id. at 191. Unfortunately, as Carter testified and the Court so finds, D.I.'s experience is “typical ... of the entire foster care system in the State of Texas,” especially in the PMC. See id. at 132-33, 194.

c. S.A.

S.A. is a class representative for the General Class and the Licensed Foster Care Subclass. At the time of trial, S.A. was 18 years old and living in a shelter after aging out of PMC. (See DX 120 at DFPS009035700-10 (filed under seal)). She is one of the 1300-1400 children who age out of PMC each year. She first came into Texas foster care at the age of five when her adoptive mother was arrested on outstanding warrants trying to cross into Mexico with her boyfriend, a registered sex offender.²⁴ See id. at DFPS # 48573, # 48597. Within four months of entering foster care, she made her first sexual abuse outcry, reporting anal penetration by an older male foster child. (See D.E. 326 at 192; D.E. 325 at 23-24; DX 120 at DFPS008344565, 00903674, 1 RFP CPS 155128 (filed under seal)). After this abuse was reported, RCCL's only response was to request that the Bair Foundation (the CPA that placed S.A. in that foster home) complete an internal investigation. (DX 120 at DFPS008344565 (filed under seal)). This investigation was labeled “Priority 3,” meaning that it related to “minor violations of the law or minimum standards that involve low risk to children.” Id. According to that report, the CPA performed an internal investigation but “did not cite any non-compliances” and planned no further action. Id. at DFPS008344566. It is hard to imagine that a five-year old could have any frame of reference to fabricate anal penetration. There is no evidence in S.A.'s 33,000 pages of case files that she was interviewed regarding these allegations or that she received any type of physical examination or medical treatment in connection with these allegations. It does not appear that RCCL conducted any independent investigation or follow-up in regard to these allegations. Because RCCL allows CPAs to keep their own files, the records of the CPA's internal investigation are not in her IMPACT²⁵ or External²⁶ case files. Therefore, no subsequent caseworker would have this information.

24 The record is unclear whether S.A. was sexually abused before coming into foster care in Texas.

25 IMPACT, or Information Management for the Protection of Adults and Children in Texas, is an automated system, included in children's case files, in which DFPS staff record casework related activities. DFPS Handbook § 1181, http://www.dfps.state.tx.us/handbooks/cps/files/CPS_pg_1170.asp (last visited December 14, 2015).

26 External case files contain supporting case documentation that cannot be entered or uploaded into the IMPACT system, including paper documents, spreadsheets, photographs, computer disks, audio tapes, and video tapes. DFPS Program Handbook § 5212, http://www.dfps.state.tx.us/handbooks/aps/files/APS_pg_5200.asp (last visited December 14, 2015).

Although the CPA Ruled Out abuse based on its internal investigation, S.A.'s foster parents asked that she be removed “to ensure her safety and [prevent] any possible contact between her and the other child.” Id. DFPS # 49623. Approximately a month later, S.A. was moved to an emergency shelter in an RTC. Id. at DFPS # 49624-5. After S.A.'s sexual abuse outcry, she began exhibiting frequent, uncontrollable, and “extreme tantrums where she tried to hurt herself and others.” Id. at DFPS # 49622, # 48997. A week after the outcry, S.A. was hospitalized due to those behaviors. See id. at DFPS008344565. Throughout her time in care, she repeatedly reported being uncomfortable around and distrustful of older, especially male, children, and male adults. See, e.g. , id. at DFPS # 49435-36, # 49438-40, # 49464. S.A told her caseworker that she did not trust CPS “to keep her safe.” Id. at DFPS # 49435. There are mentions throughout her records that she received therapy to address past sexual abuse. S.A. entered PMC at the age of six in late 2002, already exhibiting “sexual acting out,” which was graded as “extreme.” Id. at DFPS009037819.

Expert and fact witnesses for both parties testified that once children are sexually abused, or “sexualized,” that behavior is ongoing and destructive to themselves as well as to the other children with whom they come in contact. (See D.E. 299 at 36, 70-71; D.E. 303 at 8; D.E. 304 at 51; D.E. 326 at 15-16, 105-06, 113-15, 162-63; D.E. 328 at 161). As discussed infra , because DFPS does not track child-on-child abuse in an aggregate or centralized manner, it is possible that these children will again be placed with the same children who participated in their earlier sexualization. See Section IV.B.1. Although Defendants' and Plaintiffs' experts agreed that sexually abused children should live in single-child placements, none of S.A.'s 13 foster homes was a single-child placement. (See D.E. 328 at 161; D.E. 329 at 161; D.E. 303 at 8).

S.A. is one more child that entered the foster care system at a Basic-level of care and exited from an RTC some 13 years later, severely damaged. She had over 45 placement changes and lived in at least 33 different placements throughout the State. (D.E. 325 at 21-22; see DX 120 at DFPS009030795-96 (filed under seal)). Many of these placements were in psychiatric hospitals where she was restrained and heavily medicated. Several of these placements were inappropriate, some lasting as little as two days. She was shuttled between different placement types including foster homes, restrictive RTCs, and emergency shelters. The reason for moving S.A. was generally a decreased level of care or displaying aggressive and self-abusive behavior. (See DX 120 at DFPS # 48435-568, # 48993-99, # 49002-21, # 49495, DFPS009030795-96 (filed under seal)). Because of the significant amount of time that she spent in RTCs, at the age of 17 S.A. did not have “basic adult living preparation skills,” and did not know how to wash dishes or set a table. Id. at DFPS009035666.

S.A.'s longest stay at a placement was at the Hill Country Youth Ranch (“HCYR”) RTC in Ingram, Texas, where she was placed in December 2003. Id. at DFPS # 49008, # 48716. In July 2004, her service plan indicated that DFPS would “look at placing [S.A.] in another RTC since she has made minor improvement and the current facility staff feel there is no more that [S.A.] can benefit from their program.” Id. at 1 RFP CPS 177486. However, the department did not remove her from that RTC until nearly three years later, in April 2007, because her level of care was reduced. Id. at DFPS # 49103. By then the placement change appears to have been contrary to S.A.'s needs. By 2007, S.A. had made remarkable progress socially, emotionally, and academically at the RTC and appeared to finally have some stability in her life. See id. at DFPS009043113-14, 009043119-20. Although her improvement was significant, multiple professional staff members at the RTC, including the school director, believed that she should not have been removed when she was, especially in the abrupt manner in which CPS conducted the removal. See id. In letters to a Cameron County district judge, the staff members said they were “shocked and saddened by [S.A.'s] sudden removal” and admonished CPS for their failure to fully consider the propriety of the removal. Id. In her letter, the RTC school director wrote:

I am unclear as to how the decision to remove [S.A.] was made, and I do not know who made it or the criteria used to reach this judgment. I was informed of her impending removal this morning immediately before her departure. In the midst of many tears and much confusion, we attempted to say our goodbyes and to provide her with some measure of support and comfort.

It was and still is painful to think that this child was removed from a positive and fruitful program that was clearly advantageous to her.

I truly have no words to describe the depth of my objections to this decision, as well as my disappointment at the manner in which her removal was done.

This child may have been well-served if the person or persons making placement decisions would have sought and obtained objective data and information from the staff who were effectively working with her.

Sadly, her amazing struggles, effort and eventual success were dismissed and disrespected in this life-changing decision.

Id. at DFPS009043119-20 (emphasis added). One of her former instructors stated that S.A. had made much progress after a very difficult beginning and was not “ready to be taken from here.” Id. at DFPS009043114. The instructor further stated that she was disappointed that S.A. was being removed just as she was “showing improvement, and ... starting to blossom.” Id.

After this removal, S.A. was placed in a therapeutic foster group home that lasted only two days because she became “violent and defiant.” Id. at DFPS # 48716. She was removed to another therapeutic foster home where she remained for 11 days before being admitted to a psychiatric hospital for self-abusive and suicidal behavior. Id. While in this hospital a psychologist evaluated S.A. and noted that she had become “too aggressive and assaultive and was considered to be dangerous to the other children in the home” and that “she had a history of not getting along with other children” in foster homes. Id. at DFPS # 52873-74. That psychologist determined that she was “unable to live in a community setting” and recommended she be placed in an RTC to meet her level of care, which was up to Specialized. Id. at DFPS # 52876-77. In spite of the psychologist's recommendations, S.A. was again placed in a therapeutic foster home with other children. Id. at DFPS # 48810, # 49500. This placement lasted six days, as she became confrontational, taunting, and negative, particularly toward the other children in the home, and ended with her readmission to a psychiatric hospital after an episode of uncontrollable yelling, screaming, and crying. Id.

S.A. had many other inappropriate placements throughout her time in care. For example, after several failed foster home placements, S.A. was placed in an RTC in July 2007. In September 2007, S.A.'s I See You caseworker was concerned about S.A.'s continued restraint pattern and that the RTC staff did not appear to be “dealing with her behaviors very well.” Id. at DFPS # 49505-06. Earlier that same month her caseworker observed and reported that S.A. was being inappropriately restrained in that RTC when a staff member bent and pushed S.A.'s arms backward to the point that the caseworker was “concerned that her shoulders/arms would be pulled out of [their] sockets or her elbows broke[n].” Id. at DFPS # 49507. The caseworker noted that she thought the RTC was unable to provide for S.A.'s needs at that time and an RCCL investigator who visited S.A. concluded that she did not belong at that RTC. Id. There is nothing in S.A.'s files, though, about the outcome of that investigation or whether the offending staff member was ever admonished for the improper restraint. S.A. was not removed from that RTC until December of that year. See id. at DFPS # 49045, # 49511-18.

As another example of an inappropriate placement, in mid-2010 S.A.'s psychological evaluation stated that, because she was “vulnerable to self-destructive and aggressive behaviors,” she required “ongoing placement in a locked facility to help manage her mood and behavior.” Id. at DFPS009041897. Yet, two months later, she was placed in a therapeutic foster home. Id. at DFPS # 49087. After approximately four months, which included almost daily emotional breakdowns and two psychiatric hospitalizations, she was moved to a new RTC. See id. at DFPS # 49087-94, # 49403-13. Finally, in October 2011, she was returned to HCYR RTC where she had previously been successful, but her evaluating psychologist noted that she was “increasingly aggressive especially toward young children” and habitually tried to hurt herself and others. Id. at 1 RFP CPS 006986. This psychologist recommended that she be placed either with a family trained in appropriate techniques without any other children or an RTC “where there is a high level of restriction, rules, consequences, structure, routine, emotional support, modeling, praise, and help with academic, developmental, and psychiatric needs.” Id. at 1 RFP CPS 006992. This psychologist recommended that DFPS make every effort “to NOT move [S.A.] again before she is an adult.” Id. Nonetheless, a week later S.A. was moved to a different RTC for two weeks, whereupon she was placed in another psychiatric hospital. Id. at 1 RFP CPS 177322-23. In July 2012, she was moved to a therapeutic foster home that had other foster children. Id. at 1 RFP CPS 177980, 1 RFP RCCL 03635.

Along with the many placement changes, S.A. experienced a revolving door of caseworkers. She had approximately 28 different primary and secondary caseworkers. See id. at DFPS009033366-942. Her records reflect that she was often upset to learn she had a new primary caseworker, and usually her previous caseworker left without giving her notice. See, e.g. , id. at DFPS # 49437, # 49439. During her first few years in foster care, her caseworkers visited her in person only four times a year. See id. at DFPS # 48689, 1 RFP CPS 177485. In 2005, an I See You worker began to visit S.A. on roughly a monthly basis. S.A. would sometimes go several months without any contact with her primary caseworker and would complain about this to her various I See You workers. See, e.g., id. at DFPS # 49386, # 49511, # 49516, # 49519. At least once when her primary caseworker changed without any notice, S.A. continued to try to contact the former caseworker without any response. Id. at DFPS # 49519. Eventually, after months without contact from her primary caseworker, one of her I See You workers checked IMPACT and discovered that her primary caseworker had changed. Id.

Throughout her time in care, several possible adoptive placements never materialized, in part due to significant paperwork delays. See, e.g. , at DFPS # 49445-61, # 49535. The most common obstacle in placing S.A. with potential adoptive families was her caseworkers' failure to update her health, social, educational, and genetic history report (“HSEGH”). This report must be shown to prospective adoptive parents before any adoption can go forward. (See id. at DFPS # 49123; PX 1964 at 22). The delays appear largely attributable to a frequent turnover in her primary caseworkers. (See DX 120 at DFPS # 49450 (filed under seal)).

In May 2005 and again in February 2009, delays in updating her HSEGH derailed several possible adoptive placements. See id. at DFPS # 48694, # 49445-61, # 49535. S.A. experienced a significant amount of worry and anxiety throughout the potential adoption delays, and the continued failure to update the reports was particularly devastating for her. See id. at DFPS # 49436-41, DFPS # 49447-48. In spite of this, she continued to hold out hope of being adopted. See, e.g. , id. at DFPS # 49366-67, # 49475, # 49491. In 2011, a psychologist noted in his evaluation of S.A. that her prior failed adoptions was one of her major psycho-social stressors and that her behavior became increasingly worse after this. Id. at 1 RFP CPS 006986-90.

S.A. significantly deteriorated while in care, as reflected in her psychological evaluations. Shortly after entering State custody, she received an initial psychological screening and was provisionally diagnosed with Disruptive Behavior Disorder, but no learning disorders, ADHD, or PTSD. Id. at DFPS # 52311, # 52321. She received a full psychological evaluation in early 2002, which found that her intellectual functioning was approximately “in the upper end of the Low Average range,” which was likely an “underestimate of her actual intellectual level of functioning.” Id. at DFPS00941860. At that time, she was diagnosed with ADHD and Intermittent Explosive Disorder. Id. at DFPS00941861. In her psychological evaluation in late 2003, she was diagnosed with Bipolar Disorder, mixed with psychotic features, ADHD, Reading Disorder, Mathematics Disorder, and Disorder of Written Expression. Id. at DFPS009041870. By early 2004, S.A. was additionally diagnosed with PTSD and Reactive Attachment Disorder. Id. at DFPS009043286. Her psychosocial stressors included “potential sexual molestation” and “failed placements.” Id. In July 2005, a psychologist diagnosed her with Cyclothymic Disorder and Oppositional Defiant Disorder, and said she was susceptible to “a full-blown mood disorder such as Bipolar Disorder II or Major Depressive Disorder.” Id. at DFPS009041874. In April 2007, shortly after she was removed from HCYR RTC, a hospital psychologist concluded that S.A. was “a danger to herself and others” and that she should be “committed for long-term psychiatric care.” Id. at DFPS009041884. At this psychiatric admission, she was diagnosed with Bipolar Disorder, Conduct Disorder, and Dysthymia. Id. at DFPS009041879-83. In 2008, she was diagnosed with all of the above plus ADHD and Impulsive Disorder. Id. at DFPS # 52113. By this point, S.A. was prescribed Lithium Carbonate,²⁷ Trileptal, Risperdal, and Vyvanse, to “help improve mood swings, sleep, attention and to reduce aggression, hyperactivity and impulsivity.” Id. In 2009, S.A. reported during a psychological evaluation that she “cannot sleep and has bad dreams every other night in which she gets killed,” that she “has felt so sad that she no longer wanted to live,” and that she “tried to kill herself twice using a knife and a piece of glass.” Id. at DFPS009041886.

27 Lithium is used to treat manic symptoms. The many potential side effects include extreme thirst, weakness, restlessness, confusion, vision problems, lightheadedness, fainting, hallucinations, seizures, fever, nausea, vomiting, diarrhea, drowsiness, and lack of coordination. http://www.drugs.com/lithium.html (last viewed July 27, 2015).

During her 2010 psychological evaluation, S.A. “denied that any happy things [had] ever happened to her” and described her mood as depressed. Id. at DFPS009041893-94. The psychologist at that time found that her intellectual functioning appeared to be “in the borderline to possibly mildly mentally retarded range” and that her “overall functioning is very impaired.” Id. at DFPS009041896. The psychologist was concerned that S.A.'s intellectual functioning could deteriorate even further to “a level of clear mild mental retardation.” The psychologist noted, “In the past she was described to be functioning higher.” Id. at DFPS009041896. The psychologist further noted that S.A. “has a long history of rejection, abandonment, and betrayal by others,” which as a result “has significantly interfered with her capacity to trust others and develop meaningful human relationships.” Id. The psychologist observed that S.A. saw other humans as “perpetrators and irritants.” Id.

Her next psychological evaluation in 2011 noted that she “displayed regressed emotional functioning” and that her behavior had not improved over the past year. Id. at 1 RFP CPS 006986. At this point S.A. was 15 years old and composed the following sentences, among others, during that evaluation: “Other kids make me frustrated”; “I regret hitting little kids”; “I can't stay in a place a long time”; and “I need a family.” Id. at 1 RFP CPS 006989. S.A.'s reading skills were estimated between third and fourth grade levels and the psychologist noted that two of her major psycho-social stressors were “numerous placements” and “failed adoptions” and that “not being apart [sic] of a family is painfully disturbing to this child.” Id. at 1 RFP CPS 006990. The psychologist also found that S.A.'s “caseworkers need to be thinking of her long-term future because” she “will not be able to be on her own when she is 18, if ever” and “will likely always need the support of a closed community to keep her safe and to help her function on a daily basis.” Id. at 1 RFP CPS 006992.

According to Carter, because of her many placements, S.A. had little “sense of who she is and what her history is” and “no sense of ... belonging.” (D.E. 326 at 98-99). She could not remember all the places she has been and could not assist Carter in developing a chronology of her life. Id. She is “disgusted by the ways she has been allowed to fall through the cracks” in foster care due to the constant turnover of primary caseworkers and her many placements. Id. at 96. S.A.'s education was also significantly impacted by her many placement changes; she attended 16 different schools over 13 years. (DX 120 at DFPS009035208-13 (filed under seal)).

Based on an IQ and achievement screening that Carter administered, S.A. is “well below normal in all areas” and at the age of 17 could not “read past the third or fourth grade level.” (D.E. 326 at 99, 130-31). According to Carter, S.A. knows that she is a “psychological wreck” and not equipped to become an independent adult in this world. Id. at 130. Like almost all other current and former Named Plaintiffs, S.A. exhibits a condition of “learned helplessness,” which occurs when a “child learn[s] that [her] circumstance [is] unlikely to change and it manifests itself in the form of a helpless sense or helpless state.” Id. at 91, 132. Two months after she aged out of care and one week after Carter's evaluation, S.A. walked into traffic on a highway and was hit by a car. (DX 120 at DFPS009033437 (filed under seal)). Fortunately her injuries were not life threatening. Id. S.A., who aged out with no life skills and nowhere to go, is another example of the forgotten children in Texas's PMC. Unfortunately, as Carter testified and the Court so finds, S.A.'s experience is “typical ... of the entire foster care system in the State of Texas,” especially in the PMC. (See D.E. 326 at 132-33, 192, 194).

d. A.M.

A.M. is a class representative for the General Class and the Licensed Foster Care Subclass. At the time of trial, A.M. was 17 years old. She was placed in foster care in July 2004 at age seven due to her mother's neglectful supervision. (DX 120 at DFPS009048706 (filed under seal)). Her case files numbered approximately 36,000 pages. When she was placed in care, her father was on death row and her mother was on trial for homicide. Id. at DFPS # 3980. DFPS had previously investigated possible physical abuse and neglect of A.M. and her sister in 2003, but determined no abuse occurred. Id. at 1 RFP CPS 133080. A.M. came into care with two half siblings with whom she was very close. She was essentially the primary caregiver for her younger sisters. See id. at 1 RFP CPS 133081. A.M.'s first caseworker noted that A.M. was “consumed with caring for her little sisters,” particularly her youngest sister, whom A.M. called “my baby.” Id. at 1 RFP CPS 133130. All three siblings were initially placed together in a foster home outside of their home community, but within 50 miles. Id. at 1 RFP CPS 133131.

In August 2004, A.M.'s next younger sister was removed to a kinship placement with her biological father, where she reportedly did well.²⁸ Id. at DFPS # 3375. In October, that child was dismissed from DFPS custody and her biological father was granted permanent conservatorship of her. Id. at DFPS # 3376. Although DFPS initially made efforts to keep A.M. and her youngest sister together, they were eventually separated. See id. at 1 RFP CPS 133361.

28 The three siblings each reportedly had different biological fathers.

A.M. had two failed adoption placements. In the first, after cycling through several therapeutic foster homes, in December 2004 A.M. and her half-sister were placed in a pre-adoptive placement. Id. at DFPS # 3476, 3374-80, 3384. Although A.M. initially did well, her behavior began to deteriorate. See id. at DFPS # 3386-93. These foster parents struggled to handle her fits of rage. In February 2006, A.M.'s therapist notified CPS that these foster parents intended to separate A.M. from her sister and return A.M. to CPS. Id. at DFPS # 3360. A.M.'s caseworker reported serious concerns that these parents were in a “power struggle” with eight-year-old A.M. See id. at DFPS # 3360, # 3424. The caseworker also expressed concern that these parents refused to change their child-rearing techniques despite A.M.'s needs. Id. at DFPS # 1010. The foster father refused all of A.M.'s therapist's suggestions for how to deal with her behavior, instead forcing her to run laps in the back yard. See id. at DFPS # 3428. Nonetheless, the adoption process for only the younger sister continued. Id. at DFPS # 3360.

In April 2006, the foster parents demanded that A.M. be removed immediately as A.M. allegedly tried to hurt her younger sister. Id. at DFPS # 3435-36. Both children were removed. Id. at DFPS # 3437-38. After removal, the two year old younger sister displayed disturbing behavior, including attempting to place a crayon in her vagina and open-mouth kissing. Id. at DFPS # 3437. A.M. reported that the former foster parents punished her by making her drink vinegar, which induced vomiting. Id. at DFPS009048356-64. A.M.'s aunt reported to her caseworker that she was glad that the children were removed as the foster mother made A.M. eat alone in the kitchen, forced her to clean the house, and didn't allow her to play outside. Id. at DFPS # 3450. RCCL investigated the allegations and Ruled Out any abuse in the previous foster home. Id. at DFPS009048366, 009048440-41.

In May 2006, both children were placed in a therapeutic foster home. Id. at DFPS # 3449. A week later A.M. was placed in a psychiatric hospital for anger, rages and threats to harm herself and her foster parents. Id.

In July 2006, A.M. returned to the same foster home and for a time appeared to be improving, though she did have outbursts and crying fits. Id. at DFPS, # 3452, # 3456-57. The foster parents expressed interest in adopting both children. Id. at DFPS # 3456. At this time, the foster parents changed the first name of the younger sister. Id. at DFPS # 3457-61. A.M.'s behavior continued to deteriorate and, in December 2006, the foster parents informed A.M. that “her behavior needed to improve, or she may not be able to stay.” Id. at DFPS # 3461. The caseworker requested that she be given adequate notice if A.M. could not remain with them to avoid placing her in a shelter. Id.

In January 2007, a court agreed to separate the two siblings, thus allowing the younger sibling to be adopted. Id. at DFPS # 3350-51. A.M. apparently had no knowledge of this, as her psychologist wrote to the court in October 2007 that it was in A.M.'s best interest to remain in this placement with hopes of adoption and that her removal from this placement “would be a tremendous loss to her and highly destructive to her ability to attach to any parents in the future.” Id. at DFPS # 3188. The psychologist also warned that the younger sister's adoption should not be kept a secret from A.M. as her “inevitable realization of this deception will significantly impair her ability to trust ... and further impair her attachment process.” Id. Around this time, her caseworker noted that a separation from her sister would wound A.M.²⁹ Id. at DFPS # 3214. In May 2007, A.M. told her I See You worker that she wanted to stay with this family because she loved them and wanted to be with her younger sister. Id. at DFPS # 3468. That same month, the foster parents informed the I See You worker that they intended only to adopt the younger sibling and suggested A.M. be sent to a group home. See id. Still A.M. was left in the dark by everyone as to this second failed adoption. In December 2007, the foster parents formally adopted the youngest sister and changed her last name. Id. at DFPS # 963. A.M. asked her caseworker if her name could be changed as well. Id. at DFPS # 961. No response is noted.

29 From January to October, 2007, A.M.'s only contact with a caseworker was an I See You worker. Notes indicate that this worker spoke with the foster parents by phone and in person once during this ten-month period. A.M.'s monthly impact evaluation prepared by her primary caseworker appears to be entirely based on secondhand information from the I See You worker.

Two weeks later, A.M. had a major meltdown and was placed in emergency respite care. Id. at DFPS # 963. By January 2008, she was in a psychiatric hospital for her behavioral deterioration and for hitting her teacher. Id. at DFPS # 964. The following day, the foster parents gave notice that they could no longer provide A.M. a home even though they signed a long-term care agreement for A.M. around the time they adopted her younger sister. Id. Carter testified that A.M. “deeply regrets” being separated from her two half-sisters. (D.E. 326 at 120, 122).

After a placement change, A.M.'s behavior and emotional state deteriorated even further. According to Talley, A.M.'s next friend, A.M. never recovered from that second rejection. (D.E. 323 at 98). After two weeks in the psychiatric hospital, in February 2008 A.M. was placed in an RTC. (DX 120 at DFPS # 969 (filed under seal)). She remained there for approximately three years, and was visited exclusively by her I See You worker. See id. at DFPS # 1101-51. Despite seeing A.M. regularly over the three years, the worker seems to have never developed a relationship with A.M, mechanically asking the same 10-12 questions at every visit to which A.M. gave one-word responses. See id. at DFPS # 828-997. After A.M.'s primary caseworker dropped her off at the RTC in February 2008, A.M. did not see her primary caseworker again until February 2011, when a different primary caseworker took her to another placement. See id. A.M.'s primary caseworkers changed four times over that period, meaning that three of them never saw A.M. in person. See id. A.M. spoke with her primary workers by telephone less than once a month over that period, often initiating the calls herself. See id.

In April 2009, a person from North Carolina contacted CPS about adopting A.M. See id. at DFPS # 989. For two months, CPS prepared A.M. for the possibility of being adopted in a different state. See id. at DFPS # 991-98. In July 2009, A.M. told her therapist that she did not want to be adopted “because she would live so far away and would not see her sister.” Id. at DFPS # 998. That same month, A.M.'s therapist informed her primary caseworker that A.M. “can't seem to get out of her head that the family that adopted her sister [did] not want to adopt her and won't.” Id. at DFPS # 999. In June 2010, CPS learned that the person from North Carolina had been untruthful and was carrying on a relationship with A.M.'s father, who was still on death row. Id. at DFPS # 905-06. Because of the attempted deception, CPS immediately stopped all contact between this person and A.M. Id. at DFPS # 906. Two weeks later, CPS decided to stop all “active recruitment efforts” for a permanent home for A.M. and her permanency goal was changed to Another Planned Permanent Living Arrangement (“APPLA”). Id. at DFPS # 824, # 906. According to DFPS policy, APPLA is reserved for children age 16 and above when all other possible permanency goals have been explored and ruled out. And the caseworker must state a compelling reason why no other permanency goal is possible or appropriate. (PX 52 at 29). A.M. was 12 years old when her permanency goal changed to APPLA. On her service plan that month, A.M.'s caseworker acknowledged that A.M. may now age out of care. (DX 120 at DFPS # 329 (filed under seal)).

Since June 2010, A.M. cycled through ten more placements (including five different RTCs) and had at least three more psychiatric hospitalizations, including an unnecessary six weeks in a psychiatric hospital because CPS could not find her a placement. See id. at DFPS009043853-920. By February 2014, A.M.'s behavior had deteriorated to such an extent that most RTCs in Texas appeared unwilling to take her. See id. at DFPS009043821. In August 2014, she was discharged from the only RTC that had been willing to take her, when she assaulted a staff member and was taken into police custody. Id. at DFPS009043809, 009044256-61. She was detained at the Bexar County Juvenile Detention Center. Id. at DFPS009043809. As of October 2014, A.M. was living in another RTC, more than 200 miles from her home county. (DX 266 at 3 (filed under seal)). DFPS expected her to age out of care when she turned 18 on July 31, 2015. (PX 2003 at 106 (filed under seal)).

When A.M. first entered care, her level of service was Basic. (DX 120 at DFPS009048706 (filed under seal)). Throughout her time in care, her service levels ranged the entire spectrum, mostly alternating between Moderate and Specialized, with two extended periods at Intense. See id. at 1 RFP CPS 132864-3000. Although A.M.'s level of care had fluctuated previously, she was Basic for most of her stay with the first potential adoptive family in 2005-06. Id. at DFPS # 44-57. Approximately two months after that placement failed, her level of care raised from Basic to Specialized. Id. at DFPS # 58. A.M. fluctuated between Moderate and Specialized throughout her time with the second potential adoptive family in 2006-07; she was Moderate at the time that her sister was adopted. Id. at DFPS # 58-73. At the end of January 2008, almost immediately after A.M. was returned to CPS while her sister was adopted, A.M.'s level of care increased from Moderate to Specialized. Id. at DFPS # 74. Less than two months later, her level of care rose for the first time to Intense. Id. at DFPS # 76. Her level of care fluctuated for the rest of her time in care, remaining primarily on Specialized, and never again reduced to Basic. See id. at 1 RFP CPS 132864-3000, DFPS009043800.

A.M. changed schools at least ten times while in State custody. See id. at 1 RFP CPS 135321-25. Shortly after she entered care, A.M. appeared “to be right on target developmentally” and had “no known developmental delays.” Id. at DFPS009048706. In early 2006, approximately a year and a half later, A.M. was still doing well in school, had good word comprehension, and evenly developed math skills. Id. at DFPS009050704. In April 2006, her evaluating psychologist found that her “basic academic skills [were] better developed than those of most” children her age and believed that A.M. would do “well in a regular education classroom.” Id. A.M. was in third grade at the time and was regarded as “generally a good student, and very bright.” Id. at DFPS009048727. She remained in regular classes until 2008, at which point she began receiving special education services. Id. at 1 RFP CPS 135323, DFPS009050535. She continued to receive special education services on and off throughout the rest of her time in care, though she initially remained a good student overall. See id. at 1 RFP CPS 135321-23. For example, in the sixth grade (2008-09 school year), she received a “90” or above in all of her classes. Id. at DFPS009048993. During the 2010-11 school year (eighth grade), A.M. had a “90” or better in all of her core classes, with a “100” in Mathematics. Id. at DFPS009049166. As of July 2014, though, after being shuffled between different four RTCs over three years, A.M.'s grades fell to “below average.” Id. at DFPS009043800-01. As of October 2014, A.M. was attending a self-contained special education program for “emotionally disturbed children.” (DX 266 at 3 (filed under seal)).

During her time in foster care, A.M. had at least eight different primary, and eight different secondary, caseworkers. (See DX 120 at 1 RFP CPS 133792-4232 (filed under seal)). Because several of her placements were far from her home county, A.M. never met many of her caseworkers. See id. A.M. changed placements at least 21 times, including seven psychiatric hospitalizations. See id. at DFPS009043800. Although early on in her time in care she was placed mainly in therapeutic foster homes, since 2008 she lived primarily (five and a half of her last seven years in custody) in various RTCs, all hundreds of miles from her home county. Id. Carter opined that A.M. is “so damaged from being moved from RTC to RTC, she [doesn't] know how to function in a home.” (D.E. 323 at 97-98). Carter testified that A.M. viewed her experience in foster care overwhelmingly negatively. (D.E. 326 at 120-22). While in care, A.M. would occasionally “achieve relative stability,” during which her service level would improve, and as a result her placement would be changed. Id. at 122. She was often moved just as she became accustomed to her placement. See id. According to Carter, A.M. has not “achieved any kind of lasting stability” in State care. Id. at 120, 122. Carter testified that A.M. projects her pain caused by separation from her siblings outward by being “extremely defiant.” Id. at 122. Like most of the other Named Plaintiffs, she has “no sense of belonging” and does not know how to function in a family environment. See id. at 121-22. She is woefully unprepared “for adulthood, for marriage, family life, social settings,” or employment. Id. at 136. Carter concluded that A.M. had been “greatly hurt” in foster care. Id. at 192, 194. Unfortunately, as Carter testified and the Court so finds, A.M.'s experience is “typical ... of the entire foster care system in the State of Texas,” especially in the PMC. See id. at 132-33, 192, 194.

e. J.S.

J.S. is a class representative for the General Class and the Licensed Foster Care Subclass. At the time of trial, J.S. was 12 years old. (D.E. 326 at 95). He was removed from his mother's home in Lubbock in May 2007, at the age of five, due to neglect and his mother's drug abuse. (DX 120 at DFPS009058393 (filed under seal)). He was initially placed in a foster group home with his older sister (and her infant child). See id. at DFPS009058892. In his first psychological evaluation, in July 2007, the psychologist diagnosed J.S. with Adjustment Disorder with Mixed Anxiety and Depressed Mood, and noted that J.S. may have a learning disability, but also said that if he “remains in a stable environment, his prognosis is good.” Id. at DFPS009062646. His initial level of care was Basic. See id. at DFPS009058904. Yet, two weeks later, J.S. was moved to an emergency shelter at a GRO. Id. at DFPS009058894. Shortly after that move, J.S. made an outcry of prior sexual abuse by two of his young cousins. Id. at 1 RFP CPS 014877, DFPS009068441, DFPS # 40677. Although the allegations were investigated, it does not appear that a doctor physically examined J.S. in connection with that outcry. Id. at 1 RFP CPS 015429. J.S. did not receive any specialized counseling to address those allegations either. See id. at 1 RFP CPS 015463. In October 2007, J.S. was admitted to a psychiatric hospital and his level of care was increased to Specialized because he threatened to cut his neck, cut off his nose and brain, and shoot himself. Id. at DFPS009055638, 009062077. There, only four months after entering care, an evaluating psychologist found that J.S. was suffering from anxiety and depression, noting that J.S.'s “depression appear[ed] to have worsened causing him to act out even more.” Id. at DFPS009062182. In November 2007, DFPS moved J.S. to an RTC in Victoria, 500 miles away, separating him from his sister. Id. at DFPS009058904. After this move, J.S.'s behavior worsened and he began to act out sexually, “exposing himself to his peers and asking for them to perform oral sex on him.” Id. at DFPS009062772. J.S.'s caseworker noted that sexual acting out was now “an obvious issue.” Id. at DFPS009062811. In spite of Defendants' and Plaintiffs' experts testifying that sexually abused children should live in single-child placements, DFPS never placed J.S. in a single-child home. (See D.E. 328 at 161; D.E. 329 at 161; D.E. 303 at 8). J.S. made little progress at the Victoria facility. During that placement, a judge referred to J.S. as “unadoptable.” Id. at DFPS009059404. J.S. remained at that RTC until November 2008, at which point DFPS moved him back to Lubbock to reestablish contact with his sister, per his attorney ad litem's request. Id. at DFPS009058903. Lutheran Social Services, a CPA, placed J.S. in a therapeutic foster home. Id. at DFPS009055662, 009058906.

J.S.'s behavior initially worsened in that home, at times running out into the street and telling his foster mother that he wanted to be hit by a car. See id. at DFPS009057698. J.S. eventually improved, growing close to his foster mother. Back in Lubbock, in December 2008 J.S. began having visits with his sister, which made him very happy. See id. at DFPS9061249. In January 2009, J.S.'s caseworker noted that J.S.'s sister was still visiting him weekly and that his behavior had improved, although he continued to struggle. Id. at DFPS009061253. In May 2009, a psychologist noted that J.S.'s “behavior had improved significantly since this examiner last evaluated [J.S.].” Id. at DFPS009062463. J.S. was diagnosed with Adjustment Disorder with Depressed Mood, Dysthymic Disorder, and ADHD. Id. The psychologist noted that J.S. required “a stable, long-term placement.” Id. In August 2009, J.S.'s caseworker wrote that J.S. “appeared happy and healthy” and that he was “free from abuse and neglect [in] his current placement.” Id. at DFPS # 40524. Two weeks later, though, and for reasons that are not supported by any concrete evidence or findings, J.S. was removed from that placement. The removal appears to have been based entirely his caseworker's concern that J.S.'s foster home had a “chaotic structure” and his doctor's worry that J.S.'s foster mother “may” not have had “adequate training to deal with” J.S.'s behavior. Id. at DFPS # 40528, # 40530. There is no record that DFPS investigated these allegations. According to the caseworker's notes, J.S. liked living in that home and when he was told that he would not be returning, he became very angry and began to cry. Id. The Bair Foundation moved him to another therapeutic foster home. Id. at DFPS009063546.

J.S.'s behavior and condition declined dramatically. Although J.S. had “show[n] improvement at [his] previous school,” a month after being moved, he was suspended from his new school several times. Id. at DFPS009057707. When placing J.S., neither DFPS nor the CPA informed his new foster parents about his history of sexual abuse and sexually acting out. Other than his race, the only information that the foster parents were told prior to bringing J.S. into their home was that he was “desperate for a placement.” Id. at DFPS009062030. The foster parents, who had three other similarly aged foster children and an adopted teenage son, only learned about J.S.'s tendency to act out sexually when J.S. exposed himself to two of his foster brothers. Id. at DFPS009062030, 009061738-39. J.S. shared a room with R.R., the foster parents' 15-year-old adopted son; they slept in bunk-beds. Id. DFPS009062030-32.

Within two months of being placed in that home, J.S. made a sexual abuse outcry. On November 5, 2009, a teaching assistant overheard seven-year-old J.S. tell his eight-year-old foster brother: “I'm gonna stick my [penis] in your [behind], it's gonna hurt you but it's gonna make me feel good.” Id. at DFPS009061758. When later questioned, J.S. indicated that his foster brother had performed anal sex on him, but did not provide any other details. Id. The school's principal told RCCL that the children's behavior had recently become much worse, and the teaching assistant told RCCL that she observed R.R. “be a real bully to the other children.” (DX 120 at DFPS009062029 (filed under seal)). The Child Advocacy Center conducted forensic interviews with J.S. and his three foster brothers, but none of them made any further outcries. Id. at DFPS009062035. During J.S.'s interview, he denied that anyone touched him inappropriately, but did state that his roommate, R.R. “scare[d] him.” Id. at DFPS009062026. There is no record in J.S.'s approximately 40,000 pages of case files that a doctor physically examined J.S. in connection with these allegations. After many years of therapy, J.S. would later admit to his psychologist that a boy at a previous placement blindfolded him, urinated on him, and fondled him. Id. at 1 RFP CPS 136190. Even during his 2014 psychological evaluation by Carter, J.S. only admitted that he was “sexualized” by his foster brother, but was too embarrassed to discuss the incident any further. (D.E. 326 at 115).

During the investigation, the foster father told the RCCL investigator that, one month prior, R.R. had exposed himself to one of the other young boys in the home, that R.R. had a “sexual history,” and that another foster child had several years prior exposed himself to R.R. Id. (DX 120 at DFPS009062027 (filed under seal)). The foster father also told the investigator that due to R.R.'s “history and behaviors,” all of the rooms have baby monitors and motion detectors, although the detectors do not pick up movement in a child's bed.³⁰ Id. The investigator who reviewed R.R.'s case files confirmed that R.R. had a sexual abuse history, had acted out sexually in the past, and had only one week earlier exposed himself to one of the other foster children. Id. at DFPS009062029. Following that previous incident, DFPS implemented a safety plan to ensure that R.R. “would not be alone with any of the foster children.” Id. During the investigation, the investigator discussed with the CPA's case manager his concern that because R.R. was “much older and shar[ed] a room with” J.S., that “could have put [J.S.] at risk.” Id. at DFPS009062028. The investigator specifically noted that because R.R. had a history of sexual abuse, “the incident when he exposed himself to [another foster child] in October pose[d] a concern for the safety of any children alone or sharing a bedroom with” him. Id. The investigator informed the Bair Foundation that before any of the other foster children could return to that home, the CPA needed to “reduce the number of children in the ... foster home” and prohibit R.R. from sharing a bedroom with any other child, which the Bair Foundation agreed to do. Id. ; see 1 RFP 6 RCCL 01565-69. Yet R.R. continued sharing a room with J.S. Id. at DFPS009062029. Despite all of the above, the investigator never interviewed R.R., a decision his supervisor approved. Id. at DFPS009062031-33. The investigator Ruled Out neglectful supervision by the foster parents “with no minimum standard violations cited,” recommending only “[r]outine monitoring.” Id. at DFPS009062035. In making that disposition, the investigator noted that, aside from J.S., the “children in the foster home do not have any history of sexually acting out before or in this home,” in direct contradiction of his own investigation notes. See id. at DFPS009062026-35. The entire investigation, from outcry to final disposition, lasted less than a month. See id.

30 The baby monitors turned out to be of little use as well. As J.S. told his therapist nearly a year later, he “couldn't scream because he was too scared.” Id. at DFPS009066838 (internal quotation marks omitted).

Although DFPS Ruled Out all allegations of abuse and neglect, it did not return J.S. to that foster home and placed him instead in an RTC, again far from his home county, where he immediately began “struggling with aggressive and inappropriate sexual behavior.” Id. at DFPS009057715, 009058914. The record of his time there is replete with serious incident reports, documenting many unsettling behaviors. Besides repeatedly exposing himself to other children, staff continuously found J.S. sleeping in the same bed as, kissing, and inappropriately touching other children. See, e.g. , id. at DFPS009063328, 009067086, 009067206, 009067790, 009068214. The RTC eventually placed J.S. in his own room. Id. at DFPS009068857. Although his sister could no longer visit him, she continued to call J.S., albeit irregularly; J.S. told his caseworker that he missed his sister and liked talking to her.³¹ See id. at DFPS009061159; 009061203-18. J.S.'s next psychological evaluation, in July 2010, reflected marked deterioration. J.S. was diagnosed with PTSD, Major Depressive Disorder, Disruptive Behavior Disorder, and Reactive Attachment Disorder, all of which were new diagnoses. See id. at DFPS009062377, 009063226. In that evaluation, J.S. was characterized as “a very depressed, anxious[,] and angry youth who has suffered through a long history of chronic instability and mistreatment.” Id. at DFPS009063226. His level of care was fluctuating between Specialized and Intense. See id. at DFPS # 42140, DFPS009068857.

31 By March 2010 (and perhaps much earlier), J.S.'s sister had aged out of care, although J.S. would continue to speak with her by phone on and off until at least the middle of 2011. See id. at DFPS009061186-317. His sister attempted to reestablish contact in August 2012, but CPS determined that it would be best for contact to be “limited” until the sister was able to maintain consistency in written correspondence. See id. at DFPS9061361-80. In January 2013, J.S. was again speaking with his sister by phone, but by July 2013, due to sister's inconsistency, J.S.'s caseworker began withholding her letters from him. Id. at DFPS009061391-92, 009061416.

In September 2011, J.S.'s RTC closed so DFPS moved him 200 miles to another RTC in Nacogdoches. Id. at DFPS009055662. J.S. was placed in a room with four other boys in spite of his well-documented history of sexually aggressive behavior. Id. at DFPS009058886. After two months, the Nacogdoches RTC “requested [J.S.] be moved due to his behavior requiring more supervision than what they were able to provide.” Id. at DFPS009058919. In November 2011, J.S. was moved to an RTC 250 miles away in Goldthwaite—his ninth placement in four years, not counting respite homes. See id. at DFPS009055662. His level of care at this point was Intense (his case files are inconsistent as to how long he had been at that level). Id. at DFPS # 42140, DFPS009055662. J.S. remained at an Intense level until July 2013. Id. at DFPS009055656. Shortly before his level of care decreased, an adoption specialist described J.S. as “not looking anywhere near ready for [an] adoptive placement.” Id. at 1 RFP CPS 136258. Yet DFPS, after investing a considerable amount of effort, was able to find J.S. a pre-adoptive placement, more than 500 miles from his home county, which he was moved to in December 2013. Id. at DFPS009055662. In May 2014, J.S.'s caseworker noted that he appeared to be bonding with his potential adoptive father, but that the caseworker was dependent on J.S.'s therapist and secondary worker for all such observations as “it [was] hard to tell what is on [J.S.'s] mind and how deep the bond is from long distance.” Id. at DFPS009061444. J.S.'s adoption was finalized on July 18, 2014. Id. at DFPS009056970, 009056973.

Throughout his time in care, J.S. had at least 15 different primary and secondary caseworkers. See id. at DFPS009059443-527. J.S. rarely saw, or even heard from, his primary caseworkers and had little opportunity to establish stable or meaningful relationships. See id. On the rare occasion that J.S. developed a relationship with a caseworker, it generally did not last very long and sometimes resulted in additional pain for J.S. For example, in May 2011, J.S. told his therapist that he was sad because he missed his caseworker—whom he had not seen in over a year at that point—and asked if he could write her a letter. Id. at DFPS # 42124. Both primary and secondary caseworkers failed to visit J.S. regularly. No caseworker visited or even called J.S. in October 2009, immediately prior to his outcry of sexual abuse at the hands of another child in his placement. See id. at DFPS009059467-69. Most egregious, after his outcry and removal to an RTC, his primary caseworker did not visit him for over four months. See id. at DFPS009059469-74. Even when secondary caseworkers visited J.S. regularly, he was subjected to a flurry of different ones, each a complete stranger. For example, in a six-month period in 2010, J.S. saw six different workers, most of whom he had never met. Id. at DFPS009059443-527. In 2011, J.S. saw seven different caseworkers, some for the first time. See id.

Not including several different respite homes stays, J.S. had ten different placements across Texas during his seven years in care. Id. at DFPS009055662. He attended eight different schools over that time, each in a different county. See id. at DFPS009060987-89. At the age of 12, J.S. was “a virtual non-reader.” (D.E. 326 at 115). According to Carter, J.S. has “severe attachment deficits” due to his experiences. Id. at 114. Although J.S. likely came into care with some attachment deficiencies, they worsened during his time in care “simply because ... he was provided no stability whatsoever.” Id. Carter opined that J.S., due to his experience in substitute care, has “become an aggressor” and will likely hurt himself and others. Id. at 114-16. J.S. was “hurt by his experiences in the State['s] care.” Id. at 190. Unfortunately, as Carter testified and the Court so finds, J.S.'s experience is “typical ... of the entire foster care system in the State of Texas,” especially in the PMC. See id. 132-33, 190, 194.

f. H.V., J.V, and P.O.

H.V. is a class representative for the General Class, the Licensed Foster Care Subclass, and the Foster Group Home Subclass. P.O. is a former class representative for the General Class and the Licensed Foster Care Subclass.³² J.V. is not a Named Plaintiff, but his case files were included with, and linked to, the case files for H.V. and P.O. The Court relies on J.V.'s experience in the same way it relies on the evidence from former foster children Arce, Bentley, Jackson, Virgil, and Sharp. See supra p. 31.

32 The Court holds that certain Named Plaintiffs are no longer adequate class representatives. See infra Section III.

H.V., J.V., and P.O. are brothers who DFPS removed from their family's care in Levelland in October 2004 following allegations of neglectful supervision, physical neglect, and drug abuse by both parents. (DX 120 at 1 RFP CPS 035047-48, DFPS009027864 (filed under seal)). Their case files constitute more than 60,000 pages. All three boys have the same biological mother, but P.O. has a different biological father. Their mother had a long history of drug abuse and it was suspected that H.V.'s and J.V.'s father was using and selling drugs. Neither parent had a permanent home. Id. at 1 RFP CPS 035048-49. P.O.'s father also had a history of drug abuse, criminal behavior, and incarceration. Id. at 1 RFP CPS 034327, 034997, DFPS009026635. All three boys had lived off and on with their maternal aunt and grandmother. Id. at DFPS009027864. The brothers also had an older sister, who lived with the grandmother. Id. at DFPS009028131.

H.V. was five years old and in kindergarten when he entered care with DFPS. Id. at DFPS009027864. J.V. was three years old and in daycare, and P.O. was one year old when they entered the DFPS system. Id. at 1 RFP CPS 032017-18. All three boys' initial level of care was Basic. 1 RFP CPS 148315, 1 RFP CPS 034595, 1 RFP CPS 033430. After removal from their parents, DFPS temporarily placed H.V. and J.V. in a Basic foster home and P.O. with his maternal aunt. Id. at DFPS009026217, DFPS009026134, DFPS009026453. After about a week, DFPS moved H.V. and J.V. to the aunt's home as well, where they stayed for over a year. Id. at 1 RFP CPS 032553. While in that home, H.V. was described as a smart and active boy, outgoing, friendly, and mischievous, without any developmental delays or behavior problems. Id. at DFPS009027864. J.V. usually was observed playing with his brothers or watching television, and P.O. was described as active, giggling, happy, and playful. The children were in play therapy with a psychologist. Id. at 1 RFP CPS 032549.

In November 2005, allegations were made that the aunt tied J.V.'s hands behind his back and made him stand with his nose on the wall, spanked him, and put him in a black hole, later determined to be the basement. It also was alleged that the aunt hit J.V. in the head with a paddle, causing an injury that required stitches, and that he was not fed regularly and was frightened. Id. at 1 RFP CPS 032554, DFPS009028691, 009025491. The boys were removed from that home and placed with their grandmother. While there, J.V. alleged that the grandmother tied his hands with a jump rope and that another member of the household threatened him with a knife. Id. at DFPS009028691. An investigation resulted in a finding of RTB regarding the allegations against the aunt and UTD regarding the allegations against the grandmother. Id. at 1 RFP CPS 037378-79. In the meantime, the parents' custody rights were terminated. DFPS obtained permanent managing conservatorship of H.V. and J.V. on February 10, 2006, and of P.O. on January 9, 2007. Id. at 1 RFP CPS 031078. After the investigation, DFPS moved all three boys to an emergency shelter at the Panhandle Assessment Center on February 17, 2006. J.V. and P.O. were very upset during the transport to the shelter and cursed, screamed, hit, and kicked at workers. Id. at DFPS009028690-91.³³ The boys' time with DFPS is notable for the ready use of psychotropic medication to control what often was age-appropriate behavior, or at the very least, behavior expected from children exposed to violence and upheaval in their young lives. In February 2006, when H.V. was still living with the grandmother, he was seen by a psychiatrist at the request of his caseworker. He was diagnosed with Adjustment Disorder, although it was noted that there was no need for medication at the time and no evidence of ADHD. Id. at 1 RFP CPS 036687-89. While H.V. was at the Panhandle Assessment Center, a psychological evaluation showed that his overall intellectual functioning was in the average range with high average performance skills and high average reading and spelling skills. He was capable of responding in a compliant, cheerful, and appropriately interactive manner, though he frequently responded in an argumentative, dominating, antagonistic, and aggressive manner toward his brothers and residents in the home. He was diagnosed with an adjustment disorder, with disturbance of mood and conduct and a severe history of abuse and neglect. His prognosis was “hopeful,” as long as he could be placed in a well-structured, nurturing, consistent, and enriching environment. Id. at 1 RFP CPS 036725-29. Nevertheless, within one month, H.V. was prescribed Risperdal, an antipsychotic medication. Id. at 1 RFP CPS 036698, 036701. In May 2006, H.V. was doing well and staff at PAC reported no problems. Id. at 1 RFP CPS 032566. He was continued on the Risperdal because of a tendency to be defiant and talk back. Id. at 1 RFP CPS 036701.

33 As early as December 6, 2004, P.O.'s paternal grandparents expressed an interest in adopting him, but inexplicably, their home was not considered as a potential placement until eight years later. Id. at DFPS009028104, DFPS009024714.

In January 2006, two months after making physical abuse outcry, five-year-old J.V. had his first of many visits to a psychiatrist to address behaviors of kicking, hitting, and defiance. He was prescribed Focalin for ADHD, although there is no record of an evaluation. His therapist recommended that J.V. see the psychiatrist. Id. at 1 RFP CPS 031868. At a March 2006 psychological evaluation, J.V. exhibited occasional impatience, demanding behaviors, emotional sensitivity and withdrawal. Id. at 1 RFP CPS 037195. He also struggled to listen and follow directions, and was fidgety and overactive, but was responsive to redirection. His overall intellectual functioning was in the average range. Id. at 1 RFP CPS 037193-99.

P.O. was moved to a therapeutic foster home run by PAC. Notes show that P.O., who had just turned three years old, had tantrums when asked to do something, interrupted adults when talking, used curse words, ignored adults when they talked to him, resisted taking naps and going to bed, and pouted, sulked, and whined when told it was time to clean up or go to bed. Id. at 1 RFP CPS 034589-90. In April 2006, P.O. was prescribed Albuterol for bronchitis. In June 2006, P.O. was taking Sudafed for reasons that are not clear in the record. Id. at 1 RFP CPS 034490, 034609. Both medications list nervousness, restlessness, and sleeplessness as side effects.³⁴ Later in June, based on P.O.'s foster mother's report that he was hyperactive, aggressive, and had trouble sleeping, a psychiatrist prescribed Clonidine to help him sleep. There is no mention in the psychiatrist's notes that P.O. was taking other medications at the time. Id. at 1 RFP CPS 034496. The next month the psychiatrist increased the Clonidine dosage. Id. at 1 RFP CPS 034506-07.

34 See http://www.drugs.com/albuterol.html (last viewed on June 27, 2015); http://www.webmd.com/drugs/2/drug-6573/sudafed /details/list-sideeffects (last viewed on June 27, 2015).

P.O. continued to have tantrums, hit other children over toys, resisted going to time-out, bit his nails, played in bed when he was supposed to be sleeping, whined, interrupted adults, played with his food, and argued, although he also had good days and behaved well at times. Id. at 1 RFP CPS 034624-66. In August 2006, P.O.'s foster mother requested an increase in his level of care from Basic to Moderate based on aggression and not wanting to take naps. Id. at DFPS009028702. However, he remained at a Basic level of care. Id. at 1 RFP CPS 146859. In September 2006, the psychiatrist added Focalin and Risperdal to P.O.'s medication regimen. Id. at 1 RFP CPS 034508-09. P.O. had not undergone a psychological evaluation or intellectual assessment. Id. at 1 RFP CPS 034334. The next month, P.O. began to have occurrences of enuresis (involuntary urination ) and encopresis (involuntary defecation). Id. at 1 RFP CPS 034337. Nothing in the record indicates that any of P.O.'s caregivers or health care providers questioned the possible link between his new medications and his urinary and bowel troubles. These conditions plagued him for years.

P.O. underwent a psychological evaluation on October 10, 2006, where the psychologist noted that the amount of Clonidine he was taking had been reduced. P.O. behaved well during the assessment. His intellectual functioning fell within the average range and his motor behavior was slightly excessive. Based on reports from his foster mother, P.O. was defiant, destructive, disobedient, aggressive, resistant to toilet training, and unresponsiveness to punishment. He was diagnosed with Oppositional Defiant Disorder, ADHD, and Relational Problems (not otherwise specified). Id. at 1 RFP CPS 034352-62. The psychologist recommended that P.O. continue placement in a therapeutic environment that provided adequate stability, consistency, and emotional responsiveness. Id. at 1 RFP CPS 034574.

In October 2006, P.O.'s caseworker submitted a request to raise his level of care to Moderate, but he remained at Basic. Id. at 1 RFP CPS 033428. In March 2007, P.O. was doing well, was cooperative and pleasant. He was easy to redirect and well behaved. The psychiatrist reduced the amount of Clonidine he was taking. Id. at 1 RFP CPS 034535. In May 2007, his foster mother said she would like him to be taken off his medication at some point. Id. at DFPS009028633. P.O. had a difficult month in June 2007, exposed himself to another boy and tried to urinate on him, and talked back to his caregivers. Id. at 1 RFP CPS 034877. In December 2007, his therapist remarked that P.O.'s social skills appeared improved and he was calm and compliant. His therapy was reduced to once per month. Id. at 1 RFP CPS 034587. He was four years old.

In April 2008, P.O.'s therapist said his behavior had improved and he was showing more compliance and less aggression. Id. at 1 RFP CPS 037734. After the boys were placed in the adoption unit, P.O. was moved to a different foster home, in Lubbock, so that he would be in the same city as H.V. and J.V. Id. at DFPS009028658659. P.O. did well in that home. In October 2008, P.O.'s foster mother said he never got in trouble at home or school. He was taking Focalin and Risperdal. Id. at 1 RFP CPS 031816.

In the meantime, in May 2006, H.V. and J.V. were moved to a GRO in Amarillo. After the move, H.V. began acting out violently and was rebellious toward his caregivers, according to his psychiatrist, who increased the amount of Risperdal H.V. was taking. The psychiatrist noted no side effects from the Risperdal, even though listed side effects include aggressive behavior and agitation. Id. at 1 RFP CPS 036703. Another listed side effect is enuresis ; H.V. had begun to wet the bed at night. Id. at 1 RFP CPS 038078.

In May 2006, J.V. was described as shy, quiet, fearful, and aggressive, although he did not display the aggression that H.V. displayed. Id. at 1 RFP CPS 036932, 036957. His therapist noted that nurturance and safety were J.V.'s biggest needs. Id. at 1 RFP CPS 037206. In September 2006, J.V.'s therapist stated that he had become more emotionally defiant with his house parents, cursing at them and trying to provoke them. The therapist also said it was important for staff to evaluate J.V.'s behavior in light of his medications. Id. at 1 RFP CPS 037211. As of November 2006, J.V., five years old and in kindergarten, was taking Strattera,³⁵ Focalin, and Risperdal. Id. at 1 RFP CPS 036946. His level of care rose to Moderate at the GRO's request. Id. at 1 RFP CPS 033430. GRO staff described H.V. and J.V. as a handful, but said that everyone who met the boys enjoyed them and that even though they were active and could be a problem, they were pleasant and enjoyable to be around. Id. at 1 RFP CPS 032579.

35 Strattera is used to treat AHDH in children. Side effects include confusion, hallucinations, unusual thoughts or behavior, feeling aggressive or hostile, painful or difficult urination, drowsiness, and insomnia. http://www.drugs.com/strattera.html (last viewed July 1, 2015).

There was discussion about allowing the aunt and grandmother to visit the boys. J.V. indicated that he wanted to see the aunt but also was afraid of her. Id. at 1 RFP CPS 037209. After J.V. had a visit with the aunt, he complained of a stomachache and was withdrawn and depressed. He chewed his nails, which his therapist attributed to anxiety or a side effect of his stimulant medication. Id. at 1 RFP CPS 037217. J.V.'s behavior began to deteriorate after the aunt and grandmother started visiting him twice per month. He started having wetting and soiling accidents, became distraught, was emotionally out of control and explosive. Id. at 1 RFP CPS 037220. By January 2007, J.V. was very aggressive with other children, hitting, kicking, and biting them. He responded poorly to timeouts and loss of privileges. His therapist said J.V. needed as much individual parenting as possible and the houseparent structure at his GRO was not meeting his needs. Id. at 1 RFP CPS 037224.

In January 2007, the GRO requested that J.V. be moved because of his difficult behavior. He began to improve, but his behavior dramatically regressed after the visits from his aunt and grandmother, and he was completely out of control at home and at school. He was anxious, angry, and explosive, and acted out in a sexual manner. Since starting a drug for enuresis and a stool softener medication, J.V. stopped wetting himself. Id. at 1 RFP CPS 033430-31. On March 13, 2007, J.V. spontaneously said he did not want to have more visits with the aunt or grandmother. Id. at 1 RFP CPS 037234. As of March 26, 2007, Trileptal and Clonidine were added to J.V.'s medication regimen. Id. at 1 RFP CPS 037011

H.V. and J.V. stayed at the GRO until March 2007, when J.V. was moved to a therapeutic foster home. Id. at DFPS009028630. Following the move, J.V. became very disruptive at school. Id. at 1 RFP CPS 037094. J.V. continued to see his aunt and grandmother; his therapist noted it was unclear whether the visits were helping or hurting him. Id. at 1 RFP CPS 037238. In a letter to DFPS dated May 24, 2007, the therapist remarked that the temporary placements of J.V. and H.V. looked more and more destructive. He also expressed frustration that home studies had not been done on the aunt and grandmother, since the boys had at least some sort of connection with them. He added that J.V. was “regressing by the bucketfuls” and “showing more and more attachment dysfunction [with] each placement.” Id. at 1 RFP CPS 037240. The therapist also noted that J.V. “needs a mother and special treatment” and that he “cannot be medicated into health, only managed chemically.” Id. The therapist asked DFPS, “What do you need to establish a permanency plan?”; it is unclear if he received a response. Id.

H.V. was moved to the therapeutic foster home where J.V. was living in April 2007. Id. at DFPS009028632. H.V. did well in the home, but his foster parents commented that he needed to handle himself better and not act so aggressively. Id. at DFPS009028634. H.V.'s level of care was raised to Moderate in April 2007. Id. at 1 RFP CPS 036006. In June 2007, at eight-years-old, H.V. was taking Invega,³⁶ Vyvanse, and Clonidine, though it is not clear when those medications were prescribed or when Risperdal was discontinued. Id. at 1 RFP CPS 031073. That same month, the foster parents gave notice that they could no longer care for H.V. and J.V., in part because of their disruptive behavior. H.V. and J.V. were then moved into a different therapeutic foster home managed by the Bair Foundation in Lubbock. Id. at DFPS009028635-37. After H.V. and J.V. were moved, the foster mother said she thought J.V. was over-medicated because he was very quiet and had trouble listening and focusing. Id. at 1 RFP CPS 037241. A psychological assessment indicated that J.V.'s overall intellectual functioning was in the low average range and his prognosis was “guarded.” Id. at 1 RFP CPS 033831.

36 Invega is an antipsychotic medication used to treat schizophrenia in adults and teenagers who are at least 12 years old. Common side effects include uncontrolled muscle movements, drowsiness, and agitation. http://www.drugs.com/invega.html (last viewed June 30, 2015).

Following a visit to a psychiatrist in July 2007, J.V.'s medications were changed to Invega, Vyvanse, and Clonidine. He also took Desmopresson Acetate for bedwetting. Id. at 1 RFP CPS 033939. In August 2007, he was doing well and his caseworker reported that his eyes were brighter since his medications had been reduced. Id. at DFPS009028642. In October 2007, J.V. was still having a difficult time following directions at school and at home, talking back to his teachers and foster parents. He threw tantrums lasting up to 20 minutes, though he was making progress in not wetting the bed. Id. at 1 RFP CPS 037048.

H.V. did not like the new placement and accused the foster parents of being racist. He resented moving and blamed J.V. for the move. It was noted that H.V. needed a stable environment where he knew someone loved him unconditionally, and that many of H.V.'s negative behaviors were attributable to being upset at changing placements so often. Id. at DFPS009028643-44, 009027904. In October 2007, H.V.'s foster parents called the police because H.V. became angry and threw chairs against a glass door. His foster mother wanted to have H.V. moved but keep J.V., who was doing well in the home. Id. at DFPS009028645.

H.V. and J.V. lived in that home for approximately 18 months until December 23, 2008 when they were moved to another therapeutic foster home (also managed by the Bair Foundation) in Muleshoe. P.O., who was still in his earlier foster home, was moved into this home to be reunited with his brothers. The decision to move H.V. was made because his foster parents had given notice that they could no longer care for him. J.V. and P.O. were moved to the same placement so that the brothers could be together. Id. at DFPS009028543.

P.O.'s previous foster parents were extremely distressed that P.O. was moved because he was doing very well in their home and because he was moved during the holidays. Id. at 1 RFP CPS 035257. P.O.'s behavior deteriorated rapidly after the move. He acted out, broke other children's toys, stepped on their fingers, screamed, threw tantrums, pinched his own skin to the point of bleeding, urinated on himself, and constantly fought with his brothers. Id. at 1 RFP CPS 031822, 032097. On March 18, 2009, P.O.'s foster mother noted that he “got his meds regulated for behavior.” Id. at 1 RFP CPS 033545. On March 27, 2009, P.O.'s level of care was still Basic and he was taking Focalin and Clonidine. Despite his behavior at home, he was doing well in kindergarten. Id. at 1 RFP CPS 031077.

H.V. and J.V. also struggled after the move. H.V. displayed violent behavior with kicking, screaming and not following directions. Id. at 1 RFP CPS 031822. J.V. had bonded with his previous foster mother, but in the new placement talked back, was aggressive, struggled to follow directions, and wet the bed every night. Id. at 1 RFP CPS 033929. In March 2009, H.V. was taking Clonidine and Abilify. He was described as bright, manipulative, and quick to anger. Id. at 1 RFP CPS 031077. In April 2009, a psychologist evaluated H.V. and diagnosed him with Disruptive Behavior Disorder. He had symptoms of ADHD and Oppositional Defiant Disorder, but there was insufficient evidence to support a diagnosis of those disorders. His prognosis was “fair.” There were no contraindications to adoption, but it was recommended that H.V. be adopted with his brothers and that caregivers should assume that any adoption would not go smoothly for the first six months. Id. at 1 RFP CPS 036789-99. H.V. continued to have emotional and violent outbursts and was overheard talking in a sexually inappropriate way about the household dog. Id. at 1 RFP CPS 038579, 031851. However, he also appeared to be trying to control his temper and behave better. Id. at 1 RFP CPS 031858. Though, in June 2010, H.V. started a fire in his grandmother's garage during a visit. Id. at 1 RFP CPS 036843.

In April 2009, J.V.'s foster mother reported that when J.V. was angry he cursed and yelled and threw tantrums, although they did not last very long. He often did not follow the house rules and could be argumentative and oppositional, lied, and occasionally stole things. She described him as hyperactive and impulsive. Testing showed J.V. in the average range of intelligence, functioning near age-level expectations in spelling, math, and reading. J.V. reported some auditory hallucinations and said he frequently heard a voice talking to him at school but could not understand what it said. J.V. was diagnosed with ADHD, but was ruled out for psychotic symptoms. The therapist said that any potential adoptive parents should be made aware of J.V.'s behavioral difficulties and should have an opportunity to form a relationship with J.V. and his brothers prior to placement. Id. at 1 RFP CPS 034109-18.

On September 1, 2009, a safety plan was implemented following an incident that occurred on August 28, 2009. An incident report was not located in the record, although it may be related to a conversation the foster mother overheard between J.V. and H.V., where J.V. said he was going to urinate in the dog's mouth. The safety plan required that H.V. and J.V. not be allowed around pets without adult supervision, that H.V. and J.V. have a responsible adult present when they were playing outside, and that their foster parents report any inappropriate touching or talking to the CPA immediately. Id. at 1 RFP CPS 033527, DFPS009028565.

In February 2010, J.V.'s level of care was Specialized. Id. at 1 RFP CPS 033829. He was having extreme tantrums when he was asked to do anything, although he was not having behavior problems at school. He continued to urinate on himself and wet the bed at night. Id. at 1 RFP CPS 033833. Due to his aggressive behavior, J.V. was to be closely supervised by adults during waking hours. Id. at 1 RFP CPS 033842. In June 2010, J.V. continued to have difficulty controlling his bowels and would hide his soiled clothes. Id. at 1 RFP CPS 034077. In July 2010, J.V. said he had reservations about being adopted. Id. at 1 RFP CPS 034144. Not even two weeks later, despite all the behaviors noted above, CPS staff stated in an adoption readiness summary that J.V. was ready for adoption, although there might be an adjustment period and time of behavioral issues once he was placed with the family. Id. at 1 RFP CPS 035792.

P.O. underwent a psychological evaluation on April 13, 2009. His foster mother described disruptive behavior and said that he was hyperactive and impulsive. He had average intellectual functioning and was functioning consistently at a Kindergarten level in reading, spelling and arithmetic. He was diagnosed with ADHD and Adjustment Disorder with disturbance of conduct. His prognosis was “fair” and it was recommended that every effort be made to keep P.O. and his brothers together. It was also recommended that any potential adoptive family be made aware of his behavior difficulties and be allowed to form a relationship prior to adoption. Id. at 1 RFP CPS 037564.

P.O. continued with his disruptive behavior and, in September 2009, touched a pet in an inappropriate way. He was doing well at school but was confrontational and aggressive at home. Id. at 1 RFP CPS 037555. In October 2009, P.O. touched another boy inappropriately. When his foster mother talked to him about it, P.O. said that in his former foster home he used to touch all the little girls in their private areas. The foster mother said she had monitors in the bedrooms and was “pretty certain” there was no sexual acting out going on between the boys. Id. at DFPS009028569-70. P.O.'s level of care was Moderate as of October 1, 2009. Id. at DFPS009028474. As of April 2010, P.O. was taking Focalin and Clonidine. His foster father reported P.O. was doing well in school but would get in trouble for not taking schoolwork home. When he did take it home, he “[ate] it on a daily basis.” He had trouble sleeping but did better with medication. He cried frequently and fidgeted when he was sent to time-out. He regularly lied and stole. He angered easily, used foul language, screamed, hit, kicked, bit, destroyed property, and was cruel to the family dog. He was seven years old and in the first grade. Id. at 1 RFP CPS 036453-56. Despite his behavior, P.O. functioned at age-level expectation in school and appeared to possess adequate resources for performing well academically. He underwent a psychological evaluation in April 2010, still diagnosed with ADHD and Adjustment Disorder with disturbance of conduct. His prognosis was “fair” and the psychologist noted that his behaviors worsened since being placed with his brothers, but continued to recommend that every effort be made to keep the brothers together. Id. at 1 RFP CPS 036457-61.

In May 2010, a visitor to the foster home reported to CPS that P.O. told her that an older boy, who shared a room with P.O. and his brothers, hurt P.O.'s “wee wee.” The visitor also reported that the foster mother was verbally abusive to a girl in the house and would not let any of the children talk to visitors. A CPS investigator interviewed all the children and the foster parents, and they all denied that any inappropriate touching or other abuse. P.O. was examined by a doctor who saw no evidence of sexual abuse. The CPS investigator Ruled Out sexual abuse, but the older boy was moved to a room by himself. Id. at DFPS009030587-601.

On August 30, 2010, all three boys were placed in a pre-adoptive home in Blanco. Id. at 1 RFP CPS 038603. At that time, H.V. was 11, J.V. was 9, and P.O. was 7. This adoption was doomed from its inception. The adoptive parents wanted to adopt two children under the age of eight, and preferably under five. They wanted to adopt girls but were open to adopting boys. They wanted to adopt children needing Basic-level care; H.V. and P.O. were at Moderate levels of care and J.V.'s level of care was Specialized. Id. at DFPS009028610. The parents also did not want to adopt children with severe emotional or mental issues. Id. at 1 RFP CPS 035735, 035746. The person who prepared the home study stated that children who had severe violent behaviors, sexual acting out behaviors, animal cruelty, or fire setting—all behaviors that the boys had exhibited—should not be referred to this family. Id. at 1 RFP CPS 035735. At the adoption staffing, all of the DFPS staff members were concerned about letting the family adopt the boys. The concerns included the family's preference for girls and children under the age of eight; an incident between the father striking an adopted daughter who later became estranged from the family; that all three boys would have to share a room; that the family had dogs; and that they kept guns in the home. Id. at 1 RFP CPS 035876-88. Nonetheless, the decision was made to place the boys in that home.

The first pre-placement visit went well. During the second visit, though, J.V. had meltdowns at bedtime and in the morning when getting up. He appeared to want one-on-one attention. H.V. behaved well and P.O. had some minor episodes. Id. at 1 RFP CPS 032723. On the first day they were actually placed in the home, J.V. ran off but came back. He said he wanted to live with his mother and father and that H.V. was mean to him. J.V. stole some items from the foster parents' son. Id. at 1 RFP CPS 032724.

The adoption disrupted in stages. One of the first things the foster parents did was give the boys new first names. In the first few days of the adoption, J.V. was having anger outbursts and destroying property. Id. at 1 RFP CPS 034128. Six days after being placed in the home, J.V. was admitted for acute care at a mental health treatment center in San Antonio for homicidal ideation and physical aggression. At the foster parents' request, the staff stopped all of J.V.'s medications and he was released to their home five days later. Id. at 1 RFP CPS 034133. J.V. continued with outpatient therapy, but was placed in an emergency shelter a few days later after threatening to harm his foster mother; he never returned to that home. Id. at 1 RFP CPS 034241. For J.V., the adoption was over two weeks after it began.

H.V. was slow to bond with the foster parents but did appear to bond with their biological son. The parents began tapering him off his medications soon after he arrived in the home and said they did not notice any change in his behavior. He did well in school and began making friends but he threw tantrums at home and screamed when corrected or told to do chores. Id. at 1 RFP CPS 032702. H.V. blamed the foster parents for the removal of J.V. from the home and his behavior worsened after J.V. was removed. Id. at 1 RFP CPS 032711. H.V. also began bullying their biological son. On November 12, 2010, H.V. was removed from the home and placed in a therapeutic foster home in San Antonio. He appeared sad but not distressed about being removed from the pre-adoptive placement. Id. at DFPS009028610-12.

P.O. resisted attaching to the foster parents and did not let his guard down. He made some progress at home, but did poorly in school. He had to be monitored around their pets because he could be cruel to them. Id. at 1 RFP CPS 032710. The foster parents took him off his medications. Id. at 1 RFP CPS 038655. P.O. was removed from the home along with H.V. and placed in the same therapeutic foster home on the same day H.V. was removed. P.O. was sad about being removed from the home and blamed H.V. Id. at DFPS009028612.

In addition to J.V. entering a mental health care facility following the disrupted adoption, H.V. and P.O. also showed signs of intense mental distress over the next several months. H.V.'s foster mother in San Antonio described him as “a child who is lost, had a lot of bravado to cover his fears and is protective, yet disdainful of his little brother.” He had a bad temper and responded poorly every time he was disciplined. He was aggressive, threatened other boys, did not sleep well, argued, talked about sexual things, did not want to shower or brush his teeth, and arrived with a stolen cell phone. He was stealing from the other children and had no boundaries. On November 19, 2010, he was prescribed Risperdal and Concerta. The next day, H.V.'s “hands were flip flopping around, his body was contorting into all kinds of poses and he simply could not sit still.” That behavior continued for more than three hours. Two weeks after being placed in the foster home, H.V. was lying openly, talking about his sexual past, threatening older, bigger boys, and having tantrums every other day. Id. at 1 RFP CPS 038658-59. In December 2010, H.V. had trouble sleeping and was very moody and provoking others in the home. The Risperdal was discontinued and he was prescribed Seroquel. Id. at 1 RFP CPS 036888. H.V.'s level of care was increased to Specialized in December 2010. Id. at DFPS009028482.

A week after the disrupted adoption, P.O. became aggressive at school, disrupting the classroom and threatening other children, and made suicidal statements. When the counselor walked him to her office, he twice tried to trip her. While in the assistant principal's office he jumped off a desk approximately 25 times, threw a chair, pulled off baseboards, and banged on a windowsill. He threatened to kill himself and told the counselor he would kill her too. His foster mother picked him up from school and took him to a psychiatric hospital the next day for evaluation. Id. at DFPS009028610, 1 RFP CPS 037973. While at the psychiatric hospital, P.O. was prescribed Risperdal and Vyvanse. Id. at 1 RFP CPS 037579.

By December 17, 2010, all three boys had been moved to a GRO in Lubbock. H.V. shared a cottage with J.V., and P.O. lived in another cottage on the campus. Id. at DFPS009028614. On January 15, 2011, records show H.V. was taking Abilify and Clonidine. Id. at 1 RFP CPS 038372. On January 18, 2011, H.V. had a very violent episode after becoming angry with another boy. H.V. went throughout the house turning over furniture, breaking pictures and window blinds, throwing chairs, and threatening to harm the other boy. The house parents called the sheriff, who took H.V. into custody. Id. at 1 RFP CPS 038704-05.

At a court hearing on January 6, 2011, it was determined that H.V. would be placed with his grandmother and he was placed there on January 24, 2011. Id. at 1 RFP CPS 032274. However, that placement ended on May 31, 2011 when H.V. was moved to an emergency shelter at a GRO in Lubbock “due to his behaviors.” Id. at DFPS009027953. Notes indicate that there was a lot of conflict between H.V. and his sister, who had lived with the grandmother for years. Id. at 1 RFP CPS 036627. H.V.'s level of care was decreased to Basic on May 31, 2011. Id. at 1 RFP CPS 036005. H.V. was prescribed Intuniv in July 2011. Id. at DFPS009028501-02.

J.V. did well at the GRO and, in January 2011, staff generally had no issues with his behavior. See id. at 1 RFP CPS 032277. His level of care was reduced to Moderate in April 2011. Id. at 1 RFP CPS 032282, DFPS009028482. In an August 2011 psychological evaluation, J.V. was noted to anger easily and throw tantrums when upset, but many of his behavior difficulties had improved. Id. at 1 RFP CPS 036175. The psychologist did not find significant support for J.V.'s previous diagnosis of ADHD, but because he was taking psychotropic medication, it was assumed that the ADHD was present but responding favorably to medication. Id. at 1 RFP CPS 036174-75, 036178-79.

In January 2011, P.O. was enjoying the GRO and, other than a tantrum the first weekend he was there, his behavior had been appropriate. Id. at 1 RFP CPS 037465. His level of care was Moderate. Id. at DFPS009028474. He was described as obedient, helpful, and sweet-natured. Id. at 1 RFP CPS 037466. He was taking Clonidine and Focalin. Id. at 1 RFP CPS 037469. P.O. underwent another psychological evaluation on August 30, 2011. His medications were changed to Vyvanse and Risperdal. Staff reported that P.O. angered easily, particularly when asked to do something he did not want to do. Although he occasionally screamed and threw tantrums, he was more likely to be quiet when angry, and had told staff in a quiet voice that he was going to kill them. When he was not upset or angry, he was sweet and would express remorse for his actions. He did well on his medication regimen. The psychologist believed that P.O. had failed to develop an appropriate attachment relationship to any caregiver due to the long history of instability in his placements. He was diagnosed with ADHD, although the psychologist said he did not find support for the diagnosis on the day of the examination and assumed that P.O.'s medication was controlling the symptoms. His prognosis was “fair to good” and it was recommended that he be placed for adoption with J.V. Id. at 1 RFP CPS 036437-49.

After the emergency shelter, H.V. returned to the GRO with his brothers and stayed there for four months, until September 27, 2011, at which point he was moved to a therapeutic foster group home. (Id. at DFPS009028503-04; PX 2015 at 90; D.E. 215 at 6). On October 14, 2011, P.O. and J.V. were placed in a pre-adoptive placement in Wichita Falls with a single woman who had worked for CPS for many years. J.V. initially did well in the placement, but P.O. had problems from the beginning, and was defiant and not focusing in school. (DX 120 at 1 RFP CPS 032735 (filed under seal)). J.V. was sad because he was never going to live with his grandmother. Id. at 1 RFP CPS 032736-37. Six weeks after placement, the adoption disrupted when the therapist told the adoptive mother that neither J.V. nor P.O. were ready for adoption. P.O. had been urinating in the trashcan, wiping feces in the bathroom, and refusing to do his homework. When the adoptive mother tried to redirect him, he was defiant. Id. at 1 RFP CPS 032738. In December 2011, P.O. was moved to the therapeutic foster group home where H.V. was staying. P.O. was happy to see H.V., but sad about leaving the adoptive home. Id. at 1 RFP CPS 032323. J.V. was taken to the Texas Boy's Ranch and cried for half an hour because he had to leave the adoptive home. J.V. told his caseworker that he was moved because he could not behave and could not stop stealing. Id. at 1 RFP CPS 032323.

Following this second disrupted adoption, J.V. acted out in increasingly alarming and violent ways. In February 2012, he was suspended from regular classes after he held a girl down and told her he was going to rape her. He refused to follow directions or go to bed on time at home. Id. at DFPS009028524. J.V. was discovered looking up “rape” and pornography on the Internet and was suspected of “huffing” paint, although he denied it. Id. at 1 RFP CPS 036161. On March 6, 2012, J.V. told his caseworker that he did not know why he was acting out, but then said that he was acting out so that he would be moved to live with H.V. He also said that he wanted to go back to his adoptive placement and asked why no one asked him about moving before he was moved. Id. at DFPS009028527. A request to raise J.V.'s level of care to Specialized was submitted on March 7, 2012. Id. at 1 RFP CPS 031085.

J.V. was admitted to the Canyon Lakes Day Treatment Program on March 19, 2012 because of escalated behaviors and the threat of rape he made to the girl at school. Id. at 1 RFP CPS 036060. On March 29, 2012, one of his therapists said J.V. might be acting out because he wanted to return to his adoptive home and recommended his caseworker contact the adoptive mother to ask about the possibility of adopting only J.V. Id. at 1 RFP CPS 032334. There is no indication in the record that the adoptive mother was contacted about this possibility.

J.V. returned to the Texas Boy's Ranch and, in May 2012, a psychological assessment showed that he was still having behavior problems, the previous weekend stating he wanted to kill himself and refused to eat, play, or take his medication. He was taking Vyvanse, Clonidine, and Lexapro.³⁷ He was still and focused during the assessment and produced drawings that were well done and had significant detail. He was tearful and talked about being in trouble and feeling he would not be able to “start over.” J.V. had significant emotional distress with symptoms suggesting a major depressive disorder. The depression was believed to be a greater factor in his difficult behavior than the ADHD. There were indications of depression and sadness related to his experiences of abuse, neglect, loss or abandonment, and repeated placement changes and failed adoptions. According to the assessment, his acting out and sexualized behaviors may have been attempts to get his needs met, using fantasy and seeking out ways of demonstrating strength and power. It was believed he met the special education criteria in the category of emotional disturbance due to sad mood, difficulty establishing and maintaining relationships, and inappropriate behaviors and feelings under normal circumstances. Id. at RFP CPS 064556.

37 Lexapro is an antidepressant used to treat anxiety in adult and major depressive disorder in adults and adolescents who are at least 12 years old. Side effects include racing thoughts, unusual risk-taking behavior, and feelings of extreme happiness or sadness. http://www.drugs.com/lexapro.html (last viewed July 7, 2015).

In May and June 2012, J.V. did better, but his level of care was raised to Specialized in September 2012. Id. at DFPS009028411. In November 2012, J.V. did much better. His foster mother commented that he would be an amazing only child. Id. at DFPS009028420. In March 2013, J.V. was succeeding at home and at school. His house parent did not think he had been stealing and he had asked for an air freshener for his room because it smelled like urine. It was considered a big step that he had admitted urinating in the closet. Id. at DFPS009028451.

In December 2011, in his first week in the group home with H.V., P.O. repeated episodes of encopresis and did not shower. Id. at 1 RFP CPS 032321. By May 2012, P.O. was doing much better and had developed a bond with his foster parents. He made great improvement with his tantrums and was able to control his anger. He was taking Vyvanse, Risperdal, and Intuniv. Id. at 1 RFP CPS 036468. P.O. received court-ordered neuropsychological evaluations in May and June 2012 with an educational psychologist. P.O. was nine years old and in third grade. P.O. had reached all developmental milestones, but continued to struggle with encopresis and enuresis. His foster mother was most concerned about his lying and stealing. His behavior at school had improved. He had no learning disabilities, but did have some areas of weakness that interfered with learning to an extent. P.O. was described as a very complex child with many negative experiences in his young life. The psychologist said there was no doubt that frequent placement changes interfered with his emotional and psychological health. He had learned to not become too attached to people. His prognosis was “guarded.” Id. at 1 RFP CPS 036411-25.

P.O. stayed at that foster group home until he was placed for adoption in the home of his paternal grandparents in December 2012. Id. at 1 RFP CPS 147612. It is unclear why placement was not sought with the grandparents when P.O. first entered the DFPS system, considering that the grandparents inquired about adopting P.O. as early as December 2004 when P.O. was not yet two years old. Id. at DFPS009028101-04. Incredibly, P.O.'s grandparents were not notified until late 2012 that P.O. was in the State's PMC and available for adoption. See id. at 1 RFP CPS 153525. They had no contact from P.O. (or DFPS), and had been under the impression that he was living with his aunt ever since being removed from his home. Id. Placing P.O. with his grandparents eight years earlier might have saved P.O. years of trauma. Instead, he suffered through ten placement changes and two failed adoptions. Id. at 1 RFP CPS 146624. He also attended seven different elementary schools between kindergarten and third grade, where, despite the upheaval, his school performance was overall “good.” See id. at 1 RFP CPS 035991. He additionally had eight primary and secondary caseworkers. See id. at DFPS009027500, 009028630-97, 009028711-94.

H.V.'s behavior improved in his more recent years in foster care. He struggled with responding respectfully to his foster mother, but did respond to redirection by his foster father. He was excited when P.O. moved to the foster group home. In December 2011, H.V. was taking Seroquel, Intuniv, and Concerta. Id. at 1 RFP CPS 036582. In June 2012, P.O. moved into H.V.'s room at the group home, which seemed to go well. Id. at 1 RFP CPS 032352. H.V. was sad when P.O. was placed in his grandparents' house and said at that time that he wanted to be adopted with J.V. Id. at DFPS009028428. In February 2013, CPS decided to place H.V. and J.V. separately for adoption. H.V. was angry and sad about the decision, but acknowledged that it might be for the best. His behavior improved and he was no longer arguing with his foster parents. Id. at DFPS009028429-32. By June 2014, a psychological assessment of H.V. showed that he made tremendous progress. He no longer angered easily and did not typically exhibit aggressive or destructive behaviors. He was diagnosed with Adjustment Disorder; it was noted that while he showed some characteristics of ADHD, there was insufficient evidence to support a diagnosis. His prognosis was “good.” Id. at DFPS009022130. As of August 2014, H.V. was not taking any medications. Id. at DFPS009022099-100. Also in August 2014, H.V. was moved to a foster home that was intended to be a permanent placement. H.V. was friends with the family's biological son. Id. at DFPS009028196, 009022075. In his ten years in DFPS care, H.V. experienced 13 placements and one failed adoption. See id. at 1 RFP CPS 146607. He attended 11 schools in nine years, including five different schools while in the sixth grade, yet remarkably managed to perform (sometimes exceptionally) well academically. See id. at 1 RFP CPS 146634, DFPS00902210. He had eight primary and secondary caseworkers. See id. at DFPS009028711-94.

A CPS placement review in J.V.'s record indicates that during his visit to a psychologist in January 2014, he made a sexual abuse outcry. Id. at DFPS009022102. The psychologist's assessment from that visit was not in the record and no other mentions of sexual abuse were found. There is also no evidence in the record as to whether those allegations were investigated. After that outcry, J.V., who had previously demonstrated troubling and inappropriate sexualized behavior, remained in multiple-child facilities. As of August 2014, J.V. was living at the same GRO, although there is a note that a potential adoptive family had been selected for him. Id. at DFPS009022101. As of the end of the trial record, J.V. had experienced at least 13 placements and two failed adoptions during his time in care. Id. at 1 RFP CPS 146616. J.V. changed schools nine times in nine years, but generally did well academically. See id. at 1 RFP CPS 14663. J.V. has had nine primary and secondary caseworkers. See id. at DFPS009028711-94, 009026196. Unfortunately, as Carter testified and the Court so finds, H.V.'s and P.O's experiences are “typical ... of the entire foster care system in the State of Texas,” especially in the PMC. (See D.E. 326 at 132-33, 194).

g. Z.H.

Z.H. is a class representative for the General Class and the Licensed Foster Care Subclass. At the time of trial, Z.H. was 14 years old. He entered 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. (DX 120 at DFPS # 33580 (filed under seal)). DFPS investigated multiple reports of abuse and neglect regarding Z.H. between 2003 and 2009. See id. at DFPS # 32996-97. In 2007, DFPS investigated a report of sexual abuse of Z.H. and his younger brother. Id. at DFPS # 37143, # 32996. At that time, two unrelated men, both sex offenders with backgrounds in prostitution who had allegedly stated that they “liked little boys,” were living with Z.H.'s mother. Id. Following the investigation, DFPS found RTB that Z.H. was sexually abused by one of the men. Id. DFPS, though, Ruled Out neglectful supervision by Z.H.'s mother because she “denied knowing [the] criminal history of her homeless friends.” Id. DFPS thus did not remove Z.H. at that time. There is no indication in Z.H.'s 38,000 pages of case files that he received therapy for that sexual abuse incident once he entered foster care.

Between 2007 and 2008, Z.H. was psychiatrically hospitalized on four occasions and had been diagnosed with several significant psychological disorders. Id. Z.H. was removed from his mother's custody in 2009, at the age of eight. The removal was precipitated by a hospitalization at the end of 2008, during which it was determined that he was malnourished and that his mother was not administering his required medication. Id. at DFPS # 33580. Z.H.'s two brothers (five years old and ten months old, respectively) were both removed at that time as well. Id. at DFPS # 32391, # 33059. Shortly thereafter, the middle brother was placed with his biological father while the youngest brother was placed with a foster family. Id. at 33093. Z.H. was placed in an emergency shelter.

Over his six years in care, Z.H. was moved at least nine times, spending time in RTCs, therapeutic foster family and group homes, emergency shelters, and psychiatric hospitals. See id. at DFPS009116244-51. For most of his time in care, Z.H. was more than 100 miles from his home county, which made sibling visits difficult. His very first placement, in February 2009, was in a GRO emergency shelter, 60 miles from his home county. Id. at DFPS # 32998. Z.H. cycled through several more placements until he was eventually, in June 2009, placed in an RTC 150 miles from his home county. See id. at DFPS # 33187-202, # 33244, # 33580, DFPS009116244-47. In addition, many of Z.H.'s placements were clearly inappropriate, certain to fail from the outset. For example, despite his violent outbursts, violence toward himself, and history of sexual and physical abuse, Z.H. was placed in at least two therapeutic foster family and foster group homes with other foster children. See id. at DFPS009116749. Despite Z.H.'s history of sexual abuse, he was never placed in a single-child home as he should have been according to Defendants' and Plaintiffs' experts. (See D.E. 328 at 161; D.E. 329 at 161; D.E. 303 at 8).

Z.H. mainly lived in RTCs where he was heavily medicated. See, e.g. , id. at DFPS # 33588-89, # 37320-34, DFPS009119337, 009119380. He was once admitted to a hospital because he “could not function due to being over medicated.” Id. at DFPS # 37282. Although Z.H. made some progress in his first RTC, he was moved after less than a year, partly because he was being overmedicated and partly because the RTC could not meet his needs. The next RTC was approximately 300 miles from his home county. See id. at DFPS009116247-48, 009116745. Z.H. remained at that RTC for nearly three years, again making some progress, until he was moved to another RTC, primarily to be closer to his home county. He remained there for approximately one year. See id. at DFPS009116745. In June 2014, at the age of 14, he was moved into a kinship placement in his home county. Since then, he has had contact with his middle sibling, but not his youngest sibling, who was adopted out of care in May 2011. His initial level of care was Specialized, although it was increased to Intense after four and a half months in care. See id. at DFPS009116744-45. Z.H. fluctuated mostly between Specialized and Intense throughout his time in care. See id.

Z.H.'s first psychiatric evaluation in care, in March 2009, contained diagnoses of PTSD, Mood Disorder, ADHD, and Reactive Attachment Disorder. Id. at DFPS # 37147. The evaluating psychologist noted that Z.H. was a victim of physical and sexual abuse. Id. Two of the sentences that Z.H. completed during that evaluation were: “The thing I do worst is ‘not getting hurt,’ ” and, “The worst person who ever lived was ‘my caseworker.’ ” Id. at DFPS # 37146. The psychologist stated that Z.H. “needs to reside in a secure stable living environment,” and to be “enrolled and remain in the same school” throughout the school year in order to receive “the benefit of a consistent educational experience.” Id. at DFPS # 37149. The psychologist recommended “long-term placement in a foster-care home or residential group home” and noted that Z.H. “will need intensive monitoring by caretakers” due to risk of suicide. Id. In April and June 2009, he was diagnosed with Bipolar Disorder, ADHD, Generalized Anxiety Disorder, PTSD, and Oppositional Defiant Disorder. Id. at DFPS # 37208, # 37213. He continued to experience serious psychological and emotional issues throughout his time in care. Repeated placement changes and caseworker turnover—Z.H. had at least nine different primary and secondary caseworkers—contributed to disruptions in his medication regimen, which resulted on at least one occasion in a psychiatric hospitalization that exacerbated Z.H.'s already-disturbed condition and behaviors. See id. at DFPS # 33580.

While in care, Z.H. attended at least seven different schools across six counties. His education records, though, appear to be incomplete and his IMPACT file contains conflicting notes. Z.H.'s Education Log in IMPACT does not reflect what schools he attended for the third, fifth, or sixth grades. See id. at DFPS009118757-58. His Education Log seems to indicate that Z.H. went directly from second grade to fourth grade, although there does not appear to be any explanation for that in the record. See id. at DFPS009118757-58. Notes in his IMPACT file also seem to say that Z.H. repeated the fourth grade in 2010-11, information absent from his Education Log. See id. at DFPS009124720-25. Caseworkers' notes in IMPACT indicate that Z.H. attended schools on campus at different RTCs between fall of 2010 and spring of 2013. See id. at DFPS009124637-25036. The notes, though, are inconsistent as to what grade Z.H. was in—some notes indicate that he was in sixth grade as early as September 2012, while others indicate that he was still in fifth grade as late as September 2013. See id. at DFPS009124903, 009125020. In the fall of 2013, Z.H. was in the seventh grade. See id. at DFPS009126275-94. In the fall of 2014, Z.H. entered the eighth grade. See id. at DFPS009124956-65. When Carter evaluated him, in the summer of 2014, Z.H. was unable to read a second grade level book. (D.E. 326 at 99-100). He also had “poor impulse control” and was aggressive with both other children and himself. Id. at 100-01. Carter found that Z.H. did not have a good sense of attachment, was “well below normal” from a “cognitive and school achievement standpoint,” and was “seriously disturbed.” Id. at 100, 186. At the 2014 evaluation, Carter found that Z.H.'s condition had not appreciably improved since his 2012 evaluation. Id. at 186. Unfortunately, as Carter testified and the Court so finds, Z.H.'s experience is “typical ... of the entire foster care system in the State of Texas,” especially in the PMC. See id. at 132-33, 194.

h. K.E.

K.E. is a former class representative for the General Class and the Licensed Foster Care Subclass.³⁸ At the time of trial, K.E. was 19 years old. She aged out of care in 2013. (DX 120 at 1 RFP CPS 141226 (filed under seal)). She was removed from her mother's home in Odessa, Texas in 2003 at age eight because of neglectful supervision, emotional abuse, and physical neglect. Id. at DFPS # 44605, 1 RFP CPS 140538-47. During her ten years in care, K.E. had over 25 primary and secondary caseworkers. See id. at 1 RFP CPS 141859-62. She had at least 27 placements in a variety of settings, including psychiatric hospitals, therapeutic foster family homes, therapeutic foster group homes, emergency shelters, GROs, and RTCs. Id. at 1 RFP CPS 141120, 142037, 1141778-81, 140191-92. K.E. also attended at least 15 schools all over the State. Id. As with Z.H., K.E.'s Education Logs are incomplete. There is no record of K.E.'s sixth or eighth grades and there are date gaps after leaving one school and beginning another. See, e.g. , id. at 1 RFP CPS 142602-06; 019904-07, 092721-25. In the 2008-09 school year alone, K.E. “was in 4 school districts.” Id. at DFPS # 47130.

38 The Court holds that certain Named Plaintiffs are no longer adequate class representatives. See infra Section III.

The State placed K.E. and her three half-brothers in foster care at the same time in 2003. During the initial investigation, her brothers admitted to physically abusing K.E, which K.E.'s mother confirmed. See id. at 1 RFP CPS 140545-46. K.E.'s father was incarcerated at the time, and K.E.'s mother reported that K.E. was abused when she lived with her father. Id. at 1 RFP CPS 140547. DFPS found Reason To Believe that K.E.'s mother neglectfully supervised all four children and that K.E. was physically and emotionally abused. Id. at 1 RFP CPS 140717.

K.E. was initially placed in a therapeutic foster home with one of the brothers who had physically abused her.³⁹ She spent three months in that placement, during which time she was hospitalized twice for being “out of control.” Id. at 1 RFP CPS 142037. In June 2003, after her second hospitalization, CPS raised K.E.'s level of care to Intense and moved her to an emergency shelter. See id. CPS then placed her in an RTC approximately 270 miles from her home county. Id. at 1 RFP CPS 141120. Not until after that placement change did CPS disallow visits between K.E. and her brothers because of the earlier abuse. See id. at 1 RFP CPS 140594. In August 2004, K.E.'s three half-brothers were placed in the conservatorship of their biological father and his wife.⁴⁰ Id. at DFPS # 46830-42. K.E. was placed in Texas's PMC. See id.

39 The other two brothers were placed in a different therapeutic foster home.

40 The half-brothers had the same mother as K.E., but a different father who was married. The boys were placed with their father and step-mother. Both of K.E.'s parents were imprisoned at that time. K.E.'s mother was unavailable due to her incarceration, half-way house term, and home confinement for 2 years. Id. at DFPS # 45352.

CPS was notified that K.E. was sexually abused, likely by multiple perpetrators, before entering care. Yet her case files do not indicate to what extent any of those allegations were investigated. K.E.'s first Common Application noted that she may have been sexually abused by her biological father's girlfriend. Id. at 1 RFP CPS 140710. Shortly after being placed at an RTC in late 2003, K.E. made an outcry of sexual abuse against her brothers. See id. at 1 RFP CPS 140726, 141829. Yet nowhere in K.E.'s 38,000 pages of case files is there mention of a sexual abuse investigation. There is also no record of K.E. receiving a medical examination in connection with either of those allegations. Her next Common Application, which appears to have been updated soon after her first one, stated that K.E.'s sexual abuse by two of her brothers had been confirmed.⁴¹ Id. at 1 RFP CPS 140726-29. The application also noted “sexual abuse by brothers and other adults” as a factor to consider in determining K.E.'s level of care. Id. at 1 RFP CPS 140728. K.E. also made an outcry of sexual abuse against her mother shortly after she entered PMC in late 2004. Id. at 1 RFP CPS 141833. There is also no documentation of any investigation into this allegation.

41 The application also stated that K.E. had to be removed from the home because the brothers who sexually abused her still lived there. Her brothers were actually removed from the home and placed in foster care at the same time as K.E. Id. at 1 RFP CPS 140726-29; 140538-47.

Despite documentation of sexual abuse allegations in 2003 and 2004, K.E.'s September 2011 service plan review stated, “No determination was ever made with regards to whether the [sexual abuse by her brothers] ever took place,” although K.E. “insist[ed] that it did take place.” Id. at 1 RFP CPS 140646. It is unclear whether the inconsistencies in K.E.'s record are a result of sexual abuse investigations being overlooked, sloppy and careless caseworker notes, constant caseworker turnover, or some combination of the above. Not until 2011 did a caseworker notice that K.E.'s files did not contain documentation of an investigation into the alleged sexual abuse by her brothers, which prior caseworkers accepted had occurred. It seems that no caseworker read K.E.'s case files all the way through during her first eight years in care or else they would have noticed the absence of this critical record.

K.E.'s Common Applications are similarly inconsistent. When K.E. was initially removed from her mother's home because of neglect, emotional abuse, and physical abuse, she made no allegations of sexual abuse, nor was sexual abuse investigated. See id. at DFPS # 46137-49. Yet, by January 2009, K.E.'s Common Application stated that she had been removed because her mother failed to protect her from sexual abuse by K.E.'s brothers. See id. at 1 RFP CPS 140873. Even singular documents were contradictory. For example, in July 2009, K.E.'s I See You worker noted that K.E. has eight siblings but that “[o]nly one of [K.E.'s] siblings is in care.” Id. at 1 RFP CPS 141766-67. In the very next paragraph the worker noted that K.E.'s siblings “are all out of care.” Id. Those conflicting statements were copied and pasted into every contact narrative for the next five months. See id. at 1 RFP CPS 141767-75.

As K.E.'s caseworkers continued to change, incidents that were previously described as allegations came to be accepted as fact, while references to other incidents and allegations disappeared completely. For instance, in June 2007, K.E.'s I See You worker e-mailed K.E.'s primary caseworker to inform her that another child in K.E.'s foster home placement hit K.E. and “bloodied her nose.” Id. at 1 RFP CPS 141798. There is no further mention of this incident, or any resulting medical attention, in the record. Further, in May 2012, K.E.'s caseworker noted that K.E. received stitches in her hand at the emergency room, after she broke a window at her RTC. Id. at 1 RFP CPS 141661. Later that month, when K.E.'s I See You worker visited, the worker's notes do not mention any injuries but instead state that K.E. had no marks or bruises. Id. at 1 RFP CPS 141662. Neither worker appears to have followed up with the RTC about the incident or K.E.'s medical treatment.

Inadequate casework also contributed to DFPS's failure to find K.E. a permanent home. Between mid-2003 and March 2005, DFPS identified K.E.'s paternal aunt as a possible kinship placement. Separate caseworker notes during this time said that the aunt would like to be considered as a placement, that she would like to get to know K.E. and have K.E. live with her, that she would not provide a home for her, that the aunt could not care for K.E., and that the aunt made no effort to get to know K.E. Id. at 1 RFP CPS 140577, 140595, 141066, 141095, 141807, 141810, 018073019238-39, 019261, DFPS # 45354, 46685, 46718, 46846-47. After 2005, K.E.'s aunt is no longer mentioned. The case files do not indicate a home study was ever conducted for K.E.'s aunt.

More disappointing, DFPS appears to have given up on K.E. almost as soon as she entered care. Her second Common Application, updated in late 2003 when K.E. was eight years old, stated, “Long term foster care, [K.E.] will likely age out of care.” Id. at 1 RFP CPS 140729. In March 2005, when K.E. was ten years old, DFPS officially changed her long-term goal to APPLA, six years before DFPS policy allows. Id. at 1 RFP CPS 140593. K.E.'s October 2005 service plan review noted that adoption was not possible due to her “emotional/behavioral problems.” Id. at 1 RFP CPS 140597-99. K.E. received no other adoption-related services or efforts until 2009, when only after a court order did a caseworker note that K.E. would meet with staff from the adoption unit to “discuss adoption and everything that entails.” See id. at 1 RFP CPS 140623-25. Nothing came of that meeting; no other adoption efforts were made for K.E.

K.E. experienced little placement stability while in care. The only exception was a two and a half year stay at her first RTC, between September 2003 and March 2006. See id. at 1 RFP CPS 141120. K.E. was removed from that placement because the RTC felt that she had “not made any improvements while there.” Id. at 1 RFP CPS 141778. When K.E. initially learned that CPS was looking for a different placement for her, she became significantly more aggressive. RTC staff reported that she was “trying to sabotage the placement in order to stay because [she was] afraid to leave.” Id. DFPS still moved her 270 miles to a different RTC, more than 500 miles from her home county, in the middle of the school year. Id. at 1 RFP CPS 141779-81. After that move, K.E. experienced 22 different placements over the following seven years, never spending more than a few months in any facility. See id. at 1 RFP CPS 140191-92.

Further, many of K.E.'s placements did not comport with her needs, exposing her to multiple incidents of sexual and physical abuse. During most of K.E.'s time in care, she displayed “extreme behaviors that require[d] 24-hour supervision.” See, e.g. , 1 RFP CPS 140591, 141781, 141844, 142748, 142809. Yet on multiple occasions, K.E. appeared to have no supervision at all. For example, in February 2004, DFPS was notified of allegations of inappropriate sexual conduct, and possible sexual abuse, between K.E. (eight years old at the time) and three other female residents of her RTC, including allegations of oral sex and anal penetration. Id. at DFPS # 46603. The intake report reflected that the whereabouts of the staff on duty at the time was “unknown.” Id. K.E.'s case files contain no documentation or mention of an investigation into these allegations.

That same year, in September 2004, DFPS was notified of different allegations of inappropriate sexual conduct between K.E. (nine years old at the time) and a 14 year-old female resident of the RTC that included digital penetration, which both children admitted. See id. at 1 RFP 7 RCCL 02151-58. The intake report notes that both children had a history of sexual acting out. Id. at 1 RFP 7 RCCL 02154. RCCL investigated those allegations for possible neglectful supervision, including making “contact with” K.E., the other child, and multiple RTC staff members, but the “contact” is not described or summarized in the report. See id. 1 RFP 7 RCCL 02158. The investigation closed five days after it was initiated. Id. The report stated:

The alleged neglectful supervision was ruled out, because a preponderance of evidence does not exist to support that the alleged neglectful supervision occurred. The alleged perpetrator was an unknown alleged facility staff member. The specific date and time frame of the alleged incident was unknown. The alleged victims did not have a history of being sexual perpetrators. The alleged victims have been separated, a safety plan has been put in place, and they were receiving therapy. During the investigation, no indications of abuse or neglect were found.

Id. Numerous statements in this report, including that K.E. and the other child involved did not have a history of being sexual perpetrators, are misleading. DFPS refuses to track child-on-child abuse, thus labeling children as sexual abuse perpetrators, instead investigating child-on-child abuse for neglectful supervision. Therefore, it is impossible that any child in DFPS's system would “have a history of being sexual perpetrators.” In the end, DFPS placed two children with a sexual abuse history in close proximity (even though these children need single-child homes), both children acknowledged the sexual incident, and RCCL Ruled Out any violation.

It is also not clear if K.E.'s caseworkers were notified of incidents of sexualized conduct and possible sexual abuse. There are no references to the 2004 incidents in K.E.'s primary or secondary caseworker's notes in. K.E.'s service level throughout 2004 was Intense. See id. at 1 RFP CPS 140315-21.

After the 2004 incidents, as happens to many sexual abuse victims, K.E. became a sexual aggressor. See, e.g. id. at 1 RFP CPS 141609, 141676, 141767. Yet her case files do not indicate if she received therapeutic treatment to address sexual abuse and sexualized behavior until late 2008. See id. at 1 RFP CPS 141756. Not until 2009, is “sexual acting out” mentioned in her comprehensive treatment plan, which already listed “anger/aggression/assaultive behavior ... suicidal ideation/gestures ... depression ... oppositional disorder ... [and] impulse control.” Id. at DFPS # 48311-15. DFPS changed her placement in mid-2009 “due to sexually acting out.” Id. at 1 RFP CPS 141676; 141265-66. DFPS changed her placement again in December 2011 after she sexually assaulted a resident at one of her RTCs. Id. at 1 RFP CPS 141643.

K.E. also suffered multiple instances of child-on-child physical abuse while in care, eventually becoming an aggressor in that regard as well. See e.g. , id. at 1 RFP CPS 141756, 141778, 142756, 142780. Carter testified that K.E. became an aggressor toward others because she had been “victimized considerably” while in State care. (See D.E. 326 at 103-07). There was also at least one allegation by K.E. that she had been physically abused during a restraint by an RTC staff member; that allegation was adequately investigated before being Ruled Out. (DX 120 at RCCL000000289-428 (filed under seal)).

When K.E. entered care, she had already been diagnosed with Explosive Disorder, for which her mother received Social Security benefits. Id. at 1 RFP CPS 140546-47. According to her second grade teacher, K.E.'s Explosive Disorder was a result of environmental, not genetic, factors. Id. In 2003, during K.E.'s initial evaluation, a psychologist diagnosed her with Conduct Disorder, ADHD, and Learning Disorder. Id. at 1 RFP CPS 140578. The psychologist also noted K.E.'s prior diagnosis of Explosive Disorder. K.E.'s initial level of care was Specialized, though it fluctuated primarily between Specialized and Intense throughout her time in care. See id. at 1 RFP CPS 140306-415. By 2008, K.E. was diagnosed with Oppositional Defiant Disorder, Bipolar Disorder, ADHD, Learning Disorder (not otherwise specified) and Relational Problem (not otherwise specified), although she was no longer diagnosed with Explosive Disorder. Id. at DFPS # 47779. Two months later, though, upon admission to a psychiatric hospital, she was only diagnosed with Bipolar Depression and “Questionable PTSD.” Id. at DFPS # 47785.

K.E.'s education also suffered while she was in care. Educational disruptions occurred primarily after 2006 when K.E. was discharged from her first RTC. Her initial service plan stated that she was “developmental[ly] on target.” Id. at 1 RFP CPS 140579. When K.E. was in the second grade, she performed at a first grade level, but her teacher told CPS that K.E. would “make [educational] progress with the right stimulation.” Id. at 1 RFP CPS 140546-47. In third grade, K.E. had “A's and B's in all subjects except science.” Id. at 1 RFP CPS 141825. Her caseworker noted that K.E. maintained “very good” grades and had no absences in the fourth grade. Id. at 1 RFP CPS 141845. After 2006, however, her intellectual functioning deteriorated, although it is difficult to gauge because of her refusal to cooperate with psychological testing. See id. at DFPS # 47696-98, 47702-05, 47770-77, 47864. By 2009, at age 14, K.E. was academically stalled at a second-grade level. See id. at DFPS # 48306-11. She failed the tenth grade due to excessive absences caused by psychiatric hospital stays, and a refusal to do make-up work. See id. at 1 RFP CPS 141643-44, 141723.

Carter testified that K.E. had been in so many placements that she stopped counting because the number no longer mattered to her. (D.E. 326 at 104). K.E. was often removed from placements for behaviors that her caregivers could not handle, placed in more restrictive environments until her level of care would drop, and then moved back to less restrictive environments, whereupon the cycle repeated. See id. at DFPS # 44886-938. When K.E. was in foster home placements, rather than RTCs, K.E. was never placed in a single-child home, as Defendants' and Plaintiffs' experts testified that a child with a history of physical and sexual abuse and aggression should have been. (See D.E. 328 at 161; D.E. 329 at 161; D.E. 303 at 8). When placed at RTCs, K.E. often shared rooms with other girls. In March of 2013, after ten years in care, K.E. turned 18 and aged out of the system. Id. at 1 RFP CPS 141226. In short, K.E.'s time in care was marked by educational disruptions, abuse, and numerous placements changes often to facilities unfit for her needs. Unfortunately, as Carter testified and the Court so finds, K.E.'s experience is “typical ... of the entire foster care system in the State of Texas,” especially in the PMC. (See D.E. 326 at 132-33, 194).

i. L.H. and C.H.

L.H. and C.H. are former class representatives for the General Class, the Licensed Foster Care Subclass, and the Basic Care GRO Subclass.⁴² L.H. and C.H. are sisters who are part of an eight-sibling group from San Benito, Texas. They were 16 and 12 years old respectively at the time of trial. See id. at 1 RFP CPS 115993. The sibling group's case files in the record, numbering approximately 16,500 pages, are undoubtedly incomplete. First, no RCCL files, which are kept in the CLASS database, were provided.⁴³ Second, the case files for this eight-member group are significantly shorter than those for a single foster child (e.g., M.D.'s was 33,000 pages, J.S.'s was 40,000 pages), which suggest that the IMPACT and External files are incomplete. All eight children are discussed at once even though they were often placed separately and had different caseworkers, making the files difficult to follow. Both L.H. and C.H. received Basic-level services for most of their time in care, although C.H. spent some time at Moderate. See id. at 1 RFP CPS 131917-47, 132327.

42 The Court holds that certain Named Plaintiffs are no longer adequate class representatives. See infra Section III.

43 RCCL files pertaining to abuse and neglect investigations are organized in CLASS, which stands for Child Care Licensing Automation Support System. CLASS is the electronic case file system utilized by CCL (similar to the IMPACT system utilized by CPS). (D.E. 328 at 112-13; see also DFPS, Definition of Terms, http://www.dfps.state.tx.us/handbooks/Licensing/Files/LPPH_px_ Definitions_of_Terms.asp (last visited December 16, 2015).

DFPS became involved in L.H.'s life in 2001 when CPS found RTB that her mother neglectfully supervised L.H., her older brother, and younger sister. Id. at DFPS009984773. CPS provided services until September 2002. During that time, CPS again found RTB neglectful supervision of L.H. and her two siblings. Id. In February 2003, a report alleged that four children, including L.H. and C.H., were left at home unsupervised and were hungry and dirty. Id. at 1 RFP CPS 122869, 122877. The children were removed that day, but were returned in April 2004. Id. at 1 RFP CPS 122961, DFPS009984773. The children were removed again in November 2004 when CPS received a report that the mother struck C.H., who was less than two years old, due to frustration with her eldest son. Id. 1 RFP CPS 122905. The children were returned to the mother in November 2005. The children cycled in and out of their mother's care until CPS removed them in May 2008, moving them into separate emergency placements. Id. at DFPS009984766-832, 1 RFP CPS 115863, 122099, 123292-93, 123347, 123506, 123533, 125554. L.H. and C.H. entered Texas's PMC in April 2009. Id. at 1 RFP CPS 130056-57. L.H. and C.H. shuttled through kinship placements (including their mother), foster family homes, foster group homes, and a GRO until they returned to their mother in May 2013. Id. at 1 RFP CPS 123413-17, 123441-44, 123480-513, 123529-37, 123626-33, 125427. Unlike many of the Named Plaintiffs, L.H. and C.H. remained geographically close to, and in touch with, their siblings and family while in care. See e.g. id. at 1 RFP CPS 116582, 116668-709, 122322.

After she was removed in 2003, L.H. claimed that her brother physically abused her and her sister and that he “used to pick up her shirt and pull down her pants and touch her.” It is unclear if these allegations were formally investigated. In 2003, CPS did investigate an incident, but Ruled Out any misconduct. See id. at 1 RFP CPS 124782, 125007. L.H. was forensically interviewed, but did not “make any indication that she was ever sexually abused” and “seemed to change the subject often to avoid answering some questions.” Id. at 1 RFP CPS 124838. Notes elsewhere in the case files indicate that CPS confirmed that the brother had “sexually molest[ed] other children,” without specifying whom, and was himself a victim of sexual abuse. See id. at DFPS009984771, 1 RFP CPS 124667.

In July 2008, L.H. was diagnosed with General Anxiety Disorder. Id. at 1 RFP CPS 131531. In August 2008, one of the siblings made allegations of sexual abuse against their mother's boyfriend stating that the boyfriend had bitten her and touched her inappropriately. Id. at 1 RFP CPS 120562, 120568. Another sibling alleged that she saw her mother and her mother's boyfriend “take off their clothes and start kissing” and that she saw the boyfriend touch three of the other siblings inappropriately, including C.H. Id. at 1 RFP CPS 120568-69. Following an investigation, CPS found RTB neglectful supervision and sexual abuse of two of the siblings (not including C.H.), but R/O abuse and neglect as to the other children. See id. at 1 RFP CPS 117218, 116168. The investigator noted significant risk factors in the mother's home, but overall found little risk of further abuse because the children were in foster care and the mother's boyfriend had been arrested. Id. Their mother continued to be allowed unsupervised weekend and overnight visits with her children. In September 2008, at the age of six, C.H. was placed in a therapeutic foster home separate from her siblings because she “bullied” them. Id. at 1 RFP CPS 123295. In April 2009, C.H. was diagnosed with ADHD, Anxiety, and Mood Disorder. Id. at 1 RFP CPS 131533.

In May 2009, a new caseworker retrieved C.H. and her sister from an overnight visitation with their mother. Id. at 1 RFP CPS 120577. Both girls made a sexual abuse outcry against their mother. Id. Although the caseworker appeared to be aware that the boyfriend was not supposed to have access to the children, she did not know about the past sexual abuse allegations. See id. The boyfriend told investigators that the mother allowed him access to her children and admitted that he and the mother spent a night at a hotel with her children. Id. at DFPS00998257. The caseworker concluded that the children were at an ongoing risk of sexual abuse. See id. at 1 RFP CPS 120577. The mother and boyfriend both appear to have been arrested following the outcry. See id. at DFPS009985222.

CPS investigated the allegations and found RTB neglectful supervision of all eight siblings by their mother, but UTD sexual abuse of any of the children by either the mother or her boyfriend. See id. at DFPS009985257-58, 1 RFP CPS 116146-47. The children did not make a further outcry of abuse when interviewed by an investigator, but they also denied having had contact with their mother's boyfriend even though the boyfriend admitted he had spent nights with the children when they visited their mother on weekends. See id. at 1 RFP CPS 116168. In a post-investigation risk assessment report, CPS noted that the children's mother established a “pattern of abuse.” Id. at DFPS009985258. The report found that the mother created an extreme level of risk because she had “been deceitful [with] the Department,” had “allowed a previously RTB'd perpetrator around the children,” and continued “to have a relationship with a man who sexually abused her children.” Id. at DFPS009985261-62. In what appears to be a separate investigation of the same allegations, CPS found RTB sexual abuse of two of the siblings, but not L.H. or C.H., by the mother's boyfriend. See id. at DFPS009985217.

In June 2009, CPS moved C.H. and several siblings, but not L.H., into a therapeutic foster family home in Brownsville. Id. at 1 RFP CPS 131533. CPS returned L.H. and three siblings to their mother in October. Id. at 1 RFP CPS 131873. The following day, the mother moved into an emergency shelter in Brownsville with the children. Id. C.H. and three siblings remained in the foster home and had weekly visits with the other siblings and their mother. See id. at 1 RFP CPS 116157. In November 2009, C.H.'s foster parents requested that her level of care be increased because she misbehaved and struggled at school. Id. at 1 RFP CPS 131871.

In January 2010, a school nurse visited L.H. at home and found her and her siblings “outside barefoot” “without jackets.” Id. at 1 RFP CPS 116130. She also observed that they were dirty, appeared malnourished, and that there was little food in the home. Id. The nurse called statewide intake to report her observations. The intake report noted that the mother's boyfriend “was charged with abusing” L.H. and another sibling, but it was “unknown how [the boyfriend] was abusing” them. Id. at 1 RFP CPS 116170. The report also noted that the boyfriend was “not supposed to be around” the children, but was “still in the picture.” Id. at 116170. After an investigation, CPS Ruled Out all allegations of abuse and neglect, and notified the nurse that CPS “will not be offering services to the family at this time.” Id. at 1 RFP CPS 116132-34, 116171-74. During that investigation, L.H. informed the investigator that she was abused while in foster care seven years earlier by another child in the foster home and she reported the abuse to her foster mother, “but nothing was done about it.” Id. at 1 RFP CPS 116157. There is no mention of an investigation into those allegations.

In February 2010, L.H. was placed in a therapeutic foster group home, along with one of her younger brothers. Id. at 1 RFP CPS 123355-57. C.H. and two other siblings were not moved. The mother was still allowed to visit with all of her children each Saturday at a nearby mall. See, e.g. , id. at 1 RFP CPS 132057-65.

A new caseworker began working on L.H. and C.H.'s cases in July 2010. Id. at 1 RFP CPS 132124. At their first meeting, the mother said that the frequent caseworker changes made it difficult to know who was the actual caseworker. Id. at 1 RFP CPS 132123. On July 8, the children's therapist told the caseworker that the siblings needed to be moved all together, that C.H.'s foster mother was a bad parent, and that she was doing more damage to the children than the alternative of adjusting to a new environment with all the siblings. Id. at 1 RFP CPS 132128.

L.H., C.H., and all of their siblings except the oldest brother were placed in a GRO on July 19, 2010 despite being Basic-level children. Id. at 1 RFP CPS 123302. The GRO was the only placement available that allowed the sibling group to remain intact. Vasquez, the attorney ad litem and next friend for L.H. and C.H., advocated that they be placed together because their ongoing trauma would have worsened if DFPS disrupted their tight sibling relationships—the only source of stability in their lives. (D.E. 188 at 161). Vasquez visited L.H. and C.H. multiple times in that placement. Id. at 167-68. She described the GRO as “an institution, a facility that just house[d] them,” with blank walls and “no personalization.” Id. Vasquez testified that the GRO was not at all like a home, and that the children did not view it as one, but was instead “a box that [they were] thrown” into. Id.

In August 2010, L.H. and C.H.'s caseworker visited them at the GRO. (DX 120 at 1 RFP CPS 132147 (filed under seal)). C.H. was anxious about wanting to return to the foster home where she had previously lived and refused to discuss anything else. Id. L.H. also stated that she missed her previous foster parent, but otherwise informed her caseworker that she liked the new placement. Id. L.H. also discussed the GRO's many strict rules and very specific, detailed schedules for all of the children's activities. See id. The siblings spent most of the day in their own room and were not allowed to visit each other's rooms. Id. Two of their siblings struggled to adjust to the strict rules and the caseworker “tried to get [the caregiver] to be a little more lenient with the rules and be more understanding about [the children's] move and give them some time to adjust,” but the caregiver was adamant about his methods. Id. The children continued to have weekly visits with their mother. See id. at 1 RFP CPS 132216.

L.H. and three other siblings returned to their mother “as a monitored placement” in January 2012. Id. at 1 RFP CPS 132256. C.H. and two other siblings were placed with a maternal aunt who lived in the same city. Id. at 1 RFP CPS 123306. In March 2012, L.H. was arrested after a fight broke out between L.H. and her mother. Id. at 1 RFP CPS 120600. On the way to the police station, L.H. said that the fight started because her mother told L.H. to take nude photographs of her mother to send to the mother's boyfriend. Id. at 1 RFP CPS 125111. L.H. was released from police custody the same day. Id. CPS moved the children to their aunt's home a few days later. Id. at 1 RFP CPS 124564, 125115-16. CPS investigated the allegations, which the mother confirmed, and made an RTB disposition. Id. at 1 RFP CPS 121433, 125116.

In April 2012, L.H., C.H., and their siblings were reunited and living with their maternal aunt. See id. at 1 RFP CPS 132346. However, their mother was regularly present in the home. Their mother was arrested for Assault on a Minor/Family Member after she got in an argument with the aunt's 15-year-old daughter. See id. While in their aunt's care, the children reported that their cousins were verbally and physically aggressive toward them. Id. Their mother also alleged that her niece was abusing the children and using illegal drugs and had exposed L.H. to those drugs as well. Id. at 1 RFP CPS 132249. During an investigation into these allegations, L.H. admitted that she smoked with her cousin, but was not sure whether it was marijuana or a regular cigarette. Id. The caseworker had L.H. drug-tested and made arrangements to move the children. Id. In May 2012, L.H., C.H., and their siblings (except the oldest brother) were moved back to their previous GRO. Id. at 1 RFP CPS 132347-52.

In June 2012, the caseworker transported L.H. and C.H. to visit their uncle in Hitchcock, Texas, in hopes that they could be “transitioned” into a kinship placement with him. Id. at 1 RFP CPS 132337. It is unclear why DFPS wanted to place the children with the uncle, as an October 2010 home assessment recommended not placing the children with him due to, among other factors, his precarious financial situation, his unstable romantic relationship, and his uncertain immigration status. See id. at 1 RFP CPS 116539-56. It does not appear that another home assessment was completed. At one visit during the “transition stage,” the maternal uncle cut the visit short because C.H. would not stop crying. Id. at 1 RFP CPS 132194-95.

On August 28, 2012, a court terminated parental rights over L.H., C.H., and five of their siblings. Id. at 1 RFP CPS 122140, 132173-78. In October 2012, their mother's rights were reinstated after appeal, although the children were not to be returned to the mother at that time. Id. The children continued to live at the GRO with reunification and alternative relative placement as their permanency goals. Id. at 1 RFP CPS 132184, 122139. In December, CPS allowed visitation with their mother. Id. at 1 RFP CPS 132170. That month the children and their mother flew to Houston to spend Christmas break with the uncle. Id. The Department's plan was to place the children with their uncle pending a new home study. Id. During the extended visit, however, the uncle locked L.H. and C.H.'s younger sister in the bathroom for an unknown amount of time after she started to cry during a scary movie, and the uncle threatened to hit her with a riding crop. Id. at 1 RFP CPS 120614. The caseworker's report from that date expressed “grave concern regarding these reported statements at the children's uncle's home, to which they will be going within [the] month.” Id. at 1 RFP CPS 120617. An investigation was opened, but the service plan for L.H. and C.H. completed on January 16, 2013, made no mention of the allegations or impending change in the children's placement. Id. at 1 RFP CPS 122140.

In February 2013, the caseworker noted that the children had mixed feelings about living with their uncle because he “curses too much and is very strict.” Id. at 1 RFP CPS 132371. Nonetheless, the department still sought to place the L.H., C.H., and their siblings with their uncle, who stated his willingness to take care of all of children except the eldest brother. Id. at 1 RFP CPS 132370-76. Additionally, the caseworker noted that their mother moved into the uncle's house. Id. at 1 RFP CPS 132354. In March, the caseworker wrote that the children were not moved into the uncle's home as planned due to the open investigation, which made the children, especially L.H., “visibly upset.” See id. at 1 RFP CPS 132354. After the investigator Ruled Out allegations of abuse, CPS again moved toward placing the children with their uncle. See id. at 1 RFP CPS 132355.

On April 9, 2013, the children were placed with the uncle, who signed a kinship agreement. Id. at 1 RFP CPS 123309, 132378. Days later, the uncle reported to CPS that the children's mother had taken all of the children to another relative's home and did not bring their medication. Id. at 1 RFP CPS 132380. On April 30, 2013, the uncle contacted the caseworker and said he no longer wanted legal custody of the children. Id. at 1 RFP CPS 132379. That same day, C.H., L.H., and their siblings were placed with their mother. Id. at 1 RFP CPS 123318. The report for each child stated:

This placement is not the most appropriate to meet this child's needs but there are no other acceptable placement[s]. The Department has diligently made efforts to place with relatives, terminated [sic] parent's rights but without success. Court has ordered that children remain with [their mother] and the Department will continue to monitor until dismissal date of 6/14/13.

See, e.g. , id. at 1 RFP CPS 123318. The report for each child further noted that it was “unknown if this placement will be able to meet the child's long term needs” and that DFPS “was not in favor [of leaving] the children with [their mother]; however, court and attorneys were all in agreement that no other placement was available and it would be too traumatic to remove the children and place in foster care.” Id. On May 8, 2013, the caseworker appears to have visited the children for the last time and noted that the mother continued “to have a hard time ensuring that her house is well kept” and “to demonstrate that she is not able to cope with the number of children that she has.” Id. at 1 RFP CPS 121564.

Between L.H.'s first removal from her mother in 2002, and her eventual return in 2013, she experienced 13 placements. C.H., who was born in 2002, had 15 placements over the same period. Despite being Basic-level children, they spent a significant amount of time in a GRO. L.H. attended at least 12 schools between pre-kindergarten and eighth grade, including three different schools during seventh grade. C.H. attended at least eight schools between prekindergarten and fourth grade, including five different schools during fourth grade. Between 2008 and 2013, L.H. and C.H. had at least eight primary and secondary caseworkers. See id. at 1 RFP CPS 125069-843. Their caseworkers did not visit them regularly, and often wrote short, recycled, and uninformative notes after visits. See, e.g., id. at 1 RFP CPS 131940, 132062, 132223-93, 132322-76. Carter did not review L.H.'s or C.H.'s case files, or interview them.

j. J.R., M.R., and S.R.

J.R., M.R., and S.R. are former class representatives for the General Class.⁴⁴ J.R. (six months old), M.R. (five years old), and S.R. (six years old) were removed from their mother in June 2010 after J.R. was hospitalized with seizures because her mother failed to properly mix her baby formula. Id. at 1 RFP CPS 71031-37. The mother was determined to have physically neglected and neglectfully supervised J.R. Id. at 1 RFP CPS 71079. The three children were in foster care for approximately three years before they were adopted by their maternal aunt. Yet, during that relatively short period, they had 19 caseworkers.

44 The Court holds that certain Named Plaintiffs are no longer adequate class representatives. See infra Section III.

The three siblings were usually placed together, although not initially. After being discharged from the hospital, infant J.R. was placed with a relative, but was moved within ten days due to a risk of abuse. She was placed with a therapeutic foster family, but after four days of continuous crying, the family requested her removal. J.R. was then placed in an emergency shelter at a San Antonio GRO until her aunt could take care of her. Id. at 1 RFP CPS 71031-37. S.R. was originally placed at an emergency shelter at a GRO in Tyler, Texas. Id. at 1 RFP CPS 71407. M.R. was originally placed with a relative in Houston, but was quickly moved to his maternal aunt's home. Id. at 1 RFP CPS 71419-23. Each child spent fewer than six months in care with someone other than a family member. Except for emergency placements, J.R. and S.R. were placed in the Houston area. M.R.'s placements were all in Houston. Because they were so young while in care, their education was not disturbed.

CPS determined that the children's mother was unable to care for them due to her own mental health problems. Both S.R. and M.R. were Basic-level children, while J.R. was Moderate because of her congenital heart defect and neurological problems. Id. at 1 RFP CPS 71555. S.R. was diagnosed with developmental problems of unknown cause, in part because she did not speak at age five. S.R. and J.R. were both special needs children and characterized as autistic. Id. at 1 RFP CPS 72480. M.R. had no health issues. Id. at 1 RFP CPS 71686. Genetic testing was recommended for S.R. in July 2010 and ordered by the Court in August 2010, but the records do not reflect that the testing was ever performed. Psychological evaluations were recognized as necessary for all the children in June 2010, but evaluations were not performed until June 2012. See Id. at 1 RFP CPS 72498, 075450-55. The evaluation determined that S.R. had a full scale IQ of 50, along with significant developmental delays and was diagnosed with ADHD and mental retardation. Id. J.R was determined to be functioning at an 11-month-old level, although she was two and a half years old. Id. at 1 RFP CPS 075705-09. J.R. was diagnosed with Pervasive Development Disorder.

In August 2011, one of the children's caseworkers stated in S.R.'s Child's Service Plan, “caseworker will ensure that child is referred for a psychological assessment to ensure that all of his [sic] mental health needs are met.” Id. at 1 RFP CPS 108213. The next Plan, done by a different worker, in December 2011 says, “should the child require additional/testing evaluations, the caregiver will schedule the appointment and provide the agency with the completed report.” The Plan does not indicate that any mental health evaluations took place in the three months between Plans, nor is there a report in the files. Id. at 1 RFP CPS 108318-23. Similar disjointed records exist for J.R. See id. at 1 RFP CPS 108349-60. It appears that as the caseworkers changed, the need for psychological evaluations was overlooked until the change of placement to a foster family caused the issue to come up again. Unfortunately, as Carter testified and the Court so finds, J.R.'s, M.R.'s, and S.R.'s experiences are “typical ... of the entire foster care system in the State of Texas,” especially in the PMC. (See D.E. at 132-33, 188, 194).

k. S.S.

S.S. is a former class representative for the General Class.⁴⁵ At the time of trial, S.S. was 11 years old. The records provided to the Court on S.S. are incomplete. They are only 776 pages long, which includes files on her father and one of her brothers. The Court was supplied with only External files, no IMPACT or RCCL files. As a result, the Court's ability to evaluate the casework for S.S., as well as her experience in foster care, is substantially impaired.

45 The Court holds that certain Named Plaintiffs are no longer adequate class representatives. See infra Section III.

S.S. was eight years old in 2011 when she made an outcry that her 14-year-old brother had been sexually abusing her for four years. Id. at 1 RFP CPS 163969-79. S.S.'s parents contacted the police and CPS after the mother had S.S. examined by a doctor in Mexico. Id. at 1 RFP CPS 163861. The family lived in the Brownsville area; the parents were separated and the children lived with their father. During CPs's investigation, S.S. lived with an aunt and uncle in Los Fresnos, Texas, and her brother was in a Houston emergency shelter. Id. at 1 RFP CPS 163912-13, 164027.

S.S.'s August 2011 Placement Summary reflected an immediate need for counseling. Id. at 1 RFP CPS 164464-66. That same report identified a therapist and states, “needs initial visit.” Id. Yet the Service Plan for S.S. dated September 12, 2011 does not mention any therapy or counseling. It only notes that S.S. is being “watched closely to determine if she has any specialized emotional needs as a result of being sexually abused.” Id. at 1 RFP CPS 164015. A year later in June 2012, S.S. still did not have regular therapy even though it was needed “to address any emotional or behavioral effects that [S.S.] may be experiencing as a result of the abuse she suffered ....” Id. at 1 RFP CPS 164153. The Court cannot determine whether the absence of therapy for an eight-year-old girl who was sexually abused for half her life was the result of overloaded caseworkers or otherwise. The available records do not reflect the frequency of S.S.'s contact with her caseworkers or any effort to arrange therapy.

The records do reflect that DFPS found RTB neglectful supervision by the father as to S.S. and one of her brothers, RTB medical neglect by the father as to one of S.S.'s brothers, and RTB sexual abuse by the father as to one of S.S.'s brothers when he was five years old. Id. at 1 RFP CPS 164052, 163932. Yet DFPS did not oppose returning the son to the father, as long as S.S. was not present. See id. at 1 RFP CPS 164046-50. S.S.'s brother was returned to the father's custody by Court Order dated June 20, 2012. Id. at 1 RFP CPS 164161-65.

S.S. entered PMC in June 2012. A year later, S.S. was moved to the substitute care of her aunt and uncle with standard visitation by her father. Id. at 1 RFP CPS 164221-36. It is not clear if S.S. was the only child in her aunt's and uncle's home, as Defendants' and Plaintiffs' experts agreed is necessary for sexually abused children. (See D.E. 328 at 161; D.E. 329 at 161; D.E. 303 at 8). Overall, S.S. was in care for two years and was placed within 15 miles of her father and siblings. Id. at 1 RFP CPS 164472. S.S. changed schools during the summer break in 2011 between first and second grade and remained in that school until third grade when she moved in with her aunt and uncle. Id. at 1 RFP CPS 164476. Carter did not review S.S.'s case files, or interview her.

l. A.R.

A.R. is a class representative for the General Class. At the time of trial, A.R. was two years old. Id. at 1 RFP CPS 162647. She was born prematurely to a mother who smoked marijuana while she was pregnant. A.R.'s parents did not marry and had a volatile relationship that exposed A.R. to drugs and to domestic chaos and arguments. A.R. came into foster care in November 2011, at five months old, because of a neglectful supervision complaint placed by hospital staff. A.R. was removed after she fell out of her car seat onto the pavement and sustained multiple injuries, including a head injury. A.R.'s mother did not strap A.R. into the car seat and was not using the car seat's carrying handle when the mother had a seizure and dropped the seat causing A.R. to fall into the street. A.R.'s mother had a long-standing seizure disorder, but did not take medication. Id. at DFPS009071852-53. DFPS found RTB neglectful supervision, but R/O physical abuse. Id. at DFPS009073216.

A.R. first moved into a foster family home, where she remained for a year and a half. Id. at 1 RFP CPS 168864. She then transferred to a kinship placement with her paternal uncle and his wife. Id. While there, A.R. developed extreme tantrums that included throwing objects, extreme anger with screaming fits, violent responses to being denied anything, and food stealing and overeating to an extreme degree. See, e.g. , id. at 1 RFP CPS 169153. The placement fell through after a year, and A.R. returned to her original foster family, where she stayed until that family adopted her in April 2014. Id. at 1 RFP CPS 168859-63, DFPS009074287.

During her time in care, A.R. was placed within 50 miles of her parents and saw them during supervised visits. A.R. had an older half-brother who was voluntarily placed with their maternal grandmother. A.R. had visits with him while she was in care. Her level of care was Basic much of the time, but increased to Moderate for a period. Id. at 1 RFP CPS 168865. As with the other Named Plaintiffs, A.R.'s case files were often incorrect and inconsistent. For example, one of A.R.'s Service Plans noted that she was an only child, even though the same Plan mentioned that she had a relationship with her brother and identified him by name. Compare id. at 1 RFP CPS 168833, 168837, with 168856.

Unlike many of the Named Plaintiffs, A.R.'s time in foster care was relatively short. She was adopted three and a half years after being placed in care. A.R. had the same caseworker for most of her time in care, but when A.R.'s case was transferred to a new caseworker, the worker admitted to the foster family that she had not seen A.R.'s file. See DFPS009074339-41. Carter did not review A.R.'s case files, or interview her.

III. PRELIMINARY ISSUES

Defendants contend that this lawsuit is precluded for a number of reasons: (1) Named Plaintiffs lack standing; (2) class certification is improper; (3) this Court should follow the abstention doctrine; and (4) the suit is moot because some Named Plaintiffs are no longer in care. The Court reaffirms its previous rulings on standing, class certification, and abstention, see M.D. II , 294 F.R.D. 7, and finds that the claims of the General Class, the Licensed Foster Care Subclass, and the Foster Group Home Subclass are not moot.

In their post-trial brief, Defendants seem to assert that Plaintiffs lack standing. However, they fail to make an actual argument to this effect. (D.E. 359 at 51-52). Instead, without an explanation, Defendants cite three seminal standing cases and declare that Plaintiffs do not meet their requirements. Id. The Court is not convinced and reasserts its finding that Named Plaintiffs have standing. In a similar cursory manner, Defendants revisit their class certification and abstention arguments, which this Court has already decided. Id. at 50-52. The Court certified the General Class and three of four subclasses after conducting the “rigorous analysis” that Fed. R. Civ. P. 23(a) and Wal – Mart require. M.D. II , 294 F.R.D. 7. Defendants have not given this Court any reason to amend its prior decision. Likewise, Defendants have not presented new evidence or arguments that convince the Court to overturn its holding that abstention is not appropriate under Younger or Burford . See M.D. v. Perry , 799 F.Supp.2d 712 (S.D.Tex.2011).

Defendants raised one new issue, intimating that some Named Plaintiffs are no longer appropriate class representatives. (D.E. 360 at 47-49). Indeed, some have been removed from foster care. See id. Generally, that would moot those Named Plaintiffs' claims and preclude them from meeting the “case or controversy” requirement for standing. See U.S. Parole Comm'n v. Geraghty , 445 U.S. 388, 396, 100 S.Ct. 1202, 63 L.Ed.2d 479 (1980) ; Fontenot v. McCraw , 777 F.3d 741, 747 (5th Cir.2015). Named Plaintiffs who lack standing are usually dismissed from the case and cannot serve as class representatives, thus causing class decertification. The Court holds that decertification is not proper for the General Class, the Licensed Foster Care Subclass, or the Foster Group Home Subclass. However, decertification is proper for the Basic Care GRO Subclass.

Seven Named Plaintiffs—M.D., S.A., A.M., J.S., H.V., Z.H., and A.R.—representing the General Class, the Licensed Foster Care Subclass, and the Foster Group Home Subclass, were in their respective classes when they were certified on August 27, 2013. M.D. II , 294 F.R.D. 7. It is well settled that upon class certification, “the unnamed class members ‘acquired a legal status separate from the interest asserted by [the Named Plaintiffs].’ ” Zeidman v. J. Ray McDermott , 651 F.2d 1030, 1045 (5th Cir.1981) (quoting Sosna v. Iowa , 419 U.S. 393, 399, 95 S.Ct. 553, 42 L.Ed.2d 532 (1975) ). “In essence, the act of certification brought the unnamed members of the class before the court for Article III purposes; so long as their claims [are] justiciable, the mootness of the original named plaintiff's claim would not require dismissal.” Zeidman , 651 F.2d at 1045. The Court finds that the case and controversy remains “very much alive” for the unnamed class members who are still in the certified classes. Sosna , 419 U.S. at 401, 95 S.Ct. 553. As a result, the General Class, the Licensed Foster Care Subclass, and the Foster Group Home Subclass have adequate representatives and remain certified.

Before discussing the Basic Care GRO Subclass, the Court notes that D.I. remains an adequate representative of the General Class and the Licensed Foster Care Subclass under the relation back doctrine. The doctrine applies in two situations: (1) where named Plaintiffs' substantive claims are “by their nature ... so transitory that no single named plaintiff with such a claim could maintain a justiciable case long enough to procure a decision on class certification”; and (2) when individual named Plaintiffs' claims are mooted “by the defendant's purposive acts.” Zeidman , 651 F.2d at 1050. In the second case, the relation back doctrine prevents defendants from purposely avoiding litigation on legitimate class claims. See Sandoz v. Cingular Wireless , 553 F.3d 913, 919 (5th Cir.2008). The Court finds that the relation back doctrine applies here under the second exception. Defendants purposely “picked off” D.I. in the hope of mooting his claims. Zeidman , 651 F.2d at 1050. Buried in D.I.'s case files are handwritten notes on his May 30, 2012 Adoption Preparation Form, which read: “*He's under the Child's [sic] Rights Lawsuit” and “***(expedite adoption—will be nonsuited after ).” (PX 185 at 3079 (filed under seal)) (emphasis added). On D.I.'s May 31, 2012 Adoption Records Form under “special circumstances the redactor needs to be aware of,” the form states, “Please expedite. Call supervisor for details.” Id. This call to arms came after a May 2, 2012 status conference where the Court notified both parties that the class certification hearing would be scheduled around December 2012. D.I. was adopted on October 24, 2012. Likewise, Ricker testified that J.S. and H.V. both received significantly more attention after becoming Named Plaintiffs. (D.E. 326 at 219-20). J.S. was assigned two caseworkers, one of whom had a law degree. Id. at 220. CPS also managed, after much concerted effort, to locate an adoptive placement for J.S. even though a judge had called him “unadoptable” due to his behavior. (DX 120 at DFPS009059404 (filed under seal)). As part of that effort, one caseworker asked a supervisor for money to purchase “brochures with good quality photographs in color” of J.S. The supervisor needed to clear this request with the Program Director because CPS did “not normally do” that. Id. at 1 RFP CPS 054667-72. Although the Court would otherwise applaud DFPS for over-the-top efforts to find a troubled child a permanent home, the apparent motivation is disconcerting. The idea that Defendants “may short-circuit a class action” by undermining the standing of Named Plaintiffs “deserves short shrift.” Roper v. Consurve , 578 F.2d 1106, 1110 (5th Cir.1978).

The Court did not find this level of malfeasance toward L.H. and C.H., the Basic Care GRO Subclass representatives. The Court recognizes that L.H. and C.H. were placed with their uncle shortly before the Class Certification Order, four years after they entered foster care, and in spite of their caseworker expressing “grave concerns” about this placement because of an open investigation against the uncle for threatening to hit L.H.'s and C.H.'s younger sister with a riding crop. (DX 120 at 1 RFP CPS 120617 (filed under seal)). The Court also recognizes that days after moving in with their uncle, CPS signed off on transferring custody to the children's mother—for whom DFPS had found multiple RTB neglectful supervision of the children—even though, “This placement is not the most appropriate to meet th[e] child[ren]'s needs.” Id. at 1 RFP CPS 123318. The Court is wary about the timing and motivation of these events, but the Court has not seen sufficient evidence that DFPS purposively “picked off” L.H. and C.H. Therefore, without adequate representatives, the Basic Care GRO Subclass must be decertified.

The Court notes that eight proposed Named Plaintiffs—P.O., S.R., M.R., J.R., K.E., and S.S., L.H., and C.H.—are no longer adequate representatives. In addition to L.H. and C.H. leaving foster care in April 2013, P.O. was adopted by his grandparents in December 2012, custody over S.R., M.R., and J.R. transferred to their aunt in August 2012, K.E. aged out of care in March 2013, and custody over S.S. transferred to her aunt and uncle in August 2013. Some of these events are undoubtedly suspicious. P.O. was adopted by his grandparents eight years after entering PMC, one month before the class certification hearing, even though they were willing to adopt him much sooner. Likewise, DFPS transferred S.S. to her aunt days before of the Class Certification Order. Neither case file, however, reveals the blatant motivation present in D.I.'s records. That said, the Court does not need to determine if the State picked off these Named Plaintiffs. P.O. was a representative for the General Class and Licensed Foster Care Subclass, which are both certified without him; S.S. was a representative for the General Class, as well as the proposed Unverified Kinship Subclass, which the Court did not certify. Although these eight children are not adequate class representatives, the Court can rely on their foster care experiences as evidence of Plaintiffs' claims in the same way that the Court can rely on the experiences of former foster children Arce, Bentley, Jackson, Virgil, and Sharp. Their experiences are proper evidence of the case and controversy that is very much alive for the remaining certified classes.

IV. DISCUSSION

To succeed on their claims, Plaintiffs must demonstrate that the challenged policies or practices were “the moving force[s] behind, or actual cause[s] of, the constitutional injury.” Harris v. Serpas , 745 F.3d 767, 774 (5th Cir.2014) (internal quotations omitted). In this case, Plaintiffs must prove that Texas's policies or practices were the moving force behind, or actual cause of, the harm and unreasonable risk of harm to foster children in the certified classes. The “question of causation is intensely factual.” Morris v. Dearborne , 181 F.3d 657, 673 (5th Cir.1999). “Circumstantial evidence and reasonable inferences therefrom are a sufficient basis for a finding of causation.” Goodner v. Hyundai Motor Co., Ltd. , 650 F.3d 1034, 1044 (5th Cir.2011) (citation omitted). Where “fault and causation [are] obvious,” demonstrating that a public agency's policy or practice is unconstitutional is sufficient to establish the agency's liability. See Bd. of Cnty. Comm'rs of Bryan Cnty., Okla. v. Brown , 520 U.S. 397, 406, 117 S.Ct. 1382, 137 L.Ed.2d 626 (1997).

As discussed supra , Plaintiffs must also show that Texas's policies or practices that caused the unreasonable risk of harm were maintained with the requisite culpability—either deliberate indifference or substantial departure from professional judgment. See Section II.C.

A. General Class

The General Class consists of all children now, and in the future, in Texas's PMC. Plaintiffs allege that DFPS's policies toward its primary conservatorship caseworkers (“CVS caseworkers” or “primary caseworkers”) cause a number of interconnected problems, such as excessive workloads and frequent CVS caseworker turnover. These problems, Plaintiffs argue, result in an unreasonable risk of harm to the General Class in violation of their Fourteenth Amendment rights. The Court agrees.

CVS caseworkers “provide or coordinate services to families in which children have been removed from their homes and placed in the State's custody because of significant risks to their safety.” (DX 119 at 17). They are “the frontline workers of the child welfare system.” (PX 2037 at 7). CVS caseworkers are foster children's “lifeline,” their “connection to everything.” (D.E. 323 at 33). They are the principal DFPS employees tasked with ensuring the safety, permanency, and well-being of foster children. CVS caseworkers have been described as “the backbone of the State's effort to protect children” who “make life-and-death decisions every day,” (DX 119 at 17), “critical, critical to the provision of child safety, permanency and wellbeing,” and the “eyes and ears of the State for ... PMC children.” (D.E. 305 at 11-12).

CVS caseworkers' responsibilities are extensive. They are in charge of “matching children with the most appropriate and least restrictive placement, monitoring that children are safe and their needs are being met in their placements, ensuring that children receive needed services, and developing and implementing permanency plans for children so they exit custody to permanent families as quickly as possible.” (PX 2037 at 7-8). CVS caseworkers are also “responsible for monthly face-to-face contact with the child,” “the foster parents,” and “all other parties involved in the case,” including the child's caregivers and attorneys and guardians ad litem , to ensure that the child's needs are being met. (D.E. 324 at 11; PX 885 at 133). In addition, primary caseworkers attend court hearings and service plan meetings, keep children's medical records updated, prepare court documents, create monthly summaries, and document and maintain relationships with their foster children and the parties involved. (D.E. 324 at 11). At the core, a CVS caseworker's primary job is to “ensure the safety of children in foster care.” Id. Foster children look to their CVS caseworkers “for support, guidance, [and] understanding .... That's the person the child comes to rely on to get them through [the foster care] experience.” (D.E. 326 at 85). Without a CVS caseworker, foster children “would be at a clear risk of harm” because “watching for safety issues” is one of CVS caseworkers' “crucial responsibilities.” (D.E. 305 at 12). Specia testified that a foster child's primary CVS caseworker is “one of the most important workers” in the child's life. (D.E. 299 at 74). McCall explained that if CVS caseworkers “really are too busy,” foster children are at risk of harm. Id. at 12-13.

Before analyzing whether CVS caseworkers “really are too busy,” the Court notes that caseworker caseloads are still something of an open question despite years of litigation and weeks of trial. The problem is that 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, rather than by individual children. DFPS stages could 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. (DX 167). One 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. DFPS's way of counting caseloads is unique to Texas. (D.E. 302 at 50; D.E. 303 at 4-5). Defendants' and Plaintiffs' experts could barely understand the stage-counting approach, let alone explain it to the Court. (See D.E. 327 at 38-39; D.E. 325 at 124-25; D.E. 305 at 45-51). It is therefore difficult to compare DFPS caseworker caseloads to national and professional standards.

Dr. Miller, who ran the DFPS equivalents in Tennessee and Kentucky, established a primary caseworker caseload range of 14 to 17 children in Tennessee, with a maximum of 20. (D.E. 302 at 50-51). When caseloads exceed 20 children, Dr. Miller explained, the caseworkers' “work really begins to suffer and the children for whom that work is being done begin to suffer.” Id. Moreover, Dr. Miller said, setting reasonable caseload caps allows a child welfare agency to better handle caseworker turnover. Dr. Miller's 14 to 17 child caseload range provided that when a worker resigned or went on leave, their work could be distributed without causing caseloads to exceed 20, assuring that “none of those kids fell through the cracks.” Id. at 53.

Professional organizations agree with Dr. Miller. According to the 2007 Texas Appleseed report, “When those caseworkers are inadequately trained, inexperienced, or overburdened, the system breaks down and children in the system are harmed.” (PX 1966 at 11). Likewise, a 2014 audit of DFPS said, “The single most important improvement any system can make is to ensure it has a well-trained workforce with workloads that meet national standards.” (PX 1880 at 16). The CWLA, which Texas recognizes, see Tex. Gov't Code Ann. § 531.001(5), recommends a caseload range from 12 to 15 children. (PX 18 at § 3.48). The COA advocates caseloads of 8 to 15 children. (PX 2037 at 9-10). The Court would like to consider DFPS's own standards, but DFPS does not have any. (See, e.g. , D.E. 300 at 12; D.E. 311 at 17). Unlike other child welfare systems, DFPS “puts no limits on the caseload size that a conservatorship worker can carry.” (D.E. 305 at 25, 27-28).

That said, one of DFPS's Regional Directors agreed that “[a]n overloaded case worker is bad for the children they are supposed to protect.” (D.E. 311 at 21). Commissioner Specia admits that high caseloads “put[ ] a burden on the worker” and “can have a number of negative consequences.” (D.E. 299 at 55-56). DFPS also agrees that caseloads must be manageable in order to protect foster children. Before working for DFPS, Burstain said, “With caseloads at a manageable level, caseworkers could visit with children and families more often and ensure that children were safe and families were getting needed services.” (PX 1877 at 10). Likewise, “Lower caseloads reduce the paperwork load and allow [conservatorship] workers to spend more time helping families and children.” Id. Yet DFPS has not capped its caseloads. Most shocking, McCall, the DFPS officer in charge of making sure that caseworkers have manageable workloads, has “no idea what size child caseload [her] conservatorship workers should have in order to do their jobs properly.” (D.E. 305 at 31). Regardless, McCall agreed that overburdened CVS caseworkers create an unreasonable risk of harm for foster children. Id. at 12-13. Thus, national professional organizations, heads of other state's child welfare systems, both parties, and this Court agree that excessive caseloads cause an unreasonable risk of harm to foster children.

The Court does not need to establish the exact point at which caseloads become so excessive that they cause an unreasonable risk of harm. Dr. Miller believes that point is at 20 children. The CWLA and the COA believe that point is at 15. Wherever that point lies, DFPS has crossed it. As of July 2014, 21% of CVS caseworkers with at least one PMC foster child had caseloads of 21-25 children; 13% had caseloads of 26-30 children; 6% had caseloads of 31-35 children; and 3% had caseloads of 36 or more children. (PX 2129). In other words, 43% of DPFs's caseworkers have caseloads above the most lenient professional standard before the Court; and the situation is likely much worse.

Beyond its nebulous stage-counting approach, DFPS “counts [as caseworkers] people that are not there,” such as workers on maternity or medical leave. (D.E. 305 at 41). Moreover, secondary I See You workers and Centralized Placement Unit workers “are categorized in [the DFPS] computer system as conservatorship caseworkers” even though neither workforce has close to the responsibility of a primary caseworker. (D.E. 310 at 67; see also infra pp. 171-76). DFPS also counts caseworkers that only keep track of one child, part-time caseworkers, and fictive workers who are “created out of all the overtime,” which “are not actually even people.” Id. DFPS's data already shows excessive caseloads. The Court can only imagine the actual figures if make-believe people are no longer counted.

Despite DFPS's deception, the caseworkers themselves say that they are overworked. Over 55% of DFPS's caseworkers agree that “they do not have adequate time during the workday to successfully do their job.” (DX 119 at 21). More than half believe that DFPS's expectations for their job performance are unreasonable. Id. Caseworkers often fail to file court reports on time, if at all, and they “do not have enough time to do more than cursory monthly home visits,” typically lasting just five minutes. (PX 1988 at 99). Attorney ad litems Ricker and Vasquez testified that CVS caseworkers are often too busy to return their phone calls, or file necessary documents, which delays needed services to foster children. (D.E. 326 at 224-25; D.E. 327 at 189-92). As one DFPS audit put it, “caseworkers do well just to put out fires.” Id.

Former CVS caseworker B. Miller testified that she had caseloads of “forty to 60 [stages], sometimes higher.” (D.E. 323 at 34). She said this caseload was typical and unmanageable. Id. B. Miller worked approximately 50 hours in a typical week—weekly overtime was the rule, not the exception. Id. at 34-35. Her caseload caused her to experience “[e]xtreme stress, burnout, wearing down, [and] anxiety,” and affected her relationship with her family. Id. at 38. Any time a worker in B. Miller's unit left, the remaining workers' caseloads would “spike” as the leaving worker's cases were redistributed. Id. at 39. As a caseworker, Miller spent most of her time “doing triage” and was forced to prioritize certain responsibilities over others because of the size of her caseload. Id. at 35-36. B. Miller also testified that children in PMC were more affected by her high caseload than children in TMC. Id. at 37. This was because TMC children's cases have “many [more] moving parts” and “many aspects that are demanding for services,” while PMC children are generally already in a placement and at least appear to be relatively settled. Id. at 37. As a result, Miller testified that “when something else [was] blowing up,” PMC children were “the first ones to ... get pushed to the side.” Id.

To gain a perspective on what is a manageable caseload, the Court calculated how many hours it took to simply review the 20 foster children's case files that are in the record, including Named Plaintiffs'. (DX 120). Bearing in mind that a 20-child caseload is below what many DFPS primary caseworkers carry, it took the Court 462 hours just to read the 358,102 pages of case files in DX 120. Admittedly, the Court is not (or was not) well trained at reading case files. That said, any added time it took the Court to read the case files was offset by their incomplete nature: the records contained no CPA investigation files, and there were no IMPACT or RCCL files for multiple children. In short, reading case files, a necessary step for caseworkers to understand the needs of their foster children, takes around eleven uninterrupted workweeks. Considering the frequent turnover among caseworkers, see infra pp. 176-82, this onerous task is duplicated many times for each child. It is no surprise, then, that caseworkers “just don't have enough hours in their day” to focus on their children. (D.E. 326 at 200). As a result, the needs and safety of Texas's PMC children are being overlooked. (See, e.g ., PX 1988 at 10).

For example, Jackson, who aged out of foster care in July 2014, told the Court that he saw his caseworker “once every two months,” and that “it was just hard to get in contact with her [and] to even have a conversation with her ... [to] tell her I needed things.” (D.E. 324 at 186). No caseworker even visited J.S. the month before an alleged sexual abuse in his placement. After the abuse, J.S.'s primary caseworker did not visit him for another four months. S.A. similarly would go several months without contact from her primary caseworker. S.A.'s rotation of overburdened caseworkers caused significant paperwork delays, which prevented many possible adoptions. Attorney ad litem Ricker explained her elaborate protocol just to speak with a primary caseworker: “you call the caseworker and then I call the supervisor, and I call the supervisor's supervisor, and I say, I'm looking for your caseworker, pleading .... And at some point ... I instruct frankly the lowest level person on my staff every hour on the hour I want you to send a fax and make a phone call until we hear back from this person.” (D.E. 326 at 199-200). “It is a pervasive problem,” Ricker testified. Id. at 200. “They just flat-out don't have the time to get back to me.” Id. These examples indicate the obvious: overburdened caseworkers cannot protect their foster children. The purpose of professional caseload standards is to set the point at which children receive adequate care. DFPS's caseloads far exceed all standards. As such, PMC children are placed at an unreasonable risk of harm.

Further, DFPS's administrative demands, such as filling out paperwork, are particularly heavy. CVS caseworkers are only able to spend 26% of their time with children, which Specia called “disturbing.” (PX 1993 at 10; D.E. 300 at 24; see also D.E. 302 at 43-44). The Stephen Group cited this 26% figure as “clear evidence that the agency is doing more compliance than care,” and “a barrier that is qualitatively reducing the CPS caseworkers' ability to keep children safe.” (PX 1993 at 10, 37). The Stephen Group went on to say, “The consequence of these administrative burdens go beyond reducing the time that workers get to build relationships with families. They also mean longer waits to close cases, which leave families in limbo, and worker burnout that leads to high turnover and lower morale which impacts performance.” Id. at 11. In Dr. Miller's experience, CVS caseworkers should spend “at least 40 to 50 percent” of their time working directly with children and families. (D.E. 302 at 42). Dr. Miller further explained that, when you combine DFPS's excessive caseloads with its administrative burdens, “In essence, [foster] children may well not have a caseworker.” Id. at 43.

Former CVS caseworkers B. Miller and Voelkel testified that they were often too busy to keep up with their documentation responsibilities, even though they considered them vital. (D.E. 323 at 36; D.E. 324 at 16). DFPS is well aware that its caseworkers often cannot keep up with required documentation when their caseloads are high. For example, in a memorandum to the Commissioner, McCall noted that exceptionally high CVS caseloads in Region 9 in October 2012 prevented caseworkers from completing required documentation “timely, if at all.” (PX 1825 at 2). Lack of documentation was especially problematic when caseworkers left and their cases were redistributed. (D.E. 324 at 16). The remaining caseworkers could not immediately assess the needs, and appropriately monitor the safety of, the new children on their caseload if their files did not contain thorough and up-to-date documentation. See id. B. Miller at times received files from departing caseworkers that were not up-to-date, making it more difficult to ascertain the current condition and immediate needs of the children. (D.E. 323 at 41). The casefiles that B. Miller received also sometimes contained inaccurate information (such as documentation of required face-to-face visits that had never actually occurred). Id. at 42. Voelkel believed that this common practice of prioritizing more pressing tasks, and allowing documentation to suffer, exposed children to a risk of harm. (D.E. 324 at 16).

The problems of inadequate and incomplete caseworker documentation are considerably magnified by the way in which DFPS maintains foster children's case files. Children's records are not kept in a single location nor are they consistently maintained in chronological order. (See D.E. 343 at 1; see generally DX 120 (filed under seal)). Some of the children's files are kept electronically on DFPS's IMPACT casework system. Other children's files are maintained entirely in External paper files. (D.E. 343 at 1-2; supra p. 80-81 nn.25 & 26). Additionally, records relating to abuse and neglect investigations of children in foster care are kept separately by RCCL in the CLASS database. (See supra p. 141 n.43). Although CVS caseworkers have access to the CLASS database, CLASS files are not merged with IMPACT files and it is unclear whether CVS caseworkers are trained, let alone have the time, to check whether children newly transferred to their caseloads have CLASS files. (See D.E. 343 at 2). Thus, not only are foster children's case files shockingly long (358,102 pages of case files for the 20 children for whom the Court has records), they are incredibly disorganized.

Further, inherent problems with DFPS's outdated IMPACT system further impede caseworkers' ability to review important electronic case file information. According to The Stephen Group, IMPACT “is not in sync with current versions of forms that are used [by caseworkers] and forces arbitrary workarounds and repetitive entry of data.” (PX 1993 at 15). This results in “delays and considerable frustration among caseworkers and can mean that those accessing the system might not have immediate availability to the most recent updates in a particular case.” Id. This creates opportunities for important safety-related tasks to “fall through the cracks,” especially when cases are transferred between workers. See id. It is unclear how easily CPS caseworkers can access their foster children's RCCL files, and how often they do so when receiving new files. What is clear is that caseworkers' continuously fail to maintain complete, timely, and accurate documentation. The resulting widespread neglect of important tasks relating to the safety and well-being of PMC children is indicative of a system where caseworkers' workloads are unmanageable.

In addition to excessive caseloads placing PMC children at an unreasonable risk of harm, CPs's work environment encourages its primary caseworkers to prioritize TMC children over PMC children. Many reports found that CPS caseworkers work in an environment permeated with a fear, making “people feel that meeting the numbers is as important as helping children and families.” (PX 1993 at 248). In 2013, the State Auditor's Office found that the work environment at CPS is “unsupportive and punitive” and that CPS staff do not “feel safe to raise concerns or make complaints, fearing retaliation or punishment.” (See DX 119 at 30). The 2014 Sunset Commission echoed these findings. Id. Employees interviewed by the Sunset Commission generally described management practices at CPS as “unfair, unsupportive, bullying, unreasonable, and fear-driven.” Id. at 29. Many CPS caseworkers and managers even stated that they were concerned about possible retaliation for their cooperation with the Sunset review. Id. Because caseworkers at CPS have no way to make “complaints within the agency but outside their chain of command,” they are generally discouraged from making any workplace or program-related complaints at all. Id. at 33. According to the Sunset Commission, because of CPs's current management structure, there is “persistent fear of retaliation among caseworkers and supervisors.” Id.

The Stephen Group made similar findings:

Supervisors are driven to use metrics as a mechanism to make sure caseworkers ‘meet their numbers,’ not as a tool to balance workload, identify workers in need of training or recognize structural issues that should be resolved. This creates a culture of fear where caseworkers are focused more on meeting numbers and checking off boxes than on the quality

of service for the children in families under their care.

Not surprisingly, many field workers do not feel supported by their supervisors, instead feeling that there is an effort to use metrics to discipline or fire them. This adds to the stress in the workplace and ultimately reduces productivity and increases turnover.

(PX 1993 at 19). One CPS employee told The Stephen Group, “Our workers are threatened and scolded more than trained and molded.” Id. at 248. In such an environment, it is only natural that overextended caseworkers prioritize TMC children who have more deadlines and concretely tracked benchmarks.

Moreover, when a child enters PMC, courts often dismiss the child's attorney ad litem and CASA, leaving the child with fewer stable relationships and advocates. (PX 1988 at 15; see also supra pp. 7-8). This makes the child's relationship with his or her caseworker that much more important. According to the 2010 Texas Appleseed report:

So many resources are spent on getting the child to safety during the TMC phase that once the child has the last of his or her TMC permanency hearings, the stakeholders feel relieved of some of the pressure to get the child out of foster care. All too often, the move to PMC is viewed as a signal that the child is out of danger, damage control has been successful, the State has become the child's parent, and the child is “in a good situation.” ... [F]or too many children, PMC becomes a “permanent” placeholder ....

These children essentially continue on with their day-to-day lives, receiving an increased level of attention right before the six-month placement review hearing and then back to the status quo. The judges report that these children typically bide their time until they can age out and go back home to their biological families, often harboring unrealistic but persistent hopes that “all will be fine.”

(PX 1988 at 72). One judge interviewed for the Texas Appleseed study stated that foster children in PMC are “just like ... inmate[s] serving time until they've reached 18.” Id. Another judge remarked that “[t]hese are the children that even God has forgotten.” Id. at 14.

DFPS claims that its secondary Centralized Placement Unit (“CPU”) and I See You workers fill in the gaps left by overworked CVS caseworkers and ensure that all foster children receive appropriate care. The Court disagrees. CPU workers do not even interact with foster children. They only help find placements. (PX 50 at 16). I See You workers only visit foster children once a month when primary caseworkers are too busy or too far away. (PX 52 at 82). McCall acknowledged that I See You workers' main responsibility, as indicated in the position title, is “seeing the child” in her placement every month to confirm that she “is still there.” (D.E. 305 at 61-63). Although I See You workers are supposed to have “meaningful discussions” with the children they visit, they often just show up at the child's placement, ask “five questions and leave.” (D.E. 305 at 65; D.E. 324 at 186). For example, A.M.'s I See You worker, who visited her regularly for more than two years, asked the same 10-12 questions at every visit, and A.M. answered the same way. (See DX 120 at DFPS # 828-997 (filed under seal)). According to the I See You worker's notes, A.M. was almost always “fine” and “stable in her current placement” and the worker often noted that “things are going well for [A.M.] at this time.” Id. In reality, during much of this time, A.M's level of care was Intense and she was repeatedly restrained by RTC staff. Secondary workers are also “not required to follow up on any needs identified during the visit beyond communicating those needs to the primary caseworker.” (PX 2037 at 57). Nor are they responsible for children's case planning or permanency planning. (D.E. 322 at 127). A foster child's relationship with their secondary worker “is never quality.” (D.E. 326 at 88). Their existence—intended as a stopgap—is itself evidence of an understaffed CVS caseworker workforce, as well as DFPS's inadequate placement array. See infra p. 222-23.

Moreover, replacing contact from primary caseworkers with contact from secondary workers inhibits primary caseworkers' ability to form relationships with their foster children. As Burstain explained, a CVS caseworker is often a foster child's “only continuous and stable relationship.” (PX 1871 at 1). Given that PMC children have been removed from their home and likely shuttled between placements, CVS caseworkers are one of the few people that foster children look to for support and guidance. (D.E. 326 at 85). Trust is “highly important” between a foster child and their primary caseworker because children need to feel comfortable telling them their problems. Id. As McCall testified, “especially with emotional harm[,] you need to know ... the child before you can tell something is wrong.” (D.E. 305 at 62). Using secondary workers to shoulder excessive workloads thus hinders primary caseworkers' ability to protect their children. Further, foster children often do not share their problems with secondary workers who they view as “not the real thing,” akin to substitute teachers. (D.E. 326 at 88). Dr. Miller explained that workers who have no permanent relationship with a child, such as I See You workers, cannot “win a child's trust in a sufficient way to have the child actually reveal what is happening in the child's life, particularly if the child is being subjected to maltreatment.” (PX 2037 at 58). Carter testified that foster children “very rarely” see I See You workers “as somebody that is there to support them” because children “intuitively know that this person is just fulfilling a service or a requirement by looking in on them.” Id.

Further, when I See You workers visit children in their placements, they often do not see individual children in private. S.A., for example, complained at one of her placement review hearings that her I See You worker “doesn't talk to her privately.” (See DX 120 at DFPS # 49459 (filed under seal)). In response she was “told to tell the worker when she does want to speak privately.” Id. Sharp testified that when I See You workers visited him while he lived in an RTC outside of his home region, the workers would see him and all other children from his region at the same time. (D.E. 325 at 167). Sharp testified that each visit lasted five to ten minutes, that none of the visits included one-on-one time between Sharp and the worker, and that none of the workers ever asked Sharp any individual questions to properly ascertain how he was doing or whether he felt safe. Id.

Moreover, the record contains numerous examples of post-visit notes by I See You workers that are superficial, confusing, internally inconsistent, or plain useless. For example, when one I See You worker visited A.M. in July 2007, that worker indicated in IMPACT that the method of contact was by telephone, yet in the contact narrative (which was barely two lines) wrote: “I met with foster parent and child all of his needs have been met.” Id. at DFPS # 1046 (emphasis added). Another time, after a July 2008 visit, a different I See You worker wrote that A.M. “was crying and was very hard to understand” during the visit and that she was “crying so much that she refuse[d] to talk. ....” Id. at DFPS # 976. Yet in the very next line the worker concluded: “No major problems were reported. There are no concerns at this time.” Id. At least one of K.E.'s I See You workers appears to have on several occasions pre-filled the identical questions that would be asked during each monthly visit into IMPACT and then typed in K.E.'s responses: “1. Do you feel safe in this foster home? [K.E.'s] response was yes . 2. Are you pushed or hit by anyone at the foster home? [K.E.'s] response was no . 3. Are you in agreement with your continued placement at the foster home? [K.E.'s] response was yes ....” Id. at 1 RFP CPS 141727-29.

I See You workers have significantly more children on their caseload than primary caseworkers, which likely contributes to their cursory and rote interactions during visits. At the time of trial, Texas had 72 I See You workers. (D.E. 302 at 73). According to Black, 39% of PMC children are placed out of region. (D.E. 300 at 101). Each I See You worker thus has between 60 and 70 PMC children that they must visit each month, in addition to any TMC children on their caseload. The often cursory nature of these visits may also be attributable to the general work environment at CPS, which is focused more on compliance (i.e., “checking the box” for required actions, such as monthly face-to-face visits) rather than care. (See PX 1993 at 10-11, 14; DX 156 at 28).

As Burry explained, visits conducted by I See You workers were not the kind of well-planned and purposeful visits that foster children require in order to achieve permanence. (See D.E. 325 at 40-41). Further, I See You workers' inconsistent notes based on their cursory visits, which are included in children's case files, contributed to problematic fragmentation in children's casework. Id. DFPS nonetheless expects primary caseworkers to create permanency plans and goals for their foster children placed outside of their region in reliance on secondary workers' reports, “without ever seeing the child.” (D.E. 325 at 40-41). Voelkel testified that when she was a primary CVS caseworker, she had to rely on I See You workers to conduct face-to-face visits with her out-of-region children for whom she was nonetheless entirely responsible, which caused her “frustration and stress and confusion” because she had to depend on “a middle man” whom she had never met and whom she found difficult to trust entirely. (D.E. 324 at 25). She found it particularly stressful when she had to conduct case planning or testify in court based entirely on second-hand information that she could not personally verify from a person she did not know. Id. at 24-25. B. Miller testified that she could not always trust information from I See You workers, as there were several occasions when she discovered that an I See You worker had not spent as much time with a child as the worker indicated or had not met with that child privately and thus failed to discover problems that the child may be experiencing. Id. at 55. Miller would only learn about these inadequate visits well after the fact, such as when moving a child to a different placement once the prior placement inevitably failed. Id. at 56. The Named Plaintiffs' case files contain several instances of purported face-to-face visits between secondary caseworkers and children that other parts of the record contradict or fail to support. (See, e.g. , DX 120 at DFPS # 988-90 (filed under seal)) (telephone call between A.M. and her primary caseworker where A.M. explains that no I See You worker has been coming to see her). The case files also contain multiple examples of disconnect and miscommunications between primary caseworkers and I See You workers, where one would fail to provide required paperwork or information to the other, and vice versa. See e.g. , id. at DFPS009043853 (e-mail from I See You worker to A.M.'s primary caseworker stating, “I am scheduled to do the placement first thing tomorrow morning and still have not received placement paperwork”).

Dr. Miller “cannot imagine how a primary caseworker can provide quality permanency casework when she is not visiting the children for whom she is responsible.” (PX 2037 at 58). Nor can the Court. Secondary caseworkers can surely make primary caseworkers' jobs easier in certain respects, and can contribute to positive outcomes for children (although there is little concrete evidence of that before the Court). The record demonstrates, however, that secondary caseworkers do not compensate for DFPS's grossly overburdened primary caseworkers. The Court finds that secondary workers are insufficient substitutes for primary caseworkers, and do not prevent Texas's PMC children from being placed at an unreasonable risk of harm.

Even if the Court accepted secondary workers as adequate substitutes, the Court still finds that caseloads are excessive to the point that foster children are placed at an unreasonable risk of harm. DFPS already includes CPU workers and I See You workers in its caseload data (alongside caseworkers on maternity leave, caseworkers in training, and caseworkers created out of overtime hours), and those numbers reveal excessive caseloads.

Besides harming foster children in and of itself, DFPS admits, “High caseloads lead to high worker turnover, further exacerbating high caseloads.” (PX 877 at 8; see also D.E. 300 at 34, 42; D.E. 305 at 36-37, 56; PX 2037 at 12). Dr. Miller calls this the “cycle of crisis.” (D.E. 303 at 22-23). Specia admits that the “appropriate workload spread out among the workers ... will help me keep workers.” (D.E. 299 at 82). It is no surprise then that “DFPS has an extraordinary amount of turnover.” (D.E. 305 at 55).

The Stephen Group reported that yearly CVS caseworker turnover is 26.7%, and “a major organizational burden.” (PX 1993 at 16-17, 76). To compare, turnover for workers comparable to Texas's CVS caseworkers was 14% to 15% in Kentucky and 10% to 12% in Tennessee. (D.E. 303 at 28-29). The Stephen Group also noted that turnover is especially high for new CPS workers, with approximately 28% leaving within the first year, and approximately 43% within the first two years. (PX 1993 at 17-18). Likewise, Black testified that turnover is approximately 38% for first-year caseworkers. (D.E. 300 at 38-39). The Sunset Commission reported, “One out of every six new caseworkers leaves CPS within six months.” (DX 119 at 20).

Unmanageable caseloads are the main reason that CVS caseworkers leave. In a survey, 70% of the caseworkers that left listed “Workload” as the first or second reason. (PX 1993 at 306). In a 2009 article, Burstain wrote, “With respect to CVS, historically, a fairly direct relationship exists between caseloads and voluntary turnover.” (PX 1871 at 11). In support of that statement, Burstain cited data showing that when “caseloads declined 16 percent from 2006 to 2008 ... CVS voluntary turnover declined 10 percent.” Id. Another report found that heavy caseloads “contribute[ ] to high turnover rates.” (PX 1964 at 9). This finding has “remained consistent from year to year.” (PX 844 at 5).

This high turnover rate means that one out of every 11 CVS caseworker positions is vacant. (DX 119 at 19). Even when those vacancies are filled, it takes “two years” for a caseworker “to fully be up to speed.” (PX 1995 at 159). During their first three months, caseworkers are in training and do not have any cases. (DX 119 at 19). Consequently, while CPS has 1980 primary caseworkers, it needs to hire more than 500 primary caseworkers per year to retain an experienced workforce of only about 1000 who actually close most of the cases. (See D.E. 305 at 41; PX 1993 at 16-17).⁴⁶ This puts a tremendous strain on the 1000 veteran CVS caseworks, who are the front line workers for over 29,000 foster children, which is not captured in DFPS's figures.

46 The Court based these figures on McCall's testimony that there are 1980 CVS caseworkers, (D.E. 305 at 41), and The Stephen Group's finding that yearly CVS caseworker turnover is 26.7%. (PX 1993 at 76). This turnover rate means that out of the 1980 CVS caseworkers, 528 leave each year. Obviously, DFPS must hire 528 caseworkers each year to maintain staffing levels. Thus, 1056 CVS caseworkers are within their first two years and not fully up to speed, meaning that 924 CVS caseworkers handle most of the cases. (See also PX 1995 at 159; D.E. 303 at 33-34) (describing these calculations for all CPS caseworkers).

Caseworker turnover has many negative impacts beyond higher caseloads. Black admitted that turnover causes delayed investigations, a lack of continuity in providing services to families and children, a lack of consistent timely visits by caseworkers to children in State custody, and significant costs to the State in terms of recruiting, training, and lost productivity. (D.E. 300 at 34, 42; see also DX 119 at 17; PX 1995 at 159). Caseworker turnover also “delays or disrupts services and the case plan” of foster children, (PX 1871 at 2), and hinders permanency planning. (D.E. 312 at 20). Moreover, as the Stephen Group explained, “workplace turnover is endemic and institutional knowledge is stripped from across the agency.” (PX 1993 at 17). The high level of turnover at CPS “represents an extraordinary organizational challenge to replace these workers and maintain a consistent level of performance.” (PX 1993 at 16). As one audit of DFPS explained, “Numerous transitions in caseworker assignments disrupt momentum toward permanency by forcing children/youth and their families to ‘start over’ repeatedly with new caseworkers.” (PX 1880 at 5). Specia admitted that foster children are “[a]bsolutely” harmed when they do not achieve permanence. (D.E. 299 at 39).

Dr. Miller explained that foster children “have already been damaged” before entering the State's care. (D.E. 302 at 32). According to Carter, subjecting these “already fragile” children to a rotation of overburdened caseworkers only causes “despair,” “isolation,” and “helplessness.” (Id ; D.E. 326 at 91; PX 2015 at 4 (filed under seal)). Instead of becoming a stable influence in a child's life, foster children “don't want to have a relationship with [caseworkers] ... they lose confidence, they lose trust,” and see caseworkers as just a “number.” (D.E. 324 at 20-21). The 2010 Texas Appleseed report found that the high incidence of repeated turnover in PMC children's caseworkers “contributes to the child[ren]'s difficulties in establishing trust” with their caseworkers. (PX 1988 at 67). Additionally, Texas Appleseed found that when a child's caseworker “changes constantly or fails to make regular contact with the child, the child's feelings of rejection are compounded and lead to feelings of distrust between the child and the ‘system,’ ” which “makes it all the more difficult to elicit critical information needed to ensure appropriate care.” Id. at 100. “[A] child exiting foster care in 2008 had an average of 3.87 caseworkers.” Id. at 99. The longer a child spends in foster care and PMC, the more caseworkers the child will have. Id. Children who exited care in 2008 and spent at least three years in the State's PMC had an average of 6.39 caseworkers. Id. at 100. Texas Appleseed found that “ ‘[r]evolving caseworkers' make it easy for ... essential services and support for the child to ‘fall through the cracks.’ ” Id. at 68. The report also found that:

[S]takeholders in the foster care system in Texas are united in their frustration with the high rate of caseworker turnover. New caseworkers spend about six months in training before they take on a full caseload, and considering that CVS caseworkers stay in that position for such a short amount of time, it is not surprising that they often have little or no personal knowledge of the children in their care. Judges especially saw the inexperience of the CVS workers and the frequent changes in Department representation as a “big problem” or even the “biggest” CPS problem, “one that makes everyone else's job harder.” Several commented that CVS workers did not have the training, life experience, or job experience to effectively manage their cases.

Id. at 101.

Carpenter testified that significant caseworker turnover was the norm for the more than 180 former foster youths with whom she has worked. (See D.E. 307 at 16). Solis, S.A.'s attorney ad litem , testified that frequent caseworker turnover is typical for the foster children he has represented. (D.E. 323 at 18). Virgil, who had “[a]bout 10” caseworkers during her seven years in Texas foster care, often “didn't know who to go to” and “didn't know who to trust.” Id. at 200. As a result, “most of the time” when she “had issues,” she kept her “mouth shut.” Id. Former foster child Jackson testified that he had between 20 and 30 primary and secondary caseworkers over his six years in Texas foster care. (D.E. 324 at 183, 185).

Likewise, former foster child Bentley did not have relationships with her caseworkers, and often did not even known who they were. (D.E. 324 at 63). Bentley, who entered foster care at the age of two and eventually aged out of care, was sexually abused in every single foster care placement but one and was constantly physically abused as well. Id. at 63, 66. Although Bentley was relieved when a caseworker would visit her placement, such visits were rare and did not occur monthly as required. Id. at 63-64. When a caseworker did visit Bentley, it would usually be in the presence of her foster parent and her abuser (often a biological or adopted child of the foster parent), which made it difficult for her to report the sexual abuse. Id. at 64. Bentley would try to hint about the abuse because she could not talk about it openly, but the caseworker would fail to notice or understand and the abuse would continue after the caseworker left. Id. Visits from Bentley's caseworkers were often brief, “15 minutes at the most,” and her caseworkers “never had time” to connect with her or get to know her. Id. at 68. To Bentley's knowledge, none of the physical or sexual abuse that she was subjected to was ever properly investigated, even when she was able to openly report it (which was only possible after she had been moved to a new placement). Id. at 65-67. She did not feel that there were any caseworkers that she could confide in for her safety or that any of her caseworkers could protect her. Id. at 64. Moreover, Bentley testified that during her time in care, she also observed other foster children suffering “a lot” of abuse. Id. Bentley testified that, per her observations, her experience was typical of children in Texas foster care. (D.E. 324 at 202-03).

Sharp, who entered care at ten years old and eventually aged out, had just two primary caseworkers during his time in care, yet he only saw them approximately “10 or 12 times in person” over eight years. (D.E. 325 at 166). Sharp did not develop a relationship with either of them, as he found it “near [ly] impossible to form a relationship with somebody that” he had only seen “two handfuls of times.” Id. He also only saw his I See You worker two or three times during the year and a half that he was at an RTC far away from his home in Region 1. Id. at 167. When I See You workers visited Sharp during that period, it was a different worker each time. Id. The workers also saw all of the Region 1 children in that RTC simultaneously as a group, with each visit lasting five to ten minutes. Id. None of those visits included one-on-one time between Sharp and the worker and none of the workers ever asked Sharp any individual questions to properly ascertain whether he felt safe. Id. Sharp was sexually abused by one of the caregivers at that RTC, as well as by another foster child. Id. at 168-69. Although he wanted to report the abuse, Sharp was unable to because the abusive caregiver was generally present when Sharp was on the phone with his caseworker. Id. at 169. Sharp also felt that no one would believe him or do anything about the abuse even if he had managed to report it. Id. Sharp was under that impression because of his lack of trust or relationship with any of his caseworkers. Id. at 170. Sharp testified that during that time he felt like he did not have anyone who cared about him and that he would have been better off dead because he “knew that [the abuse was] just going to keep happening.” Id. at 172. Sharp testified, “I don't think you can ever go back and undo the trauma of being raped by your caregiver, of being beaten by your peers and feeling like you couldn't talk about it and you didn't have any hope.” Id. at 182.

Sharp, who after aging out of foster care began to work with current and former foster youths, stated that all the young people that he has worked with who had been placed at an RTC had been abused while in care. Id. Those youths generally had experiences similar to Sharp's: they attempted to report abuse or other safety problems in their placements but received no response or follow-up, which then discouraged them from trying to make any future reports. Id. at 170-71. Sharp also suffered physical and psychological abuse in other placements, including other RTCs, foster homes, and group homes. Id. at 171. Sharp found sexual assault between foster children to be common, especially in group homes where caregivers were simply “not able to watch everyone.” Id. at 172. Sharp described one foster group home where one young boy was sexually abused by a bigger boy “almost every night.” Id. at 172-73.

Attorney ad litem Ricker testified that “almost all” of the PMC children she has represented have been sexually abused. (D.E. 326 at 216). Carpenter testified that, for the over 180 former foster youth with whom she worked, physical, sexual, and emotional abuse in foster care was “way too prevalent” and “the norm.” (D.E. 307 at 23-24). Approximately 50% of the former foster youths that Carpenter has worked with were sexually abused in foster care. Id. at 32-33. Caseworkers for those youths also often failed to return the youths' phone calls, even when those calls were in regard to serious issues. Id. at 17. Furthermore, the youths' former caseworkers consistently failed to “show up” or “do what they said they're going to do,” and some of the former foster youths did not always know who their caseworkers were. Id.

In addition to leaving PMC children without a caseworker who can effectively protect them, caseworker turnover can on its own cause psychological harm to children. According to Carter, common to the Named Plaintiffs “was a lack of relationships and contact with their CPS caseworkers and a lack of continuity in their care due to experiencing frequent changes in [their] CPS caseworkers.” (PX 2015 at 4 (filed under seal)). Carter has a significant amount of experience in evaluating the importance of a good relationship between foster children and their primary caseworkers. (See D.E. 326 at 85). He stated that it is “[h]ighly important” that children have a strong rapport with their caseworkers and that they feel their caseworkers understand them and “have regular contact with them.” Id. at 86. However, Carter testified that children with whom he worked rarely had a strong, trusting relationship with their caseworkers, either because the caseworkers were too busy to develop such a relationship, or due to the constant turnover of caseworkers. Id. at 86-87. According to Carter, “One of the most basic needs any child has is the sense that they are safe.” Id. at 98. However, “If a child feels that the person who is directing [her] life does not know [her], does not have contact with [her] or is ever-changing, that causes [the] child to feel that her world is not a safe world,” which is “a very discouraging feeling or emotion for a child to have” and can lead to depression. Id. When foster children, who have already suffered “a disruption in their personal relationships,” experience repeated caseworker turnover it compounds their “feelings of isolation and helplessness.” (PX 2015 at 4 (filed under seal)). Carter testified that the conditions he saw in his evaluations of the Named Plaintiffs, such as psychological disturbance and “learned helplessness,” were typical for the foster care children he has worked with. (See D.E. 326 at 99-100, 114, 120-21, 125, 131-33). These conditions are generally worse in PMC children. Id. at 132-33. According to the 2007 Texas Appleseed report, “even fundamentally normal children who have been taken from their homes and families can become aggressive and ‘emotionally reactive’ due to a lost sense of trust, and their conditions are only worsened by ... frequent caseworker turnover.” (PX 1966 at 22).

Further, besides turnover harming the children who lose their caseworker, it also hurts children whose caseworkers remain and take over the departing workers' cases. (D.E. 323 at 39). This situation all but assures that every PMC child is assigned an overburdened caseworker.

Specia also acknowledged that the “longer children stay in the custody of the state the harder it is for them to achieve a permanent home.” (D.E. 299 at 63, 83; see also PX 1988 at 9). Thus, another consequence of rotating overburdened caseworkers, which disrupts permanency planning, is that 1300-1400 foster children age out of the system each year. It is widely recognized that foster youths who age out generally experience poorer life outcomes. These youths leave the system with few life skills and little, if any, support. As Burstain wrote before joining DFPS, aging out of care is an outcome DFPS tries to avoid as aged-out youths “have no permanent place to call home and often have a difficult time.” (PX 1877 at 36). Burstain reaffirmed this sentiment at trial, saying that children for whom DFPS has failed to find a permanent home and who age out are “likely ... to be harmed.” (D.E. 310 at 55). This is especially true for children who age out while living far from their home communities and support networks, often lacking the resources or ability to return. As of August 2014, approximately 60% of all foster children were placed outside their home county. (DX 183 at 4). Thus, aging out in a foreign community is commonplace. As a result, these children frequently end up homeless, participating in “criminal activities in order to survive, trespassing in vacant homes or stealing or human trafficking, prostitution, those kind of things in order to have a place to stay.” (D.E. 307 at 15; see also PX 1872 at 4).

Aged out foster youths often experience “serious, and in some cases disabling, physical and mental health care issues” and are likely to suffer from post-traumatic stress disorder (“PTSD”) due to “the traumas and frequent moves and transitions experienced in foster care.” (PX 1988 at 10). According to Casey Family Programs, former foster youths suffer from PTSD at nearly five times the rate of the general population and nearly twice the rate of United States combat veterans. Id. at 57, 71. In addition, aged out foster youths often have significantly lower educational attainment than their peers. Id. at 53-55, 58-59. Foster youth in general are “much more likely to be held back than their peers.” Id. at 54. This is due in part to the frequent school changes, which means they are “often absent for large parts of the school year, lose academic credits due to mid-semester moves, and often have incomplete school records due to missing transcripts.” Id. at 53-54. Foster children are also significantly overrepresented in special education classes—at a rate of over four times that of the general population—which may be an underestimate. See id. at 55. Many do not finish high school. Id. at 58. Fewer than 2% of former foster children complete college. Id. at 56.

Talley, who previously worked at DFPS as a Preparation for Adult Living coordinator, testified that the 800 former foster youths for whom she provided services, the majority of which were in PMC, were simply being “maintain[ed] in foster care until they aged out.” (D.E. 323 at 84-85). Carpenter testified that aged-out children lack independent living skills. (See D.E. 307 at 8, 13, 29-30). They do not know how to answer a phone, take or leave a message, cook a meal for themselves, or load a dishwasher. Id. They do not know how to fill out a job application, let alone drive a car to get to work. Id. at 29-30. According to Carter, none of the Named Plaintiffs who are on the cusp of aging out, or have by now aged out, “have sufficient adaptive living skills that are necessary for even a minimally reasonable chance at a decent” life. (D.E. 326 at 129). None of them were involved in any extracurricular activities at school or had any vocational training or employment experience. Id. None of them can “take care of themselves.” Id. In general, as Burstain wrote in 2009 before joining DFPS, youths who have aged out of care:

[A]re less likely than their peers in the general population to achieve academic milestones, including high school graduation and postsecondary education, which are the foundations of self-sufficiency. These youth are less likely to be employed and, even when they are employed, are more likely to be in jobs that do not pay a living wage. They are more likely to experience violence, homelessness, mental illness, and other poor health outcomes. They are more likely to be incarcerated, to abuse substances, and to experience early parenthood out-of-wedlock.

(PX 1877 at 36; see also PX 1988 at 9-10, 14). Carpenter testified that 49% of the young women who age out of Texas foster care are pregnant by the time they are 19 years old and 70% of those children end up in foster care. (D.E. 307 at 13-14).

All professional standards, national standards, state standards, and anecdotal evidence this Court has heard—from both sides—are in lockstep that excessive caseloads and overburdened caseworkers cause an unreasonable risk of harm to PMC children. The Court therefore finds that DFPS's policies and practices regarding its CVS caseworkers amount to structural deficiencies that violate the General Class's Fourteenth Amendment right to be free from an unreasonable risk of harm.

1. DFPS is Deliberately Indifferent Toward Caseload Levels

The central inquiry for determining deliberate indifference is whether DFPS was aware of facts from which the inference could be drawn that maintaining excessive caseloads created a substantial risk of serious harm to PMC children.⁴⁷ See Hernandez , 380 F.3d at 882. DFPS has known for almost two decades that overburdened caseworkers cause a substantial risk of serious harm to foster children. (PX 1964 at 2, 13). DFPS also admitted, “An overloaded case worker is bad for the children they are supposed to protect,” (D.E. 311 at 21), and high caseloads “put[ ] a burden on the worker” and “can have a number of negative consequences.” (D.E. 299 at 55-56). Further, DFPS's external consultants have told DFPS that manageable caseloads are crucial to foster children's well-being. Numerous reports echo this sentiment. (See, e.g. , PX 1966 at 11; PX 1988 at 16; PX 1880 at 16).

47 Foster children have the constitutional right to be free from an unreasonable risk of harm. To succeed on their claim, Plaintiffs must prove that the State violated their right with the requisite level of culpability: deliberate indifference or substantial departure from professional judgment. The “substantial risk of serious harm” language from Hernandez refers to the burden, not the right. 380 F.3d at 882 (emphasis added).

In addition, DFPS has long been aware that its caseloads are too high. As early as 1996, the Governor's Committee to Promote Adoption told DFPS that it needed to reduce CVS caseworker caseloads. (PX 1964 at 10). In 2004, the Texas Comptroller issued the Forgotten Children report, finding that DFPS had several serious problems, including “heavy caseloads and high caseworker turnover that prevented the agency from performing required visits with foster children.” (PX 1966 at 5-6). That report prompted the Texas Legislature the following year to begin efforts to comprehensively reform DFPS by passing Senate Bill 6 (“SB 6”), which, inter alia , provided CPS with $250 million in additional state funding. Id. at 6. SB 6 was aimed primarily at the issue of unmanageable investigation (“INV”) caseworker caseloads, leading CPS to hire more than 3200 INV caseworkers, supervisors, and staff, and improve INV caseworker training. Id. at 6-7. SB 6, though, did not address CVS caseworker caseloads, which were also high. See id. The additional INV caseworkers resulted in CPS removing significantly more children from their homes and into State custody (from 13,431 children in 2004 to 17,547 children in 2006—a 30.6% increase). Id. at 7. This, in turn, exacerbated the already high CVS caseworker caseloads. Id. at 7-8. In 2007, Texas Appleseed told DFPS that “the problem of increasing conservatorship caseloads and inadequate conservator caseworker staffing ... are simply getting worse.” Id. at 36.

In 2009, Burstain published an article noting, “when caseloads are high, CVS workers may not be able to do real social work because they must spend all their time completing paperwork and complying with administrative requirements.” (PX 1871 at 11). She further stated, the “Texas legislature acknowledges that reducing caseloads is important but past legislative sessions focused on investigations.” Id. at 14. Burstain explained that “CVS caseloads are too high and above recommended standards” and recommended that caseloads “be reduced to keep paperwork manageable, allowing CVS workers to focus on helping and providing services to children and their families.” Id. at 15.

The 2010 Appleseed Report found that “CPS caseworkers were routinely handling around 30 cases, even though the national standards call for between 15 and 17 cases per worker ... [and] caseworkers are commonly assigned 40-plus cases at a time.” (PX 1988 at 16). Moreover, the report pointed out that caseload “averages do not reflect spikes in caseloads that can occur when a co-worker goes on extended leave or quits.” Id. “Almost every stakeholder in the foster care system interviewed” for the 2010 Appleseed Report, “expressed the belief that CVS workers are overworked to the point of not being able to do their jobs properly.” Id. at 99. In addition, the 2010 Texas Adoption Review Committee said that caseworkers carry “extremely high caseloads” that are “often twice what is deemed best practice.” (PX 1964 at 9). That report went on to say, “Many of our recommendations are, sadly, ones that have been made in prior years,” including, “Reduce caseload to 15-17/worker in compliance with national standard.” Id. at 2, 13.

Also in 2010, DFPS told the Texas Legislature that it the needed to reduce caseloads for its CPS caseworkers. (PX 1286 at 15). DFPS acknowledged that decreasing caseloads would allow caseworkers to make “regular, meaningful contact with children in the substitute care system” and “to ensure the safety and well-being of [those children] until the permanency goal is reached.” Id. In that report, DFPS presented data indicating a correlation between decreasing “statewide average daily caseload[s]” over the prior three years and children “being seen more frequently by their caseworkers.” Id. In its Legislative Appropriations Request for Fiscal Years 2012 and 2013, DFPS acknowledged that the caseload issue was still ongoing, stating, “High caseloads cause quality casework to suffer, thus putting children in our system further at risk of harm.” (PX 878 at 105). DFPS also noted that foster child “safety is greatly improved with lower caseloads that allow workers to spend more time on each case.” Id. at 15.

Yet despite DFPS's knowledge that it needed to reduce caseloads, between 2009 and 2013 CVS caseloads only increased. (PX 885 at 18; PX 894 at 10). In its Legislative Appropriations Request for Fiscals Years 2014 and 2015, DFPS requested additional funding from the Legislature to lower caseloads to 2009 levels. (PX 885 at 18). DFPS acknowledged that allowing caseloads to go any higher would cause “the quality of casework to suffer, ... would impair the agency's ability to keep children safe,” and would result “in significant ... safety issues.” Id. Currently, though, even by DFPS's distorted data, caseloads remain well above the maximum recommended by professional standards. According to DFPS's stage-counting method, in 2014 CVS caseworkers in many counties averaged over 40 stages, with the worst averaging 88.3. (PX 650). Former CVS caseworker B. Miller testified that a “forty to 60” stage caseload is “typical” and unmanageable. (D.E. 323 at 34). Yet, in a memorandum to the Commissioner, McCall described the CPS program Regions 2 and 9 being “in a state of crisis” because average daily caseloads were 27.9 in Region 2 and 37.9 in Region 9. (PX 1837 at 3).

DFPS has also known for a long time that caseworker turnover poses a substantial risk of serious harm to foster children. A 2008 report by the University of Houston Graduate College of Social Work, produced in collaboration with DFPS, identified “Conservatorship Worker[ ] Issues” as an area of concern. (PX 1273 at 18). Under that area of concern, the report listed “continued turnover” as a specific weakness and “high caseloads” as a threat. Id. The Texas Adoption Review Committee, in its December 2010 report, found that high CPS caseloads contribute “to high turnover rates and reduce[ ] positive outcomes for children.” (PX 1964 at 9). The 2010 Texas Appleseed report found that “[c]aseworker turnover [was] high and directly impact[ed] caseload[s].” (PX 1988 at 100). In its Legislative Appropriations Request for Fiscal Years 2012 and 2013, DFPS noted that if certain already-authorized positions were not funded, average daily caseloads would greatly increase, as would turnover, “and the quality of casework would diminish, thus impairing the agency's ability to keep children ... safe.” (PX 878 at 13). An internal study done by DFPS in December 2012 found that two of the main factors contributing to CPS caseworker turnover were “poor working conditions and environment (safety and work-related stress)” and “workload concerns making it difficult to perform adequate work.” (PX 844 at 5, 10). In the Self-Evaluation Report that DFPS submitted to the Sunset Commission in 2013, the agency admitted that “[r]ecruiting and retaining high-quality talent” is “one of the largest challenges for DFPS” and a “long-standing concern.” (PX 1964 at 22). In its Legislative Appropriations Request for Fiscal Years 2014 and 2015, DFPS noted that “[e]ntry level direct delivery positions experience the highest rate of turnover within the agency. Protective services is a stressful job, made even more so when caseloads are high. About 29 percent of new caseworkers in the CPS program ... leave within the first year.” (PX 885 at 19). DFPS requested additional funding at that time to “incentivize staff to remain in their job long enough to become more competent, make better decisions, and provide the staff coverage necessary to adequately protect the children ... served by DFPS.” Id. In its Description of Revised Exceptional Item Requests for the FY 14-15 Biennium, directed to the House Appropriation Committee Article II Subcommittee, DFPS explained, “High caseworker turnover leads to higher caseloads, less experienced caseworkers and supervisors, and impacts quality casework and ultimately safety for the clients.” (PX 894 at 6, 7). DFPS further stated, “Caseworker turnover in all programs threatens the well-being and safety of clients. ” Id. In its Legislative Appropriations Request for Fiscal Years 2016 and 2017, DFPS admitted that “[r]etention and turnover issues have impacted the ability of CPS to achieve full staffing during FY 2014.” (DX 97 at 119). In 2014, The Stephen Group noted that the issue of turnover is “[a] well-known concern across CPS that was reinforced by [The Stephen Group's] operational review.” (PX 1995 at 7). Also in 2014, the Sunset Commission found that the “Legislature and DFPS have long been concerned with reducing chronically high caseworker turnover, which results in a number of problems that directly affect the agency's ability to meet its mission of protecting children.” (DX 119 at 9). Yet, “CPS turnover remains significantly higher than the state agency average.” Id.

Furthermore, as high as caseworker turnover is, the reported turnover rate appears to be understated. In an article from 2009, Burstain stated that the way in which DFPS calculates turnover, by counting “how many full time, regular employees left the agency either voluntarily or involuntarily ... is limited and does not allow for a full understanding of workforce issues and how they should be addressed.” (PX 1871 at 2). The article further states, “From a stability perspective, turnover should include every time workers with direct case management responsibilities leave their position,” which would account for not only the workers leaving DFPS but also “those who transfer to different positions within the agency, either laterally or through a promotion.” Id. The reason for this is fairly straightforward: “After all, the child or family does not care why the worker changed, only that they now must establish a relationship with someone new, which often delays or disrupts services and the case plan.” Id. Burstain's article reflects that, based on DFPS-provided data, from 2006 to 2008, CPs's total workforce turnover rates were understated by between 5% and 11% per year (accounting only for promotions). Id. Additionally, from 2006 to 2008, CPS workforce turnover, specifically for CVS workers, was further understated by between approximately 10% and 14% per year (accounting for promotions as well as for CVS caseworker transfers into other CPS units). Id. at 11. The Court has seen no evidence that DFPS has changed its method for calculating turnover since 2009.

DFPS has done little to address this problem. It has not created a caseload range, it has not hired enough primary caseworkers, and it has not instituted policies to reduce caseworker turnover. Further, 2004 is the last time DFPS conducted a work measurement study to determine what is a manageable caseload. (DX 5 at 1). In 2010, DFPS purported to “determine how many caseworkers it would take to ... make meaningful monthly visits with all children ... in foster care,” producing in July 2010 a 5-page report titled “What Should CPS Caseloads Look Like.” Id. According to the report, DFPS developed “a data driven staffing model” based on the figures from the 2004 study. Id. But the 2010 update does not explain or reevaluate any of the results from the 2004 study and instead makes upward adjustments of the previous figures based only on DFPS policy and practice changes that have been instituted since 2004. See id. at 1-5. The update assumed a very low vacancy rate (4%), which significantly underestimates the actual CPS vacancy rates: statewide CVS vacancy rates were 7.5% in May 2014, and significantly higher in some regions. (DX 71 at 10). The update also assumed that “the additional workers needed are trained and carrying a full caseload each year,” which is simply not consistent with reality: new workers, who make up 21% of all DFPS staff, do not carry a caseload for their first three months, have a reduced caseload for the next three months, and are not fully up to speed for two years. (D.E. 305 at 38; PX 1995 at 159). Using the adjusted 2004 figures, DFPS determined an average daily caseload of 24.9 would theoretically allow CVS caseworkers to “make meaningful monthly visits with all children” on their case load. (See DX 5 at 1, 4).

Yet CVS caseloads have consistently been far higher than 24.9. According to a September 2013 Self-Evaluation Report submitted by DFPS to the Sunset Commission, during the 2012 fiscal year, the daily caseload for CPS substitute care workers was 33.7. (DX 31 at 29). The State's target for average daily caseload for CPS substitute care workers for that year was 29.2, although it is unclear how that target was derived or what it was based on. Id. According to the 2014 Sunset Commission:

DFPS bases its target caseload and the corresponding caseworker performance requirements on a workload time study conducted in 2004. This study no longer reflects current workload, however, since the Legislature has significantly increased requirements by passing major reform legislation in 2005 and 2007, in addition to other bills. DFPS itself has added new policies and practices over time, likely contributing to higher workloads for caseworkers. The State Auditor's Office identified this same issue through an audit published in 2009, which found CPS workload measures were outdated and recommended an updated time study. However, DFPS did not implement this recommendation and continues to use the 2004 information.

Further, the measures themselves focus largely on casework output measures of timeliness that often bear only an indirect relationship to child safety and quality casework. ... With such a heavy focus on quantity, CPS cannot accurately gauge the quality of services provided to children and families. In fact, internal

reviews have illustrated that the focus on timeliness measures may negatively impact the quality of casework.

(DX 119 at 30-31).

Separately from DFPS's failure to conduct a workload study in the past ten years, the Sunset Commission found that DFPS does not distribute workloads equitably or consistently among workers, leading “to inefficiencies that increase travel time and workload, and possibly impact outcomes for children.” Id. at 32. DFPS's failure to implement an effective and consistent method for distributing cases creates “situations like the one in Harris County, where caseworkers routinely drive across the large metropolitan area” to perform their job duties and serve their clients. Id. This is further exacerbated by supervisors “hav[ing] the discretion to distribute cases to their caseworkers however they prefer, which an internal report indicates contributes to perceptions of favoritism and unfair management practices.” Id. The resulting inequitable workload distribution⁴⁸ can lead to caseworkers falling even more behind in their cases, “putting children at risk.” Id. The Sunset Commission noted, “Without a system to assign cases with both efficiency and fairness in mind, CPS will continue to struggle with caseworker retention.” Id.

48 Caseworkers are also sometimes penalized as a result of the uneven caseload distribution, which is entirely out of their hands. (DX 119 at 32).

In its report, the Sunset Commission recommended that DFPS “regularly do casework time studies to more accurately develop caseload goals and policies that are fair and attainable for caseworkers,” whereby DFPS would be able to “ensure that the targeted caseload and caseworker performance goals set by the agency are achievable and reasonable.” Id. at 38. The Commission further recommended that these studies be conducted once every three years. Id. In response to the Commission's recommendation, DFPS stated that it “agrees with this directive and is already planning a time study.”⁴⁹ (DX 103 at 3). DFPS also acknowledged that “[r]outine time studies are an important tool for ensuring that caseworkers are evaluated against realistic job expectations.” Id. Similarly, in its April 2014 assessment, Casey Family Programs recommended that DFPS “[p]erform a detailed[ ] time-study analysis of DFPS caseloads and workloads in order to determine objectively whether caseworkers have manageable workloads, and to identify where additional staffing may be needed in order to reduce and eliminate ... vacancies and assure an adequate workforce for all offices and work units.” (PX 1880 at 6).

49 Despite telling the Sunset Commission that DFPS was “already planning a time study,” no DFPS official could tell the Court of any concrete plans or timelines in regard to performing a CPS caseworker workload time study. (See, e.g. , D.E. 313 at 20).

McCall testified at trial that one of her responsibilities is “to make sure that the primary caseworkers throughout the State have manageable caseloads.” (D.E. 305 at 13-14). Part of her job is also “to analyze whether ... caseworkers have workable or manageable caseloads.” Id. at 17-18. McCall agreed that it is necessary to examine caseloads in order to determine whether there is a problem with them, and how extensive the problem might be. Id. at 20. McCall further testified that DFPS “would shoot to make the worker's workload be as manageable ... as possible.” Id. at 27. Yet McCall was not able to explain at trial what methods she, or anyone else at DFPS, uses to ensure that, or even evaluate whether, caseworkers have manageable caseloads. See id. at 20-28. Instead, McCall testified that DFPS has not done any study recently “to determine how many cases a conservatorship caseworker can carry and [still] complete all” of his or her required tasks. Id. at 28. Indeed, the State has placed no limits on the caseload size that a CVS caseworker can carry, nor “established any caseload size parameters.” Id. Black confirmed that CPS has not done a “full work study” in over ten years. (D.E. 300 at 19). She also acknowledged that, since 2004, CPS has implemented “policies and practices and statutes that have added to our caseworkers' workloads.” Id. at 20.

DFPS also points to “CPS Transformation” as a response to overburdened caseworkers. Launched in October 2014, three and one-half years after this suit was initiated and six weeks before trial, Transformation has three priorities: “developing and maintaining a professional and stable workforce; ensuring child safety, permanency and well-being; and establishing an effective and efficient organization and operations.” (DX 156 at 3). Like many of DFPS's efforts the Court heard about, Transformation is promised to work because it is “completely different” than the reform efforts that came before it. (D.E. 327 at 144). The Court finds that Transformation is not a reasonable response to the problems Plaintiffs allege, and therefore does not protect Defendants from a finding of deliberate indifference. See Farmer , 511 U.S. at 844, 114 S.Ct. 1970. First, Transformation was initiated mere weeks before trial, and DFPS has not shown the Court anything that proves it is working. The Court will not place blind faith in DFPS, who has ignored decades of reports that universally cite the danger of overburdened caseworkers. Second, and more importantly, Transformation does not answer Plaintiffs' claims. Transformation will not produce an updated CVS caseworker work study, or set minimum and maximum caseload standards. (D.E. 313 at 19-20).⁵⁰ Although at trial Specia stated that he planned to conduct a workload study, he did not provide a timeline for any such study and admittedly did not know what steps he would take upon the completion of the study. (See D.E. 299 at 77).

50 Transformation also does not address shortcomings with RCCL investigations or DFPS's placement array. (D.E. 313 at 21; see infra Section IV.B). Nor does Transformation deal with Plaintiffs' safety concerns in foster group homes. Id. ; see infra Section IV.C.

Moreover, DFPS officials, from Specia down, testified that DFPS has not adopted any national caseload standards. (See, e.g., D.E. 299 at 55; D.E. 300 at 12; D.E. 305 at 21). Dr. Miller explained that “the singularly most troubling finding regarding caseloads was the total absence of a 'ceiling,' that point beyond which caseloads would not be considered manageable.” (PX 2037 at 18). Yet DFPS officials testified that DFPS does not have any standards or maximums for its caseworkers' caseloads and that there is currently no plan to implement caseload maximums. (See, e.g., D.E. 300 at 12, 100; D.E. 305 at 21-22, 25-28; D.E. 313 at 20). Black confirmed that CPS does not “have standards [ ] for what the caseloads should be.” (D.E. 300 at 12). Bowman testified that she is not aware of any benchmarks, standards, maximums, or guidelines at DFPS for CVS caseloads. (D.E. 311 at 17).

Instead of responding to high workloads by implementing caseload standards or guidelines, DFPS continues to employ a reactive “squeaky wheel” approach to its caseworkers' workloads. (See, e.g. , DX 119 at 52). McCall testified that DFPS relies heavily on its front line supervisors and program directors to manage their caseworkers' caseloads. (D.E. 322 at 59, 62). She stated that supervisors are expected to go to program directors if there is a caseload problem, and that program directors are then expected, if necessary, to convey the problem further up the chain until it eventually reaches top executives, where the problem will be resolved. Id. at 59-63. Despite multiple reports to the contrary, McCall nonetheless believes that this system “works very well.” Id. at 62. McCall acknowledged, though, that it is not “a prudent or diligent form of management to simply wait for problems to be reported” or “for something bad to happen” before taking action. Id. at 31. Black confirmed that, unless one of her direct reports brings problems to her, she “assume[s] things are going fairly well.” (D.E. 300 at 14-15). However, in CPs's “unsupportive and punitive” work environment, staff do not “feel safe to raise concerns or make complaints, fearing retaliation or punishment.” (DX 119 at 30; see supra pp. 169-70).

The Texas child welfare system is “reactive,” meaning that it waits for problems to develop before responding to them, instead of being “proactive” in heading off predictable and preventable issues. (See D.E. 303 at 48-49, 57-58; DX 119 at 52). According to Dr. Miller, by waiting until problems develop and become significant enough to warrant attention, DFPS's reactive system exposes children to unnecessary risks. (See D.E. 303 at 57-58). Dr. Miller testified that a proactive system is easier to maintain than a reactive system because it “is a lot harder to clean up a mess.” Id. at 58. The goal of a child welfare system should be “minimize the possibility that something like [a child fatality] is going to happen,” which cannot be accomplished without proactive thoughtfulness and planning. Id. The Sunset Commission found that CPs's “inherent[ly] reactive approach” results “in a continuing cycle of crisis and criticism that distracts the agency from developing an effective” management structure, and prevents DFPS from “deliver[ing] desired results.” (DX 119 at 10). The Stephen Group found that although CPS gathers a massive amount of data, for which CPS management “has available 2,700 disparate data reports,” CPS does not have a method “to use these data to consider predictive trends and identify potential ‘hot spots' that require intervention before a bad outcome occurs.” (PX 1993 at 18). As a result, CPs's extensive repository of information “is useful primarily for historical analysis, not proactive decision making.” Id.

In short, CVS caseworkers are crucial to the safety and well-being of all foster children, including PMC children. They are responsible for ensuring that PMC children are reasonably safe while in the State's custody. DFPS has known for decades that its primary CVS caseworkers are overburdened to the point where they cannot perform their required duties, namely protecting their foster children from an unreasonable risk of harm. The Court does not find that DFPS has “responded reasonably” to the substantial risk of serious harm posed by overburdened caseworkers. Farmer , 511 U.S. at 844, 114 S.Ct. 1970. DFPS is therefore deliberately indifferent toward excessive caseworker caseloads.

2. DFPS Substantially Departs from Professional Judgment Toward CVS Caseworkers

Under the professional judgment standard, a decision “if made by a professional, is presumptively valid.” Youngberg , 457 U.S. at 323, 102 S.Ct. 2452. DFPS argues that it has “in place a wide variety of systems to mitigate risks associated with caseworker caseloads and to assure safety of children in the General Class.” (D.E. 359 at 32). However, DFPS fails to point to any specific system. The Court has sifted through the record in search of any decision or response regarding excessive caseloads, but remains empty-handed. Perhaps DFPS is referring to the reports and audits it has commissioned, which were intended to be baselines from which DFPS could begin to improve. Again, DFPS has done little with these reports, which recommend that DFPS “Reduce caseloads to 15-17/worker in compliance with national standard.” (PX 1964 at 13). The Court does not believe that perpetuating a system that places PMC children at a substantial risk of serious harm, and that violates all professional standards, is not a “presumptively valid” response.

The CWLA and the COA recommend that caseloads not exceed 15 children. Dr. Miller set a range between 14 to 17 children in Tennessee and Kentucky, with 20 as the maximum in times of crisis. DFPS caseloads far outpace these standards. Yet DFPS has done nothing to reduce caseloads, knowing all the while that foster children are harmed as a result. DFPS's response (or lack thereof) to its caseworkers' excessive caseloads shows “such a substantial departure from accepted professional judgment ... that the person responsible actually did not base the decision on such a judgment.” Youngberg , 457 U.S. at 323, 102 S.Ct. 2452.

B. Licensed Foster Care Subclass

The Licensed Foster Care (“LFC”) Subclass includes all members of the General Class who are now or will be in a licensed or verified foster care placement, excluding verified kinship placements. This Subclass claims two structural deficiencies. First, the State does not exercise sufficient oversight of its foster care facilities. Second, the State maintains an insufficient number, geographic distribution, and array of placements. Plaintiffs argue that both of these structural deficiencies cause an unreasonable risk of harm to LFC children. The Court agrees.

1. Insufficient Oversight

Plaintiffs' insufficient oversight allegation breaks down into two subparts. First, DFPS's abuse and neglect investigations at its licensed facilities are deficient. Second, DFPS's licensing and inspecting at its licensed facilities are inadequate. Separate and together, Plaintiffs argue that these failings amount to a structural deficiency in DFPS's oversight of its licensed facilities, which cause an unreasonable risk of harm to LFC children.

The Child Care Licensing division of DFPS is in charge of child day care and residential childcare operations. The Residential Child Care Licensing division of CCL is responsible for licensing and inspecting all licensed residential childcare operations, as well as investigating allegations of abuse and neglect at these facilities. Licensed operations include independent foster family homes, independent foster group homes, general residential operations, residential treatment centers, child-placing agencies, CPA foster family homes, and CPA foster group homes. (PX 41, 40 Tex. Admin Code § 745.37 ). Ninety percent of Texas's placements are licensed operations. Private CPAs that contract with the State run almost 90% of the licensed facilities. RCCL is the State's mechanism for ensuring that foster care providers meet minimum standards, keep children safe, and provide adequate care. The task is daunting. In 2013, there were 10,286 24-hour care facilities in Texas that RCCL was responsible for, and there were 4684 inspections and 5160 investigations, 1952 involving cases of abuse and neglect. (DX 119 at 90; DX 249).

When RCCL investigates an abuse and neglect allegation, one of four dispositions follows: (1) Reason to Believe, meaning that “A preponderance of evidence indicates that abuse, neglect, or exploitation occurred”; (2) Ruled Out, meaning that “A preponderance of evidence indicates that abuse, neglect or exploitation did not occur”; (3) Unable to Determine, meaning that “A determination could not be made because of an inability to gather enough facts. The investigator concludes that there is not a preponderance of evidence that [abuse or neglect] occurred; but it is not reasonable to conclude that abuse or neglect did not occur”; and (4) Administrative Closure, meaning, “The operation is not subject to regulation; or the allegations do not meet the definition of abuse, neglect, or exploitation.” (PX 86 at 113). CCL Assistant Commissioner Morris admits that investigators and inspectors “have to make correct decisions” in their investigations, “otherwise, victims could be subject to continued risk and perpetrators could be given more opportunities to cause harm.” (D.E. 301 at 28). RCCL Director Shaw agreed: “If you don't have a good investigative system, you can't have a good [child welfare] system.” (D.E. 304 at 20).

The Performance Management Unit division of CCL ensures that CCL acts according to policies and procedures. It performs internal quality control. At the request of CCL leadership, PMU reviewed 85 physical abuse investigations that occurred between August 1, 2012 to July 31, 2013, which resulted in either Unable to Determine (48) or Reason to Believe (37). Regarding the UTD cases, PMU found that “31 out of 48 (64.6%) of the reviewed [physical abuse cases] were incorrectly determined as UTD.” Of the 48, 36 (75%) involved an injury, with 17 (35.4%) requiring medical attention. (PX 1129 at 3). Nine of the 17 that required medical attention—with injuries ranging from minor to critical—were incorrectly determined as UTD. Eleven of 31 (35.5%) should have been RTB. The common reasons for incorrect dispositions included: “Preponderance explanation was not sufficient[;] Not all parties were interviewed[;] Not all evidence was reached to support/refute findings[;] Determination did not meet the definition[; and] Risks were not thoroughly addressed.” Id. at 4. Supervisors signed off on 30 of the 31 (96.8%) incorrect dispositions. This is a systemic failure.

PMU conducted a second review of all 111 UTD dispositions for cases involving physical abuse, sexual abuse, and negligent supervision investigations over the same period. Director Shaw separately reviewed these cases as well. Their findings were even more egregious. Shaw found that 84 of 111 (75%) UTDs were incorrect. (PX 1065). Likewise, Shaw determined that 33 of 44 (75%) UTDs for physical abuse allegations were incorrect. PMU concluded that 35 of 44 (79.5%) were incorrect. Shaw also found that 31 of 38 (81.6%) UTDs for negligent supervision, and 20 of 29 (69.8%) UTDs for sexual abuse were incorrect. Id. Many of the incorrect findings should have been RTB. A typical investigation error rate for a child welfare system is 2% or 3%. (D.E. 303 at 49-50). RCCL's is 75%. This is staggering, and it means that many abused children—for whom a preponderance of evidence indicated that they were physically abused, sexually abused, or neglected—go untreated and could be left in abusive placements. Even if by some unrelated event a child found reprieve from these facilities, the adult perpetrators can still accept new foster children with nothing in their record indicating a risk. Even with correct dispositions, however, DFPS places LFC children at an unreasonable risk of harm. After the reviews just discussed were completed, “No licenses were suspended .... None were revoked .... No penalties were established on any of these facilities ... [and] the State didn't move any of the children.” (D.E. 301 at 50). In short, little changed as a result of finding out that children were abused and neglected. The Court is convinced that RCCL “simply doesn't work. It's broken.” (D.E. 303 at 51).

In response to discovering this 75% error rate, DFPS conducted a one-day training session for CCL supervisors and program managers to explain to them that CCL has a “very high standard” for investigations, and that their “work has to be better.” (D.E. 301 at 53-56). DFPS took no action nor made any change to RCCL investigation policies or practices. RCCL did hire two additional staff to help Shaw “focus on ... professional development, additional training, ... [and] child safety issues.” Id. at 55. Morris described the above actions as “a pretty robust response.” Id. at 52. Rawls-Martin testified that the DFPS Council was never apprised of the PMU reports or their findings regarding the investigation disposition error rate. (D.E. 308 at 17; D.E. 309 at 8). Rawls-Martin saw the actual PMU reports for the first time at trial. (D.E. 309 at 10). She testified that, as chair of the DFPS Council, she “would have expected to see something this serious much sooner.” Id. Rawls-Martin agreed that the findings reflected a “serious issue” at least in part because “unreliable investigations by the licensing group” potentially result in unreliable statistics and findings in regard to abuse and neglect investigations, potentially placing children at risk. Id. at 8-9. She was concerned that such a high error rate “could call into question ... a much broader scope of investigations.” (D.E. 308 at 23).

The Court is certain that there are scores of other faulty investigations. PMU and Shaw only reviewed the 111 cases with a UTD disposition. In fiscal year 2014, only 4.2% of RCCL's dispositions were UTD. In contrast, 91.8% of dispositions—1984 of 2161—were Ruled Out. DFPS has not sought to determine the error rate for R/O dispositions. Considering that the same investigators who made the UTD findings made the R/O ones as well, and that R/O rates in other states are around “60, 70 percent of the total,” (D.E. 324 at 132), the Court finds it likely that the R/O error rate is similar to the UTD error rate. Morris agrees. (D.E. 301 at 65). So too does Defendant's expert on licensing, P. Wilson, who admitted that the high rate of Ruled Out dispositions “raised questions” for her. (D.E. 320 at 26-29). Shaw testified that she was not concerned with the R/O investigations possibly being inaccurate because she believed that “preponderance is a little more clear” in those cases. (D.E. 304 at 16). Yet Shaw had not reviewed any investigations with Ruled Out dispositions since at least 2013. Further, even if the preponderance finding is clearer in some Ruled Out cases than Unable to Determine cases, that explanation does not account for the fact that the investigators in question were failing to interview all necessary parties, ask pertinent questions, gather all evidence and key information, and address risks. (See PX 1129). As one DFPS employee stated, the “threshold in RCCL” investigations is “no Autopsy, no RTB.” (PX 1706 at 1).

Attorney ad litem Ricker testified about her experience reporting abuse and neglect in a foster care facility in Levelland. The child Ricker visited, J.R., was non-verbal, had fingernails so long that they curled forward to touch his fingers, had a purple “goose-egg” on his forehead that no one could explain, and was wearing a torn shirt and pants that he had to hold up with his hand because there was no button. The place “REEKED horribly of urine” and had “smeared feces on the wall.” (PX 2164 at 2 (filed under seal)). Ricker visited on numerous occasions, each time finding another child with a black-eye. Ricker twice reported to RCCL the abuse and neglect occurring at this facility. RCCL Ruled Out the allegation without even contacting Ricker, in violation of its policy to interview “adults who may have knowledge of the incident.” (D.E. 326 at 212-14; PX 86 at 69). RCCL made its disposition solely on information provided by the children involved, some of whom were nonverbal and intellectually disabled, and the facility's staff. (D.E. 304 at 38-40). The facility was not cited for any violations, and J.R., as well as the other children, remained in its uncorrected care.

Shaw agreed that these two Levelland reports were “pretty serious” and that the conditions at the RTC were not acceptable. Id. at 33, 42. Gilliam, the Regional Director for this facility, testified that those reports concerned her as well. (D.E. 312 at 48). Yet Gilliam did not know about either report until they were disclosed at trial, even though they were serious enough to have been brought to her beforehand. Id. at 49.

Despite these shocking figures and stories, the Court is certain that incidents at licensed facilities are underreported. First, LFC children often do not know to whom they should report abuse and neglect. Sharp testified from his experience as a foster child, it is “challenging to report things in the system .... I didn't know [if] there was a number that I could call. Even if there was, I wouldn't have access to it. I didn't have anyone who I felt comfortable reporting these things to. And I think these are the feelings of most children in the system.” (D.E. 325 at 175-76). According to Sharp, “there is abuse happening all the time that simply is not being reported.” Id. Even if children knew how to report incidents, Sharp explained, many would not because they feel that foster care staff do not believe them. Sharp, who now works with foster children who age out of the system, testified, “I've never met a young person who stayed in a [residential treatment center] that hasn't been abused, and hasn't felt like they wouldn't be able to report it and have somebody believe them.” Id. at 170. In fact, according to many foster children, reporting abuse makes the situation worse. M.D.'s next friend and attorney ad litem , Stukenberg, explained that when children complain about abuse, “sometimes the children are retaliated back against.” (D.E. 324 at 225). Stukenberg went on to say that in some facilities, “[t]here's a mentality [that] if you outcry you're going to pay for it.” Id. Sharp agreed, explaining, “We didn't feel safe in placements and then nothing happened, and so, I mean – why, why would you go through ... the process of even thinking that something would happen if you were to report something like this?” (D.E. 325 at 170-71). Because RCCL often does not act even when they confirm abuse, children remain in these placements, subject to retaliation. Sharp also explained that children are unable to report abuse because caseworker meetings often occur in the presence of their abusive caregiver. Id. at 167, 169. Moreover, it appears that RCCL outsources investigation responsibilities to CPAs, meaning that a child's outcry could be investigated by the alleged abuser. Thus, the already staggering number of abuse and neglect incidents (around 2000 each year), for which there is a 75% investigation error rate, is likely much higher because foster children either do not know who to contact, do not feel that anything will be done, or fear retaliation.

One of the more shocking revelations in this case—which is saying quite a bit—is that RCCL does not track child-on-child abuse in LFC placements. Rather, RCCL investigates child-on-child incidents for negligent supervision by the caregiver. (PX 1026 at 5). Consequently, there is no centralized data that shows which foster children have a history of physical or sexual abuse. (D.E. 300 at 47-48). The only place that RCCL keeps this information is in each child's records, which are often tens of thousands of pages long, unorganized, inconsistent, and contradictory. (D.E. 328 at 154-55). This oversight is unprecedented. Dr. Miller explained that tracking child-on-child abuse is “a critically important piece of the decision making process regarding not only subsequent placements but treatment needs.” (D.E. 303 at 8). Dr. Miller “can't imagine not tracking child-on-child abuse ... particularly with sexually abusing kids, that information has to inform all subsequent placement decision making.” Id. at 6-7. Specia admitted that child-on-child abuse harms foster children and that DFPS should “find an appropriate way to keep track of it.” (D.E. 299 at 36). Yet DFPS does not track this “critically important” information. The obvious result is that foster children are harmed.

The 16-year-old boy that sexually abused D.I. had a sexual abuse history himself. His father had sexually abused him, which led him to foster care. Once in foster care, in 2007 he sexually abused a younger child with special needs. This information was not in any centralized place. The 16-year-old's case files noted the parental abuse, but not the 2007 incident. The consequence was that D.I. was placed with this boy, who forced D.I. into anal and oral sex. The abuse of D.I. was “similar to the previous incident” in 2007. As with the 16-year-old, sexual abuse often “sexualizes” a child, causing them to initiate such incidents in the future. This occurred to D.I. As if to reinforce the lunacy of not tracking child-on-child abuse, D.I.'s next foster home had another “highly sexualized” child with a history of sexual abuse. That child later alleged that D.I., who was ten-years-old at the time, tried to have sex with her. In addition, after D.I. was adopted, he engaged in sexually inappropriate conduct with his adoptive mother's five-year-old granddaughter. Because of his experiences in foster care, D.I remains “very disturbed,” “heavily sexualized,” and “a high risk for sexually harming children.” His experience is also typical of PMC children. The evidence reveals that child-on-child physical and sexual abuse is typical, common, and widespread throughout Texas foster care. Much of this evidence comes from witnesses who themselves repeatedly experienced physical and sexual abuse at the hands of other foster children in their placements.

To be sure, when child-on-child abuse occurs, “both children are victims. Even the one we would typically call the perpetrator is him – or herself a victim.” (D.E. 303 at 8). According to Dr. Miller, it is important “that both victims get the treatment services that they need and are not placed in environments where they or other children could be [in] danger.” Id. Specia agreed, “There's two victims in a child on child maltreatment incident,” and that child-on-child abuse “can cause tremendous harm to the children.” (D.E. 299 at 36). Specia further testified, “Sexually abused children can act out on other children.” Id. at 70. Likewise, Shaw agreed that “sexually abused [children] often become abusers themselves.” (D.E. 304 at 51). Shaw also testified that not knowing a child's history of child-on-child abuse results in “a heightened risk.” Id. Yet DFPS refuses to track child-on-child abuse. Black admitted that if a foster child, who had previously sexually abused another child, is placed “into a foster group home with young children ... no one's going to know about it at that foster group home” because DFPS does not track that. Id. at 67. McCall, Morris, and Specia all testified to similar effect. (See D.E. 299 at 35-37; D.E. 301 at 80; D.E. 305 at 79). This places all LFC children at risk, making it impossible to assure safe placements. The Court finds that RCCL's faulty investigations, including its decision to not track child-on-child abuse, cause an unreasonable risk of harm to all LFC children.

RCCL is also failing its licensing and inspecting duties. DFPS can choose from 13 types of enforcement actions against a licensed facility when it finds a violation. These options include working with facilities to voluntarily correct deficiencies, probations, monetary penalties, and denial, suspension, or revocation of an operator's permit. (DX 119 at 90). DFPS, however, almost never takes an enforcement action. During fiscal year 2013, CCL oversaw 10,286 licensed residential childcare facilities. Id. Over that time, CCL cited providers for 6050 violations, but only issued 12 corrective actions and 1 adverse action. Over the past five years, CCL issued only four adverse actions against residential operations. Instead of enforcement actions, DFPS chooses “Collaborative approaches like corrective plans, probation, and evaluation periods [that] can take up to one year or longer for operations to come into compliance.” Id. at 91. In the meantime, children remain and continue to be placed in these facilities. The Court would not be concerned if DFPS's strategy was working, but it is not. The Sunset Commission found that DFPS's enforcement effort “has not helped gain compliance with requirements intended to protect children.” Id. According to the Commission, “One consequence of a more relaxed regulatory environment can be seen in a high incidence of repeat violations that can result when regulated entities perceive that they will not be held accountable for ignoring the State's requirements.” Id. at 92. What is more, the Commission found that “[m]ost of these repeat violations occurred on the highest-risk standards.” Id. This has been going on for a long time: In 2011, PMU found that 65.6% of residential care operations had repeat deficiencies. (PX 1111 at 13). In 2012, 77.6% of residential care operations had repeat deficiencies. (PX 1074 at 12).

DFPS also fails to consistently and reasonably apply safety standards. (DX 119 at 93). According to the Commission, “certain standards that directly relate to a child's safety should leave little room for subjectivity or error in how they are enforced.” Id. Yet the Commission found “variations in citing critical safety standards” between different regions. Id. As a result, not all children across the state are equally protected in their placements. Id. The Commission also noted that approximately 36 percent of all residential standards violations “are overturned during administrative reviews” due to inadequate investigations.⁵¹ Id. at 94.

51 Reasons that standard violations are overturned during administrative review include investigators' failures to gather sufficient evidence and to appropriately support decisions with documentation. (DX 119 at 94).

CCL has closed one facility in the past five years, but it is a story of horror rather than optimism regarding enforcement. The Daystar facility in Manvel, Texas had a capacity of 141 children. Between 1993 and 2002, three teenagers died at Daystar from asphyxiation due to physical restraints. In most cases, the children were hog-tied. Beyond these deaths, there were reports of sexual abuse and staff making developmentally disabled girls fight for snacks. Numerous stakeholders, including the district attorney, spoke out against Daystar, but the facility kept its license. In November 2010, a fourth child died in what was ruled a homicide by asphyxiation due to physical restraints. Daystar's license was still not revoked until January 2011. DFPS allowed this facility—that was responsible for four deaths, numerous allegations of sexual abuse, and unthinkable treatment of developmentally disabled children—to operate for 17 years. (See PX 2172-75). The Court understands DFPS's concern that enforcement might affect placement availability. The Court does not understand, nor tolerate, the systemic willingness to put children in mortal harm's way. The Court finds that DFPS's inadequate licensing and inspecting causes an unreasonable risk of harm to LFC children.

Part of the problem, as with CVS caseworkers, is that RCCL workers are vastly overburdened. Chansuthus determined that RCCL inspectors and investigators in Region 3 would need 34 work days each month in fiscal year 2011 to complete their assigned tasks. (PX 2021 at 70). Since then, RCCL workers' workloads have only increased. The number of licensed residential operations has remained relatively constant over the years: 10,361 in 2009 compared to 10,286 in 2013. Yet the number of RCCL caseworkers declined from 136 in 2009 to 89 in March 2014. (PX 2021 at 21). Caseloads have markedly increased as a result. In 2012, monthly caseloads averaged 9.6 per RCCL worker. In 2014, the average was 18.4 for monitoring workers and 16.3 for investigators.⁵² (PX 640; DX 136). As with CVS caseworkers, excessive caseloads have contributed to turnover. The yearly turnover for RCCL investigators is a staggering 41.2%. (PX 1993 at 76). RCCL recognized this crisis and asked for additional staff in its 2016-17 funding request, to “allow for the caseload to average out to a more manageable workload, which would help reduce risk to children placed in residential operations and promote staff retention.” (PX 898).

52 In 2014, DFPS began reporting inspector and investigator caseloads separately. (D.E. 301 at 14-15).

Chansuthus reviewed RCCL's work and found that its caseworkers were “struggling with follow ups, they were struggling with conducting thorough risk assessments, they were struggling with doing safety plans, they were struggling with completing a number of the tasks that they were charged by their own standards and policies with doing.” (D.E. 324 at 118). Besides being full of errors, RCCL's investigations were often late. Only 58% of investigations were completed within the required 45-day timeframe. (PX 1118). Chansuthus “attributed the deficiencies to staffing issues.” (D.E. 324 at 118). RCCL workers echo this sentiment, saying they “feel overwhelmed and unable to complete timely [investigations.]” (PX 1117 at 5).

Also similar to CVS caseworkers, DFPS has not conducted an RCCL workload study. DFPS has no idea how much time each RCCL task takes or should take. Nor has DFPS established caseload ranges for RCCL workers. DFPS does not know what is a manageable caseload. Until a workload and time measurement study is done, Chansuthus explained, RCCL “has no management basis for determining exactly what its true needs are and cannot be certain it is addressing those needs in a manner that will support quality performance.” (PX 2021 at 23). As of trial, DFPS had no plan to find out. Thus, while DFPS asked for an additional 20 inspectors, 20 investigators, 9 supervisors, 1 program manager, and 1 state office program specialist in its 2016-17 funding request, it is aiming in the dark.

In sum, the Court finds that DFPS's insufficient oversight of its licensed foster care facilities has caused harm and an unreasonable risk of harm to LFC children.

a. DFPS is Deliberately Indifferent Toward RCCL Operations

DFPS has known for over a decade about RCCL's faulty investigations and the substantial risk of serious harm they pose to LFC children. In 2004, a Comptroller Report mentioned investigation shortcomings as a major concern. A 2007 report reiterated the 2004 report's worry and recommended “thoroughly investigating all complaints, allegations or reports and making the results public.” (PX 1966 at 29). In 2012, PMU reviewed 89 investigations from the year prior and concluded that 20% “should not have been approved” by supervisors because “most of those were inadequate in the preponderance explanation.” (PX 1074 at 21). A year later, PMU and Shaw reviewed the 111 UTD dispositions discussed supra , which came back with a 75% error rate. Each year there are around 5000 investigations, 2000 of which are for abuse and neglect. Due to RCCL's systemic failures, children are left with their abusers without receiving necessary treatment, and adult perpetrators continue to house foster children with nothing indicating a risk. The substantial risk of serious harm to LFC children from faulty investigations “is obvious.” Hernandez , 380 F.3d at 881. DFPS has presented no evidence that it has acted reasonably to this known risk. Farmer , 511 U.S. at 844, 114 S.Ct. 1970.

Likewise, the Court finds that DFPS was aware of the substantial risk of serious harm posed by not tracking child-on-child abuse. The 2004 Comptroller Report recommended “keeping sexually abused children separate from other children; tracking and reporting the number of reports it receives concerning child-on-child physical and sexual abuse by facility.” (See PX 1966 at 29). The 2007 Texas Appleseed Report, which was given to DFPS, reiterated these recommendations. (Id. ; see also D.E. 299 at 66-67). In November 2012, DFPS explained in an internal memo that “Children who are removed from their homes and brought into the foster care system are victims of abuse and/or neglect. As a result of their abuse and/or neglect, these victim children may display a number of inappropriate behaviors with each other in the home or at school.” (DX 21 at 1). In the same memo, DFPS said it “must focus on safety and treatment for all child victims through proper notification to all key parties of an RCCL investigation, careful monitoring, and case management intervention.” Id. at 1-2. In July 2013, DFPS again acknowledged, “Incidents involving allegations of child victimization in foster care require immediate attention from CPS.” (DX 34 at 1). At trial, experts for both parties said that children with sexual abuse histories should be placed in single-child placements. (See, e.g. , D.E. 328 at 161; D.E. 329 at 161; D.E. 303 at 8). Likewise, Specia admitted that children who are known victims of sexual abuse will act out with other children. (D.E. 299 at 70). Yet DFPS does not track child-on-child abuse, which is unimaginable to other child welfare experts. (D.E. 303 at 6-7). As a result, children like D.I. unnecessarily experience abuse that affects them—as well as the children they interact with—the rest of their lives. Further, DFPS has no plans to begin tracking this “critically important” piece of information. What is more, it appears that DFPS is equipped to track child-on-child abuse, but simply refuses to do so. As recently as July 2010, RCCL investigation reports had checkboxes for “Child on Child Sexual Abuse” and “Child on Child Physical Abuse.” (See PX 185 at 46 (filed under seal)). Even though this information was not tracked in a centralized, easily retrievable fashion, it at least was noted in the RCCL CLASS records for each child. According to Morris and Shaw, however, those checkboxes are now outdated or inaccurate, and are no longer used. (D.E. 301 at 80; D.E. 304 at 45-46). The Court holds that DFPS is deliberately indifferent toward its faulty investigations, including not tracking child-on-child abuse, which causes an unreasonable risk of harm to the LFC Subclass.

DFPS is also deliberately indifferent toward RCCL's enforcement. A 2008 report given to DFPS listed “licensing oversight” as a problem area. (PX 1273 at 779). Yet DFPS continues to under-regulate licensed facilities. In fiscal year 2013, CCL cited licensed facilities for 6050 violations, but only issued 12 corrective actions and one adverse action. Over the past five years, CCL has issued four adverse actions against residential operations. CCL has closed only one facility in the past five years, but not until the fourth homicide. When confronted with recommendations on how to improve, CCL uniformly ignores them. For example, CCL employs risk analysts to determine “the most appropriate action for reducing the risk of harm to children in a licensed facility.” (DX 119 at 93). CCL rarely heeds their advice, and has never followed a recommendation to impose administrative penalties. Id. at 91, 93. CCL also ignores its internal quality control experts at PMU. In 2011 and 2012, CCL did not follow a single license revocation recommendation. (PX 1111 at 10, 12). As discussed supra , after PMU's and Shaw's review of UTD cases, “No licenses were suspended .... None were revoked .... No penalties were established on any of these facilities ... [and] the State didn't move any of the children.” (D.E. 301 at 50-51). The substantial risk of serious harm posed to LFC children by improperly licensed and regulated facilities “is obvious.” Hernandez , 380 F.3d at 881. DFPS produced no evidence that it has acted reasonably to respond to this risk. The Court holds that DFPS is deliberately indifferent toward its inadequate licensing and enforcement, which causes an unreasonable risk of harm to the LFC Subclass.

The Court also finds that DFPS is deliberately indifferent toward the substantial risk posed by an inadequate RCCL workforce. In 2010, RCCL staff reported that caseworker and supervisory responsibilities were “too large” and workloads too “intense” to “effectively monitor operations.” (PX 1937 at 17). Since that time, RCCL's workforce has shrunk by over 30% while total cases remain the same. RCCL knows that manageable workloads “would help reduce risk to children placed in residential operations and promote staff retention.” (PX 898). Yet DFPS has not conducted a workload study of RCCL workers, established caseload ranges for RCCL workers, or determined what a manageable caseload would be. Although RCCL has asked for additional staff, the request is not grounded in any knowledge about RCCL's or LFC children's needs. This a step in the right direction, but without evidence of how the additional staff will correct RCCL's systemic problems, the Court finds that DFPS's response is not reasonable.

In sum, DFPS is aware that faulty investigations, not tracking child-on-child abuse, inadequate licensing and enforcement, and an insufficient workforce all pose a substantial risk of serious harm to LFC children. DFPS has consciously disregarded each of these risks. The Court therefore finds that DFPS is deliberately indifferent toward the LFC Subclass.

b. DFPS Substantially Departs from Professional Judgment Toward RCCL Operations

Failure to act on a 75% error rate on abuse and neglect investigations, when an acceptable average is 2% or 3%, shows an absence of professional judgment. Not tracking child-on-child abuse even though it is “critical” to foster children's safety shows an absence of professional judgment. Issuing one adverse action for 6050 violations, contrary to recommendations by two quality control boards, shows an absence of professional judgment. Whether DFPS exercised professional judgment regarding RCCL's workforce is the only question needing more than a moment's contemplation.

RCCL asked for 51 more staff in its 2016-17 budget request, which would bring its workforce up to 140—up from 136 in 2009. All the evidence, however, reveals that 140 people cannot handle RCCL's workload. Although a decision “if made by a professional, is presumptively valid,” the Court is convinced that DFPS's response was “such a substantial departure from accepted professional judgment, practice or standards” that it was not based on such a judgment. Youngberg , 457 U.S. at 323, 102 S.Ct. 2452.

It is commonsense that an agency should know what a manageable caseload is before it determines how many workers it needs. As the CWLA explains, “Every agency should conduct a workload analysis to determine the appropriate workload standards for its child protective services staff.” (PX 2114 at § 5.9). Completing a workload analysis gives an agency “a true picture of how much time and effort is required to do the tasks required for a specific position—and to allow [an agency] to know, given that this is how much time is required to complete these tasks, how many staff [it] need[s] to have in order to handle the work.” (D.E. 324 at 120). This does not strike the Court as revolutionary. DFPS, however, has not conducted a workload study, nor does it have plans to do so.

Furthermore, DFPS's staffing request only returns the workforce to 2009 levels. Yet problems with investigations, licensing, and tracking child-on-child abuse began well before 2009. Virgil aged out of foster care before 2009. She testified that she was sexually abused by one of her foster families in a foster group home. When she reported the abuse, “nothing ever happened.” Id. at 200. Virgil does not know if the incident was even investigated. Likewise, Bentley entered foster care when she was two years old in the mid-1990s, and was sexually abused by foster parents and other foster children “Nonstop.” Id. at 63. Bentley was also physically abused, frequently showing up to school with bruises. She believes that people called CPS, but to her knowledge few investigations were ever done. Id. at 66-67. When investigations were done, and abuse was found, no actions were ever taken against the facilities. Id. In short, RCCL did not base its staffing request on any identifiable professional standard, and their proposal would only return staffing levels to a time when the problems Plaintiffs allege were rampant. The Court therefore finds that DFPS's decision was “such a substantial departure from accepted professional judgment” that DFPS “did not base the decision on such a judgment.”

Regardless, the central inquiry is not whether DFPS exercised professional judgment in regards to RCCL staffing. Rather, it is whether DFPS exercised professional judgment in regards to RCCL's operations as a whole. See Alberti , 790 F.2d at 1224 (“In determining the constitutional question, we need not separately weigh each of the challenged institutional practices and conditions, for we instead look to ‘the totality of conditions.’ ”) (citation omitted). Even if the Court finds DFPS's staffing request valid—which it does not—an uptick in workers is not enough to cure RCCL's systemic deficiencies. The Court has seen no evidence that 20 more investigators will cure DFPS's 75% investigation error rate, or that 20 more inspectors will begin enforcing licensing standards when RCCL has been ignoring two quality control boards for years, or that new workers will begin tracking child-on-child abuse. Returning the RCCL workforce to 2009 levels is not a silver bullet considering that Plaintiffs other allegations were well ingrained at that time.

The Court finds that DFPS's disregard of investigations, child-on-child abuse, enforcement, and appropriate staffing levels shows that DFPS substantially departs from professional judgment toward RCCL operations. 2. The State Maintains an Inadequate Placement Array

Plaintiffs allege that DFPS has a policy of maintaining an insufficient number, geographic distribution, and array of placements. This “inadequate placement array,” as Plaintiffs call it, is a structural deficiency that forces LFC children to be placed in settings that cause an unreasonable risk of harm. The Court agrees.

Texas policy requires that DFPS “find the best available placement option” for foster children. (PX 50 at 5). To determine that option, DFPS must consider the child's “immediate and long-term personal needs and the child's need for continuity, stability, and permanency.” Id. at 6. To that end, “the placement must be within the same county (or within 50 miles of the parents' home if the home is near the county boundary line) unless an exception is justified.” Id. at 7. When appropriate, placements should be with families as opposed to group care facilities, because “Families are considered to be least restrictive settings for children.” Id. at 4. Further, “Whenever possible,” children should be placed with their siblings. Id. at 8. In short, foster children should be placed with their siblings in the least restrictive, most family-like settings available that are close to their home community. Id. at 6. Plaintiffs allege that these policies cannot be satisfied because of DFPS's inadequate placement array.

To reiterate, Plaintiffs do not have a constitutional right to be placed in the least restrictive, most family like placement, or placed with their siblings, or placed close to their home community. Violating Texas policy does not necessarily equate with violating the Constitution. To have a cognizable substantive due process claim, Texas's inadequate placement array must violate Plaintiffs' right to be free from harm and an unreasonable risk of harm.

It is beyond question that DFPS's placement array is inadequate. First, “there is an imbalance in geographic distribution of services.” (PX 1864 at 19). In other words, while some counties have excess capacity, other counties have fewer than 0.5 beds per child entering care. (D.E. 315 at 43-44). It is not just the number, but the types of placements that are also poorly distributed. When a Basic-level child enters care in a county that only has space for higher level needs children, that child will be sent away. As a result of Texas's inadequate array, 39% of children are placed out of region and 60% are placed out of county. (Id. at 41-42; D.E. 300 at 101; DX 183 at 4). This is a “high number.” (D.E. 303 at 90). Attorney ad litem Stukenberg testified that is typical for DFPS to “ship[ ] very young children all across the State, especially to emergency shelters” in search of an appropriate placement. (D.E. 324 at 216). It is also typical for foster children to sleep in CPS offices—on inflatable mattresses without sheets or pillows—because there are no available beds. (DX 18 at 1; PX 1217 at 5; D.E. 327 at 201-02). DFPS recognizes that its “imbalance in geographic distribution of services” results in “children and youth without placements,” and “children and youth placed out of their home regions due to unavailability of appropriate placement resources within their regions.” (D.E. 315 at 34-36; PX 1217 at 8-9).

Second, DFPS's inadequate array results in “children and youth not placed with their siblings due to unavailability of appropriate placement resources.” (PX 1864 at 19; PX 1271 at 8-9). “[A]ccording to established professional standards and federal and state policy, children should be placed with their siblings unless doing so is inadvisable due to a therapeutic or safety concern.” (PX 2037 at 39). Texas policy says the same. (PX 50 at 8). Yet Texas's foster children “are separated from their siblings at an alarming rate.” (PX 2037 at 59). As of June 2014, only 64.7% of siblings groups were placed together. (DX 142 at 1). For comparison, Tennessee places 84% of sibling groups together. (D.E. 303 at 73).

Third, foster children are placed in inappropriate facilities. Specifically, DFPS relies too heavily on congregate care facilities, which “are not conducive to supporting youth in engaging activities that help them ‘practice’ for adulthood, or to helping young people build social capital.” (PX 2037 at 38-39). As of September 2012, 13.2% of children who were age 12 or younger, and who had been in care for at least 18 months (i.e., PMC), were placed in either group homes or institutions. (PX 2128 at 1). The nationwide average is 4.9%. Id. Texas ranks second worst, and the State is increasing its reliance on group care settings. (D.E. 300 at 60).

Fourth, sexually aggressive children are not placed in single-child homes, the safe and appropriate placement. As discussed supra , sexually abused children often become “sexualized” and initiate abuse in the future. See p. 207. That is why, Gonzalez explained, sexually abused children should be in homes where they are the only child. (D.E. 328 at 161). Defendants' expert P. Wilson agreed: “You look for a home where that would be the only child in that home.” (D.E. 329 at 161). DFPS, however, does not track how many single-child foster homes are available. After several prods from the Court, DFPS produced data on how many of single-child homes are in use. (See, e.g. , D.E. 320 at 50-52; D.E. 328 at 164; D.E. 329 at 162). Of DFPS's 12,000 foster homes, 262 house one PMC child, but DFPS could not say if there were other adoptive or biological children in the homes. (D.E. 330 at 127). It is unclear whether these homes were designated as single-child placements, or if that occurred by happenstance. It is also unclear if the children in these homes were placed there because of abuse backgrounds or otherwise. Defendants could not confirm how many of these homes, if any, were reserved for abused children. Considering that sexually abused children are frequently placed with other children, sometimes in the same room, the Court is certain that DFPS does not maintain enough single-child homes. See supra p. 207.

Plaintiffs also allege that DFPS's inadequate array causes frequent placement moves. The Court, however, is not convinced. Rather, placement moves are more a product of changing levels of care or behavior issues. An inadequate array may cause an initial inappropriate placement, but the Court does not follow Plaintiffs' logic in how an inadequate array causes additional moves. The Court does find that frequent moves are the norm among PMC children. As of August 2014, 28.7% of PMC children had experienced five or more placements, and 7.69% had experienced ten or more placements. M.D. had 19 placements, A.M. had 21, and K.E. had 27. Former foster child Jackson had 40 over six years. (D.E. 324 at 184). Shuttling children “from pillar to post and staying for short period of time” is undoubtedly traumatic. (D.E. 303 at 78). For example, J.S. had ten placements in seven years, some over 500 miles from his home county. He attended eight different schools, each in different counties. J.S. is now “a virtual non-reader.” (D.E. 326 at 115). According to Carter, J.S. has “severe attachment deficits” in part because “he was provided no stability whatsoever.” Id. at 114. Likewise, Carter explained that A.M. “had no sense of belonging,” in part because any time she achieved relative stability, DFPS moved her level of care improved. Id. at 122. Carter testified that the impact of frequent moves “from a psychological standpoint is potent” and that “identity disorders are rampant within that population of children.” Id. at 97, 128. Nonetheless, Texas's inadequate placement array does not cause frequent moves. Instead, the effects of the inadequate array—placements outside of home communities, sibling separation, an overreliance on congregate care facilities, not placing sexualized children in single-child homes—are exacerbated by frequent moves. The Court now turns to whether these effects cause an unreasonable risk of harm to Plaintiffs.

When children are removed from their home communities, they are uprooted from “everything ... that's familiar.” (D.E. 303 at 90). DFPS policy requires home community placements because “Children and youth who have been removed from their homes have a high need for stability and continuity of care.” (PX 50 at 7). As Carter explained, removing children from their family, friends, school, and church “may disrupt existing attachments and impact a child's ability to form new ones.” (PX 2015 at 6). A lack of stability “can be damaging to the psychological health and development of children in foster care,” and can lead to “any number of psychological issues including depression and defiance.” Id. DFPS's inadequate array causes instability. It moves children “hundreds of miles away from their home communities, siblings, schools, and other supports,” “disrupting their school and family relationships.” (DX 119 at 45; PX 1877 at 32). Regional Director Bowman agreed that placing children far from home negatively affects the bonds and connections that children need to be emotionally healthy. Yet 39% of children are placed out of their home region and 60% are placed out of their home county. The combined effects of the inadequate array and frequent moves means that any stability a PMC child may gain in her initial placement is destroyed when she is relocated. For example, the Court counts at least four occasions where J.S. was sent over 200 miles to new placements. Likewise, K.E. was twice moved over 250 miles. Because K.E.'s 27 placement changes took her all over Texas, she attended at least 12 schools in ten years. Like many of the Named Plaintiffs, K.E. had to repeat grades and functions academically well below her age.

Cross-state moves also impede primary caseworkers' ability to visit their foster children. As discussed supra , primary caseworker are foster children's lifelines. See Section IV.A. Yet, due to a lack of funding and a lack of time, primary caseworkers “can't go and meet with [children] face to face” when they are placed out of county or region. (D.E. 324 at 22). Thus, foster children are forced to rely on secondary I See You workers. McCall acknowledged at trial that I See You workers were created by DFPS ten years ago at least in part because children were too often being sent far from their home communities and primary caseworkers. (See D.E. 305 at 62). Although secondary workers do not protect foster children from an unreasonable risk of harm, they undoubtedly help Texas represent to the federal government that caseworkers visit children in foster care at least once per month. If the percent of children in Texas visited by caseworkers monthly were to drop below 90% (and after the 2015 fiscal year, below 95%), Texas would lose millions of dollars in federal aid under Title IV-B of the Social Security Act. See 42 U.S.C.A. § 624(f)(1). As discussed supra , the quality of I See You worker visits is inconsistent and often questionable. See pp. 171-76. The Court finds that moving children far from their home communities causes an unreasonable risk of harm to the LFC Subclass.

In addition, separating siblings harms foster children. Carter explained that entering foster care “is in and of itself a frightening experience.” (D.E. 326 at 119-20). When entering care, siblings often become “all one another has to rely on.” Id. at 118-19. When they “are split up for whatever reason,” foster children feel “hopeless ... [and] helpless because even the people they came into care with, their own siblings, are no longer around to provide them support and a safety net, a sense of belonging.” Id. at 119-20. Carter also explained, “Not only do [foster] children suffer the hurt of permanent removal from adult family members, but severing them from siblings re-traumatizes them and adds further severe damage to the youth's sense of belonging.” (PX 2015 at 10). For example, A.M. “expressed a great deal of angst over the loss of her relationship with her three younger half-siblings.” Id. In short, as DFPS recognized in 2008, “Children suffer trauma when separated from siblings.” (PX 1273 at 20). That is why DFPS policy requires keeping siblings together “Whenever possible.” (PX 50 at 8). DFPS's inadequate array, however, forces sibling apart at an “alarming rate.” (PX 2037 at 59). Further, when siblings do separate, DFPS policy requires that sibling visits occur frequently. (PX 52 at 87). Yet former caseworkers testified that when children are placed in separate counties or regions, “arranging a sibling visit is a nightmare” because caseworkers do not have the time or the funding. (D.E. 324 at 22-23). For example, when L.H. and her six siblings were separated, the department was not able to arrange visitations in spite of the siblings crying in court over the separation. (D.E. 327 at 195-97). The Court finds that separating siblings at an alarming rate creates an unreasonable risk of harm to the LFC Subclass.

Inappropriate placements also harm foster children. The 2010 Texas Appleseed report found that DFPS has “a history of poor foster care placements.” (PX 1988 at 10). In part, DFPS admitted, this is because “information for matching a child to a placement is often inadequate.” (D.E. 300 at 57-58). A former primary caseworker testified that “placements would sometimes be made based on the availability of a bed rather than the appropriateness of the match.” (D.E. 323 at 45). Solis, the attorney ad litem for S.A., said CPs's only response to poor placements is “that is the only one that's available.” Id. at 30. Due to DFPS's inadequate array, M.D. stayed at a facility for over a year after alleging that a staff member raped her because DFPS “didn't have anywhere else to move her.” (D.E. 324 at 223-24).

In particular, DFPS over-relies on GROs. A GRO, also called a congregate care placement, is “a child-care facility that provides care for more than 12 children for 24 hours a day.” Tex. Hum. Res. Code Ann. § 42.002(4). GROs include RTCs, halfway houses, emergency shelters, and therapeutic camps, and may be a single building or a campus with multiple cottages. (Id. ; see also D.E. 300 at 68). They are much more restrictive and less family-like than other placements. (PX 50 at 7). Dr. Miller testified that “we know” that placing children in the most family-like setting possible, as opposed to GROs, “is not only what's best for children, it's absolutely essential if those children are going to be whole and remain whole.” (D.E. 303 at 51-52). Professional literature reflects that congregate care placements for young children “lead to poor developmental outcomes,” and “qualitative studies have raised concerns about the negative impact of group care settings on older children, especially in terms of safety.” (PX 2037 at 54). Yet DFPS is second worst in the nation at placing young children in congregate care. GROs are designed to address serious therapeutic needs, which is why DFPS discourages placing Basic-level children in GROs except for short-term, emergency cases. (See PX 50 at 29-30). Nonetheless, at the time of trial, there were 356 Basic-level PMC children in non-emergency GROs. (D.E. 300 at 74-75). DFPS's overreliance on GROs effectively institutionalizes children that do not belong in these placements. Part of the problem is that, unlike children who live in foster family homes, children in GROs have to navigate significant bureaucracy to simply participate in extracurricular activities or spend time at friends' houses. It took former foster child Arce “almost a year just to get a sponsor” to visit with a friend who lived outside the GRO. (D.E. 324 at 55). Further, GROs' strict regulations keep children from learning basic life skills. Children at GROs never learn how “to use a knife to cut their food because they're not allowed to have knives where they live.” Nor do they learn “how to make a medical appointment for themselves” or “how to answer a phone or take a message or leave a message for anybody.” (D.E. 307 at 29-30). Attorney ad litem Langsley testified that D.P., who had most of her time in congregate care facilities, left foster care without knowing how to use a computer or fill out a job application. (D.E. 323 at 75). When A.M. left foster care, “She didn't know you could go to the store and buy something. It was a foreign concept for her.” Id. at 98. In Carter's opinion, A.M. aged out of care in a manner typical of many foster children, unprepared for “adulthood, for marriage, family life, social settings, [and] occupation.” Id. at 136, 194.

In addition to leaving foster care without basic life skills, DFPS's inadequate array assures that many of the 1300 children who age out of care every year leave the system far from their home communities. The most tragic consequence the Court heard comes from Sharp. During his eight years in foster care, Sharp was placed in 18 or 19 congregate care settings. When one of his congregate care facilities in Houston was closing, the only available placement—in the entire State—was in the Panhandle where he had previously been abused. Sharp told the Court, “I decided that I would rather leave the system and just try to make it on my own so I packed some things up and I left.” (D.E. 325 at 178-79). He spent the next six months homeless in Houston, 600 miles away from his home community of Dumas, Texas, “living on top of a shopping strip,” making money by donating blood and prostitution. Id. at 178-80. When he went to donate blood a third time, “they told me I couldn't do it anymore because I was HIV positive.” Id. at 180. Likewise, the last time M.D. was seen after she ran away from her Houston RTC, she was involved in the Houston sex trade, 200 miles from her home community in Corpus Christi.

It also appears that a corollary of inappropriate placements is overmedication. M.D., for example, was rotated between 16 psychotropic drugs while in care. Sharp explained that “Every foster child apparently has Oppositional Defiant Disorder, ADHD, a wide of things that certainly I don't think are true.” Id. at 178. While he was in foster care, Sharp took a number of psychotropic drugs for Bipolar Disorder, Schizophrenia and ODD. Id. Soon after aging out, Sharp visited a psychiatrist who told him to stop taking his medications because he did not have those disorders. Id. Langsley testified that D.P. went through 54 placements while in foster care. (D.E. 323 at 70). As D.P. shifted through placements, her medications were “constantly being adjusted, readjusted.” Id. at 71-72. Prior to one of Langsley's visits, D.P. was “chemically restrained,” meaning she was injected with a drug “cocktail” that a physician prescribed. Id. at 73. The physician “left it up to the staff to administer however many shots they wanted to.” Id.

Lastly, placing sexually abused children with other children in foster family homes and foster group homes causes an unreasonable risk of harm. Sexually abused children either need to be placed in GROs where there is strict oversight, or in single-child placements. (See D.E. 328 at 161; D.E. 329 at 161). Yet sexually abused children are often placed alongside other children in loosely regulated environments, sometimes in the same room.

In sum, the Court finds that DFPS's inadequate array places children far from their home communities, separated from their siblings, and in inappropriate placements. The Court also finds that the “totality of conditions” causes an unreasonable risk of harm to LFC children, which are exacerbated by the frequency with which children are shuttled around the State. See Alberti , 790 F.2d at 1224. The Court holds that DFPS's inadequate placement array causes an unreasonable risk of harm to the LFC Subclass.

a. DFPS is Deliberately Indifferent Toward its Placement Array

DFPS has known about its array problems for years. In 2006, DFPS knew there were “children without placement[s].” (DX 18 at 1). By 2007, DFPS knew that it “frequently moved children from one caregiver to another.” (PX 1966 at 6). By 2008, DFPS knew that it had “placement quality and capacity issues,” including:

• Children and youth without placements—children in settings supervised by CPS due to unavailability of appropriate placement resources and/or provider rejection of the referral;

• Children and youth placed out of their home regions due to unavailability of appropriate placement resources within their home regions;

• Children and youth not placed with their siblings due to unavailability of appropriate placement resources; [and]

• Lack of available and appropriate placement options for children and youth with challenging service needs[.]

(PX 1271 at 8-9). Another 2008 report identified “sibling group issues” as an area of concern, noting an “Increased difficulty in finding appropriate homes” and that “Children suffer trauma when separated from siblings.” (PX 1273 at 20). That report also said that DFPS lacks resources in rural areas. Id. at 23. DFPS reiterated these problems in a 2010 report to the Texas Legislature, saying, “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 1286 at 23). In 2011, CPS recognized that “Too many children in foster care are placed outside their home communities away from siblings, peers, families, school, churches and support networks.” (PX 939 at 1). Specia echoed these concerns at trial. (See D.E. 299 at 39). The Court holds that DFPS was “aware of facts from which the inference could be drawn that a substantial risk of serious harm exists.”

The Court is also convinced that DFPS drew that inference. DFPS agrees “there is a causal link between moving a child far from his home community and [his] outcomes.” (D.E. 300 at 54). At trial, Black acknowledged that moving children “outside their home communities and apart from their siblings ... lead[s] to poorer outcomes for these children.” Id. Two Directors of DFPS Regions agreed that it is “absolutely not good for children” to move away from home communities, as it makes it “more challenging” to stay connected to family and remain emotionally healthy. (D.E. 311 at 40; D.E. 312 at 34-35). The person in charge of revamping DFPS's array explained that “placing children close to home, placing children in the least restrictive placement setting, placing children with their siblings, maintaining connections for children with those that are important to them” all have to do with “the overall safety of children in foster care.” (D.E. 328 at 182-83).

The Court also finds that the risks of harm posed by an inadequate array are “obvious.” See Hernandez , 380 F.3d at 881. The Court did not need to read a volume of studies to figure out that placing children hundreds of miles from all that is familiar, separating siblings, housing children in facilities that are inappropriate for their needs, leaving children in facilities where they have been abused, and placing sexualized children in the same room as other children without proper oversight, all present substantial risks of serious harm. Reading the studies that DFPS has had access to for nearly a decade, and hearing the testimony from DFPS employees, confirmed that DFPS actually knew what was “obvious.”

The remaining inquiry to determine if DFPS is deliberately indifferent is whether DFPS “responded reasonably to the risk.” DFPS points to Foster Care Redesign as its reasonable response. Launched in January 2010, Foster Care Redesign is DFPS's solution to its inadequate array.⁵³ (PX 969 at 3). Under DFPS's old system or “legacy system,” the agency contracts with around 300 private CPAs for 90% of foster placements. (DX 119 at 45). DFPS directly provides the remaining 10%. Id. The legacy system uses an “open enrollment” process to procure residential childcare services, but it does not ensure that providers locate in places where services are needed, which creates the geographic imbalance in services that forces children out of their home communities. (PX 969 at 4). For example, there is no residential treatment center in Texas south of San Antonio. (D.E. 327 at 203; see also Attachment 2). Furthermore, under the legacy system, “Because very few providers have all the placement settings and services that are necessary to meet the needs of children regardless of service levels, children move from provider to provider when service levels change.” (PX 969 at 4). In other words, when children improve or decline, they are moved. The legacy system also creates perverse incentives by paying providers more money for children with higher service levels. It is therefore in a provider's financial interest for children to not improve.

53 DFPS was clear that Foster Care Redesign is only intended to fix its inadequate array, not caseworker turnover, not RCCL investigations or licensing, and not how foster group homes are regulated. (D.E. 316 at 6).

Foster Care Redesign does away with “open enrollment.” Instead of contracting with private providers to run operations wherever they choose, Foster Care Redesign contracts with “Single Source Continuum Contractors (SSCC) that provide a full continuum of paid foster care services designed to meet the needs of all children who enter care in the designated [geographic] area” that DFPS chooses. Id. at 5. In this way, children will not have to relocate hundreds of miles if their service level changes. Although DFPS's caseworkers' responsibilities remain the same, SSCC's are “accountable for well-being and permanency outcomes, that support placing children close to home in the least restrictive, most family-like settings.” Id. Foster Care Redesign also does away with the payment incentive structure. Lastly, Foster Care Redesign is cost-neutral, so DFPS spends no more on SSCCs than it paid providers under the legacy system.

The Court is encouraged by the idea of Foster Care Redesign, but discouraged by its results. More than five years into being, the program is active in only 2% of the counties in Texas, covering 800 foster children. (D.E. 316 at 4-5; D.E. 328 at 201). In spite of numerous recommendations, DFPS has not set a timeline or implementation schedule. (D.E. 316 at 32). In addition, DFPS has not:

• Clearly delineated and defined the case management roles and responsibilities of DFPS and the single-source continuum contractors;

• Identified training needs and addressed long-range and continuous plans for the training and cross training of staff;

• Articulated plans for evaluating the costs and tasks involved with each single-source continuum contract procurement to better inform future resource needs;

• More formally communicated plans for evaluating the performance of contractors and the foster care redesign system as a whole; or

• Reported on transition issues resulting from redesign implementation.

(DX 119 at 49-50). DFPS simply has “no roadmap for the overall effort.” Id. at 49. Until it does, “Texas remains unprepared to manage current and future foster care redesign efforts.” Id. at 50. This is particularly troubling because no other state has a similar model that DFPS can use as a guidepost. (D.E. 316 at 15-17). All told, the “foster care re-design model to outsource the administration of the foster care system is a risky endeavor.” (DX 119 at 48).

In the five years of Foster Care Redesign, DFPS has entered only two SSCC contracts—one was an abject failure, and there is no data on the other. The first SSCC, Providence Service Corporation, was hired by Specia in spite of recommendations not to hire them “based on Inconsistencies and Areas of Risk.” (PX 978 at 11). Far from placing children in the least restrictive facilities, Providence placed more children in RTCs and emergency shelters than the providers under the legacy system. (D.E. 300 at 87). Only 57.9% of children under Providence's management were placed in foster family homes, compared to 70.3% of children statewide, and 70.8% of children in the regions where Providence operated. (PX 937 at 3). Children still slept in CPS offices because Providence “wasn't finding placements for them.” (D.E. 300 at 87-88). Moreover, Providence failed to place children close to their homes—the central aim of Foster Care Design. In fact, Providence only placed 29% of children within 50 miles of their home—7% below their contract baseline, and worse than the legacy system's figures. (D.E. 316 at 17-19). Despite Providence's abysmal performance, DFPS did not end the contract in part out of fear that Providence would sue them. Id. at 27, 29. Providence quit after a year and a half because they lost $2 million and could not make the contract work.

DFPS has only contracted with one other SSCC, but performance data is not yet available. That said, DFPS almost contracted with Lutheran Social Services of the South before the Providence contract. DFPS tentatively awarded an SSCC contract to Lutheran, but backed out at the last moment because of an adverse licensing action against them. (DX 17 at 5). Defendants could not tell the Court the reason for the adverse licensing action, though Reinhardt, the Director of Foster Care Redesign, agreed that it “could have been” related to the deaths of children in Lutheran's care around that time. (D.E. 316 at 23-26).

The Court holds that Foster Care Redesign is not a reasonable response to the risks associated with DFPS's inadequate array, and therefore DFPS is deliberately indifferent. Foster Care Redesign is a proven failure that Texas has not attempted to fix. The only data available shows that Foster Care Redesign has made Texas's placement array worse. Fewer children were placed close to home. Fewer children were placed in appropriate settings. Children were still sleeping in offices. The SSCC lost $2 million in a year and a half and quit. This is not a sustainable or reasonable program. The Court does not know why DFPS's first instinct is to reinvent the wheel. This was displayed in its stage-counting approach, which masks caseworker workloads so that even DFPS officials do not know how many children each caseworker handles. It is also displayed in Foster Care Redesign, which is unlike any other child welfare privatization effort. The Court has seen no evidence that DFPS has tried to fix the program despite the failures with Providence. The Court cannot imagine how DFPS will fix Foster Care Redesign, let alone its placement array, if it must remain cost-neutral.

Furthermore, Foster Care Redesign has reached only 2% of Texas over five years. Even if it performed well, the Court would be pressed to find that a program with such a small reach—and no timetable for expansion—is a reasonable response to statewide problems. Further, Plaintiffs point out that after trial the House and Human Services Committee recommended that the program be put on hold. (D.E. 345 at 57). In sum, even though Foster Care Redesign may have been conceived with good intentions, it was defective and half-baked from the start. Foster Care Redesign is an unprecedented program that never addressed many critical areas, that never had an implementation plan, and that may no longer exist. With Foster Care Redesign as DFPS's sole response, the Court holds that DFPS is deliberately indifferent toward its inadequate placement array, which causes an unreasonable risk of harm to the LFC Subclass.

b. DFPS Substantially Departs from Professional Judgment Toward its Placement Array

The Court also finds that DFPS substantially departs from professional judgment toward its placement array. As discussed, Foster Care Redesign is a unique program. CPS testified that there is none similar. It was therefore not based on another state's professional judgment, practice, or standard. Likewise, DFPS has never done a formal assessment of its placement needs across the State. DFPS does not know how many children are at each care level, where those children are located, or its distribution of placements for each care level across the State. Thus, DFPS does not know where and in what ways its array needs to be improved. It is not that DFPS did not think of this obvious step. A 2008 report explicitly said, “To address current substitute care placement capacity issues and meet future placement needs, a strategic plan based on a comprehensive statewide assessment of substitute care placement capacity and related services is needed.” (PX 1273 at 9). The report went on to say, “Capacity needs must be assessed by reviewing the individual needs of children and comparing their needs to the availability of appropriate substitute care placement within reasonable proximity of the children's homes.” Id. DFPS never performed a comprehensive study. Instead, DFPS appended two pages onto a progress report, which included no figures on capacity or needs. (DX 17 at 13).

DFPS did conduct a “bed capacity analysis” in 2010, but that study only assessed the number of foster children and beds in each county. This study did not include service levels, which is crucial for placing children correctly. Nor did this study consider keeping sibling groups together. The study also depended on data from 2009, and noted that more recent data were needed. (PX 1277 at 1, 3).

In short, DFPS's attempted fix was an unprecedented program that was not based on any data, and has proven to be a failure. While a decision made by a professional is presumptively valid, the Court finds DFPS's response “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, 102 S.Ct. 2452.

C. Foster Group Home Subclass

The Foster Group Home (“FGH”) Subclass includes all members of the General Class who are now or will be in a foster group home. Plaintiffs allege that DFPS operates its foster group homes in a constitutionally deficient manner. Specifically, Plaintiffs claim that DFPS does not properly train its FGH staff or caregivers, has an insufficient number of FGH staff, does not require 24-hour awake-night supervision at FGHs, and does not restrict the placement of unrelated children of different genders, ages, and service levels in the same home. Plaintiffs argue that these policies and practices amount to a structural deficiency that violates the FGH Subclass's right to be free from an unreasonable risk of harm. The Court agrees.

Like many of DFPS's undertakings, foster group homes are unique to Texas. (See, e.g. , D.E. 300 at 60-61). FGHs are licensed to provide residential care for 7 to 12 children. (PX 67 at 16). The limit of 12 children is supposed to include “any biological and adopted children of the caregivers who live in the foster home.” Id. at 291. It seems, however, that those children are not counted, allowing capacity to exceed 12. (D.E. 310 at 60; D.E. 326 at 25-26). As of August 2014, 932 PMC children lived in FGHs. (DX 268). FGHs appear to be a hybrid between foster family homes (that contain one to six children) and congregate care facilities like GROs and RTCs (which do not have a capacity limit), for which DFPS plucked the least-demanding standards from each type of facility. FGHs reside in a grey-area that simultaneously provides fewer benefits than foster family homes and fewer safeguards than congregate care facilities.

It is difficult to compare FGHs to professional standards because they are unprecedented. They do “not exist in the nomenclature.” (D.E. 303 at 89). What is clear, though, is that foster care systems have stricter regulations for facilities with larger capacities because “the more children you have, the more risk that you have.” Id. at 87. Yet, for reasons that escape the Court, Texas exempts FGHs from greater regulations despite recognizing them as “group homes,” and categorizing them with congregate care facilities. (D.E. 300 at 61; PX 50 at 7). Therefore, the Court will compare FGHs to foster family homes and congregate care facilities, when appropriate, to determine if FGH children are at an unreasonable risk of harm.

Plaintiffs first allege that DFPS causes an unreasonable risk of harm by placing unrelated children of different ages, service levels, and genders together in FGHs. (D.E. 300 at 62-63, 66). As of April 2014, 64% of FGHs with seven or more children mixed foster children 12 years old and younger with at least one foster teen. Thirty-six percent of these homes mixed foster children aged five to ten with at least one foster teen. (D.E. 326 at 22). Plaintiffs' expert on FGHs, Richter, testified that a “very high percentage of ... younger children were being placed with older children. And we know that creates more risk.” Id. at 22-23. The Court saw evidence of FGHs with age ranges from 1 to 17, 2 to 17, less than 1 to 16, 2 to 16, 5 to 16, and 5 to 17. (PX 598). DFPS does not prohibit teenagers from sharing a bedroom with younger children. (D.E. 315 at 10-11). In addition, “there's no prohibition, there's no limitations, that the State of Texas puts on foster group homes between children of different service levels.” (D.E. 300 at 66). In fact, 89% of FGHs with seven or more foster children had a mix of Basic, Moderate, Specialized, or Intense service levels. (D.E. 326 at 24). Richter testified that “89% is very, very high.” Id. Further, DFPS allows for young girls and teenage boys to be placed in the same foster group home. (D.E. 300 at 63).

Richter testified that the most significant risk to FGH children's safety “is the fact they're unrelated.” (D.E. 326 at 14). “There are no bonds, there are no ties; and, therefore, there's not that instinct to protect or not harm.” Id. at 17-19. Foster family homes account for this instinct by having a lower capacity, allowing for fewer unrelated children with greater monitoring. Congregate care facilities account for this instinct by having more regulations plus greater monitoring. FGHs do not account for this instinct at all.

The most egregious problem is that FGHs lack 24-hour awake-night supervision. (D.E. 300 at 61-62; D.E. 315 at 9-10). Richter described awake-night supervision as the “minimum support that should be given” because it “can prevent children from getting up from their beds and going into bathrooms, other bedrooms for the purpose of sexual contact or aggression or bullying or initiation into the foster group home.” (D.E. 326 at 27; see also D.E. 303 at 87). DFPS requires “all the other group homes” to have 24-hour awake-night supervision, but not FGHs. (D.E. 300 at 61). As a result, you can have 12 unrelated children of different ages, service levels, and genders in a foster group home “and have the caregivers sound asleep every night of the week.” Id. at 62. There is “[n]o doubt” in Dr. Miller's mind that this arrangement is unsafe. (D.E. 303 at 89).

Carpenter testified that, in foster group homes that mix younger children with older children, sexual abuse “is usual rather than unusual.” (D.E. 307 at 16-21; see also D.E. 324 at 202). The abuse often occurs at night when there is no supervision. Sharp explained that nighttime was “when things happened” in foster group homes. (D.E. 324 at 175). In one of his ten FGHs, Sharp said that a foster boy was “sexually abused almost every night by one of the bigger boys in the home” while the caretakers were asleep on the other side of the house. Id. at 172. Because DFPS does not regulate mixing ages in foster group homes, let alone in bedrooms, the abuser and the abused actually shared a bunkbed. Id. In fact, all the foster children in that group home shared one room with six bunk beds, while the foster parents and their two biological children (each of whom had a separate room) lived and slept on the other side of the house. Id. at 173. According to Sharp, it is “common” in foster group homes for the foster children to be segregated in this way from the biological family, often living in a separate building. Id. at 173-75. Similarly, when Jackson was 15, he shared an FGH bedroom with a 2-year-old, a 16-year-old, and a 19-year-old. Id. at 192-93. The 16-year-old ended up in jail for physically abusing the 2-year-old. Id. Jackson, who lived in around six foster group homes, called FGHs “unconscionable.” Id. at 192. Because DFPS refuses to track child-on-child abuse, older children with histories of abuse are placed in FGHs with younger children. This is precisely what happened to 8-year-old D.I. who was repeatedly raped by a 16-year-old in a foster group home at night. The 16-year-old, who had a history of sexual abuse, knew that the caregiver was asleep and had one of the other children serve “as a lookout ... when the children engaged in sexual activities” to make sure the foster father did not wake up. (PX 185 at 58 (filed under seal)). The Court reviewed scores of other tragic events at FGHs, including a 14-year-old boy who abused his foster parents' biological daughter by “inserting his finger in the vagina of [the] 4-year-old and making her bleed.” (PX 495 at 254 (filed under seal)). The foster mother “later reported that two other foster children—18-year-old [mentally handicapped] female, and 13-year-old [mentally handicapped] male also reported being touched by the 14 year old foster child.” Id. at 244.

Mixing children of varying service levels exacerbates these problems. Children at different service levels have different needs. Richter testified that “when you mix service levels, you see bullying ... when you place very aggressive or self-injurious kids in a home with [Basic-level] children, you now sacrifice a path for them ... you may actually damage those children to a point that's unconscionable.” (D.E. 326 at 17). In spite of these dangers, DFPS only requires two hours of extra transportation training for a caregiver to become certified to operate an FGH as opposed to a foster family home. (Id. at 64). Also, for some unexplained reason, DFPS requires fewer staff per child at FGHs than it does at foster family homes. In foster family homes, DFPS policy states, “A two-parent foster family home may care for up to six children,” a 1:3 ratio. (PX 67 at 219). In a foster group home, however, “The number of children one caregiver may supervise ... is eight,” a 1:8 ratio. Id. at 221. Plaintiffs' expert Dr. Miller testified, “It's very difficult to monitor that many kids, especially ... when you look at the mix of age groups and the mix of disabilities or children with disabilities—or service level needs ... that are in those homes.” (D.E. 303 at 87). Richter explained, there is “too much going on” at foster group homes given the number of caregivers and their training. (D.E. 326 at 16-17). As a result, there are often “medication errors,” “appointments missed, forgot the court hearing.” Id.

Virgil described FGHs as “Hectic.” Her first FGH contained eight children with an age range of 5 to 16. Whenever Virgil asked her caregiver for help, she was told to “go take a nap.” (D.E. 324 at 202). Because of insufficient staff, all the children were forced to travel together anytime the caregiver ran an errand. The problem, however, is that the caregiver's minivan only seated six. Virgil, the smallest, sat on the floor without a seatbelt. Id. at 203. Virgil testified that once when a foster child slit her wrists late on a school night, all the children had to spend the night in the hospital while the girl received treatment. Id. at 202. Sharp told the Court:

I don't think there were any good foster group homes. I don't think you can give children the time and attention they need [in a foster group home], particularly children who have experienced some type of trauma, which all children in foster care have. Whenever you have ten-plus kids in a home and you tend to ... only their very basic needs and you put them all in one room and you segregate them from everyone else, that's not a way to treat young people and that certainly isn't a way to help them address the problems that they're dealing with.

(D.E. 325 at 174). According to Sharp, child-on-child physical and sexual abuse is “a common thing in the bigger homes,” because caregivers cannot monitor all their children. Id. at 172.

Plaintiffs call foster group homes “A Recipe for Disaster,” and the Court agrees. (D.E. 345 at 61). It is obvious to the Court that harm would result from mixing unrelated male and female children of vastly different ages and service levels in the same home without appropriate supervision. Beyond the examples cited, the record is full of physical abuse, sexual abuse, suicide attempts, and poor supervision at foster group homes. (See, e.g. , PX 372; PX 377; PX 406; PX 457; PX 495; PX 502; PX 2041 at 31-36). As such, the Court finds that DFPS's policies and procedures for operating foster group homes amount to a structural deficiency that causes an unreasonable risk of harm to the FGH Subclass.

1. DFPS is Deliberately Indifferent Toward Foster Group Homes

DFPS is aware of its policies regarding foster group homes. Black admitted that FGHs are “group homes” but DFPS “treat[s] them just like foster family homes.” (D.E. 300 at 61). DFPS also knows that no “other state in the country ... has what is called a foster group home between seven children and 12 children.” Id. at 60-61. Although FGHs are supposed to contain no more than 12 children, DFPS admitted that it does not have centralized data on how many children, including adopted and biological, are in each FGH. (D.E. 315 at 6-7). DFPS is aware that it mixes ages, genders, and service levels in FGHs. (Id ; D.E. 300 at 62-63). The State admitted that, because it also does not track child-on-child abuse, older children with sexual abuse histories could be placed in the same FGH bedroom as younger children. (D.E. 300 at 67). Further, DFPS knows that “all the other group homes in the State ... require 24/7 supervision,” but FGHs do not. Id. at 61-62.

In addition to being “aware of facts from which the inference could be drawn that a substantial risk of serious harm exists,” the State has drawn that inference. Black admitted that she would not allow her own young children to sleep in the same room as older teens with the caregiver asleep, as it is “scary” and “inappropriate.” Id. at 64-65. DFPS said in a 2013 Self-Evaluation Report to the Sunset Commission that there are “risks associated with having seven to 12 children in one home” including “risks associated with providing appropriate overall supervision, ensuring individualized attention to each child's specific needs, and managing stress when caring for large numbers of children in a foster group home.” (PX 1864 at 235). Even without these admissions, the Court finds that “the risk of harm is obvious” when you place unrelated children of different ages, services levels, and genders in the same facility, without nighttime supervision or a sufficient number of trained staff.

DFPS has presented no evidence that it tried to correct these known harms. There has been no response to FGHs' shortcomings, let alone a reasonable one. DFPS “exhibited a conscious disregard” for known severe abuses, which “by itself sufficiently demonstrates deliberate indifference to a child's right to personal security.” Hernandez , 380 F.3d at 881. As such, the Court holds that DFPS is deliberately indifferent toward FGHs' structural deficiencies, which cause an unreasonable risk of harm to FGH children.

2. DFPS Substantially Departs from Professional Judgment Toward Foster Group Homes

DFPS's actions defy the professional judgment standard. It is impossible to determine which DFPS policies violate professional standards because there are no FGH professional standards, period. That said, a review of professional standards for all foster facilities convince the Court that DFPS's foster group home operation is a “substantial departure from accepted professional judgment, practice, or standards.” Youngberg , 457 U.S. at 323, 102 S.Ct. 2452.

All professional standards reviewed by the Court advocate 24-hour awake-night supervision in group homes. The CWLA recommends having staff “scheduled, on site, available and awake 24 hours per day when children are present.” (PX 2 at 2-5). The CWLA also says that, for group homes, there should be “24-hour access” to “medical services” and “psychiatric services,” meaning that “the residential service provider can contact a person or facility and obtain the needed service at any time, 24 hours per day.” Id. at 10. COA standards require continuous supervision by staff “available on a 24-hour basis,” and, “Personnel must be awake at all times unless convincing evidence demonstrates” that a particular home does not need nighttime supervision. (PX 32 at 61). The COA also requires that a “physician or other qualified medical practitioner assumes 24-hour on-call medical responsibility.” Id. at 37. Richter testified that for foster group homes, “the minimum support that should be given, in my opinion, is awake night supervision,” and on-call medical and psychiatric personnel are “very, very important.” (D.E. 326 at 27, 29). In Richter's opinion, on-call medical and mental health staff, as well as 24-hour awake-night supervision “are basic accepted things that you do.” Id. at 31. DFPS itself demands 24-hour awake-night supervision for “all other group homes.” Yet DFPS exempts FGHs from a nighttime supervision requirement. DFPS also does not require FGHs to have on-call medical and mental health support staff. Id. at 30-31.

It should be noted that DFPS's disregard of professional standards overburdens FGH caregivers, which in turn harms FGH children. (Id. at 16, 24; see also D.E. 303 at 87). The Court strongly agrees with the report and testimony of Richter, who explained, “if you have 7 to 12 children in a home, you have to really attend to the caregivers, support them, give them the kind of training and support they need, or bad things are going to happen. And when caregivers aren't taken care of, then the kids aren't.” (D.E. 326 at 13). DFPS offers no extra support for FGH caregivers. The only extra training DFPS provides is two hours for transportation training. Further, for some unknown reason, DFPS allows an 8:1 child to staff ratio in FGHs as opposed to the 3:1 ratio in foster family homes. FGHs “stretch caregivers beyond their ability,” and FGH children do not receive adequate supervision—day or night. Id. at 13, 16. In short, Richter believes that the foster group home model:

Has the risks of traditional group care (e.g., problems that come with mixing a high number of unrelated children in one home/setting, including child-on-child victimization, contagion of negative behaviors, police contacts), but none of the perceived benefits ... (e.g., educational and training requirements for caregivers, strict adherence to and [sic] personnel resources to retain proper supervision ratios and awake-night supervision); and Has the risks of typical family foster homes where caregivers receive only basic state agency-mandated training to meet the needs of children ... but has only severely compromised benefits of living in a family setting in the community, which promotes normalcy, as a result of the high number of unrelated children living together.

(PX 2041 at 10). Richter concludes that “foster group homes operated by [DFPS] deviate from professional standards and, as currently structured, expose all children placed in those homes to serious risk of substantial harm.” Id. at 6, 22. The Court agrees. The Court has seen no decision by DFPS—professional or otherwise—to address the litany of problems with FGHs. Although FGH professional standards do not exist, the Court finds that DFPS has violated a number of relevant professional standards as laid out by professional organizations, Plaintiffs' experts, and DFPS itself. The Court holds that DFPS's policies and practices toward foster group homes “are such a substantial departure from accepted professional judgment, practice, or standards” that DFPS did not base its decision on any professional judgment at all.

V. REMEDY

Plaintiffs have proven, by a preponderance of the evidence, that DFPS's policies and practices amount to structural deficiencies that cause an unreasonable risk of harm to all class and subclass members. In order to obtain injunctive relief, Plaintiffs must also show: “(1) that [they] suffered an irreparable injury; (2) that remedies available at law, such as monetary damages, are inadequate to compensate for that injury; (3) that, considering the balance of hardships between the plaintiff and defendant, a remedy in equity is warranted; and (4) that the public interest would not be disserved by a permanent injunction.” See, e.g. , eBay Inc. v. MercExchange, L.L.C. , 547 U.S. 388, 391, 126 S.Ct. 1837, 164 L.Ed.2d 641 (2006). “The decision to grant or deny permanent injunctive relief is an act of equitable discretion by the district court, reviewable on appeal for abuse of discretion.” Id. Equitable relief is “limited by considerations of federalism, and remedies that intrude unnecessarily on a state's governance of its own affairs should be avoided.” Ass'n of Surrogates & Supreme Ct. Reporters Within City of N.Y. v. State of N.Y. , 966 F.2d 75, 79 (2d Cir.1992). “Courts nevertheless must not shrink from their obligation to enforce the constitutional rights of all persons ... simply because a remedy would involve intrusion into the realm of [government] administration.” Brown v. Plata , 563 U.S. 493, 131 S.Ct. 1910, 1928, 179 L.Ed.2d 969 (2011) (citation omitted). The Court finds that injunctive relief is appropriate in this case.

Plaintiffs may satisfy the first two factors by proving that they are “likely to be deprived of their constitutional rights in the future by the acts they seek to have enjoined.” Floyd v. City of N.Y. , 959 F.Supp.2d 668, 672 (S.D.N.Y.2013) ; see also Springtree Apartments, ALPIC v. Livingston Par. Council , 207 F.Supp.2d 507, 515 (M.D.La.2001) (“It has been repeatedly recognized by the federal courts that violation of constitutional rights constitutes irreparable injury as a matter of law.”) (citing Elrod v. Burns , 427 U.S. 347, 373-74, 96 S.Ct. 2673, 49 L.Ed.2d 547 (1976) (plurality opinion) (“The loss of First Amendment freedoms, for even minimal periods of time, unquestionably constitutes irreparable injury.”)). Texas has violated Plaintiffs' Fourteenth Amendment right to be free from an unreasonable risk of harm. Remedies besides injunctive relief—which Plaintiffs have not requested—could not compensate for the repeated exposure to physical abuse, sexual abuse, and psychological abuse. Until Texas's structural deficiencies are cured, Plaintiffs will be harmed.

The balance of hardships also tilts strongly in favor of granting injunctive relief. The burden on the Plaintiffs from continued unconstitutional harm far outweighs any administrative hardship Texas will face in correcting its foster care system. The Court is sensitive to the State's interests in controlling its core functions amidst administrative constraints. Having chosen to operate a foster care system, however, Texas is obligated to comply with the dictates of the Fourteenth Amendment. It is worth noting that funding is not a problem for DFPS. The Texas Legislature has been responsive to DFPS's needs. According to Specia, “when the State had money, [DFPS] got substantial amounts of money. When the State didn't have money, [DFPS] had fewer cuts than other people did.” (D.E. 300 at 26; see also D.E. 331 at 32-37). Burstain also testified that “the legislature has been very supportive and aligned with CPS,” allocating significant resources to support CPs's requests. (D.E. 330 at 176). Budgetary constraints, which do not absolve constitutional violations, nonetheless do not exist here.

Lastly, the public interest would not be disserved by injunctive relief. To the contrary, as the system currently stands, foster children often age out of care more damaged than when they entered. (PX 1964 at 2) (“there is increasing evidence to show that our foster care system is sometimes doing more harm to our children than good”). Years of abuse, neglect, and shuttling between inappropriate placements across the State has created a population that cannot contribute to society, and proves a continued strain on the government through welfare, incarceration, or otherwise. Although some foster children are able to overcome these obstacles, they should not have to. The public interest will not be harmed by an injunction requiring Texas to conform its foster care system to the Constitution. With all four factors met, the Court holds that injunctive relief is appropriate in this case.

“[T]he scope of a district court's equitable powers to remedy past wrongs is broad, for breadth and flexibility are inherent in equitable remedies.” Swann v. Charlotte – Mecklenburg Bd. of Ed. , 402 U.S. 1, 15, 91 S.Ct. 1267, 28 L.Ed.2d 554 (1971). That said, “judicial powers may be exercised only on the basis of a constitutional violation .... As with any equity case, the nature of the violation determines the scope of the remedy.” Id. at 16, 91 S.Ct. 1267. In fashioning a remedy, the Court notes the decades of reports aimed at fixing DFPS, and the corresponding lack of meaningful attempts at improvement.

A. Injunction

The State shall establish and implement policies and procedures to ensure that Texas's PMC foster children are free from an unreasonable risk of harm. To effect this injunction, the Court will appoint a Special Master to help the State implement the Goals outlined below. Further, the State shall immediately stop placing PMC foster children in unsafe placements, which include foster group homes that lack 24-hour awake-night supervision. Foster group homes that immediately require 24-hour awake-night supervision may continue to operate while the Special Master and the State craft and enforce the Implementation Plan.

B. Appointment of a Special Master

Because of the complexity and breadth of reforms that are required to bring DFPS in compliance with the Constitution, it would be impractical for the Court to craft and oversee each necessary change. As a more effective and flexible alternative, the Court will appoint an independent Special Master to help craft the reforms and oversee their implementation. See Floyd , 959 F.Supp.2d at 676. The State will fund the Special Master, though the Court believes that by facilitating early and unbiased detection of non-compliance and barriers to compliance, the Special Master will ultimately save the State time and resources. The Special Master will operate in close coordination with this Court, which retains jurisdiction to issue orders as necessary to remedy the constitutional violations described in this Opinion.

To select a Special Master:

(1) Within 30 days of this Opinion, the Court will hold a hearing to select a Special Master. At the hearing, the parties will present the Court with an agreed-on Special Master who is an expert in a field relevant to foster care systems, along with that person's curriculum vitae (“CV”). The agreed-on Special Master must attend this hearing. If the parties cannot agree on a Special Master, each

party will present the court with two candidates and their CVs. The Court will choose one Special Master out of the four options. The four candidates must attend this hearing if the parties cannot agree on a Special Master.

The functions of the Special Master:

(1) Within 180 days of selecting a Special Master, the Court will hold another hearing where the Special Master will present the Court with an Implementation Plan to reform Texas's foster care system. In crafting the Implementation Plan, the Special Master shall have access to all relevant information, records, personnel, and reports, and shall have the authority to require reports that would be helpful. The Special Master shall assure that the Plaintiffs and Defendants have a full opportunity to consult and be involved with the development of the Implementation Plan, though the Special Master must ultimately be guided by the Goals listed below, which are aimed at addressing the State's constitutional violations. The Special Master shall be mindful that Texas does not need to provide a perfect foster care system; just one that no longer violates the Constitution. The provisions in the proposed Implementation Plan will be viewed as recommendations. The Court will hear the parties' objections to the proposed Implementation Plan at this hearing, and decide which, if any, provisions to accept. If the Court finds any parts lacking or any constitutional violation unaddressed, the Court will ask the Special Master to present additional provisions at a later date.

(2) The Implementation Plan shall include:

(a) Implementation Dates. The Plan shall contain final implementation dates by which each provision will be fully implemented. The plan shall also include interim dates, prior to the final implementation date, by which each provision shall be partially implemented, and shall specify the degree of implementation required by each of the interim implementation dates for each provision. Provisions should be implemented at the earliest time practicable. Any need to prioritize implementation will be motivated by which provisions best address the State's constitutional violations.

(b) Implementation Steps, Processes, and Tasks. The Plan shall specify the steps and tasks necessary to achieve full implementation and to meet the interim and final implementation dates. The Plan shall designate timelines by which the steps and tasks are to be accomplished and shall designate appropriate staffing levels and personnel to assure timely implementation of the Plan. The Plan shall also identify who shall be responsible for the steps and tasks, including outside contractors. This portion of the Implementation Plan shall be updated every six months, and the updates shall list tasks accomplished and additional steps and tasks that may have been identified as additionally necessary to achieve full implementation. The updates shall include timelines for interim and full implementation, as well as the designation of personnel for these additional steps and tasks.

(3) Upon accepting the Implementation Plan, the Court will enter a Final

Order incorporating the provisions of the Implementation Plan that is binding on the parties. The Final Order shall be enforceable by the Court.

(4) After the Court enters the Final Order, the Special Master shall issue a report every 180 days concerning the status and progress made by the Defendants toward compliance with the Final Order. The Special Master's reports shall be provided to the parties and the Court.

(5) The Special Master and the parties shall not disclose any individually identifiable information concerning children, their parents, or their foster or adoptive parents. All reports concerning compliance with the Implementation Plan, the status of children, and compliance with the Final Order that do not contain individually identifiable information shall be public information.

(6) The Special Master shall be responsible solely to the Court, but shall work actively with the parties to assure the effective and prompt implementation of the Court's Final Order.

(7) The Special Master shall serve until the Court determines, upon Defendants' application, that the Special Master is no longer necessary.

C. Goals

In General:

(1) When DFPS staff visit or call a foster child, the caregiver must allow the staff member and the child to speak privately, unless the staff member agrees that it would be unsafe. If the meeting was not conducted in private, the staff must explain why in the child's case files.

(2) DFPS paperwork and electronic filing system, including IMPACT, CLASS, and the External files, must become more efficient. Each child should have a readily accessible and organized case file, comprised of all records pertaining to that child. The Court was routinely frustrated at the disorganization, duplication, and inconsistency in the foster children's case files. Caseworkers should be able to spend more than 26% of their time with foster children.

(3) DFPS shall include an updated portrait photograph of each child in their respective case files. The Special Master will recommend how frequently the photograph must be updated. Each photograph shall include the date it was taken, and be organized and easily retrievable in the case files.⁵⁴

(4) DFPS shall establish, staff, and maintain a 24-hour hotline for receiving and responding to reports of abuse and neglect. The hotline's phone number must be readily available and displayed prominently in all foster care residential facilities. Foster children must be allowed telephone access to reach out to this 24-hour system, free from observation.

(5) DFPS shall improve its programs and outreach for children who will age out of foster care so that more children take advantage of these programs.

54 The frequency of runaways while in DFPS care necessitates the availability of a recent photograph. Although DFPS policy already requires an annual photograph, (See PX 52 at 91), the fact that no one at DFPS knew what M.D. looked like is troubling. See supra pp. 68-69. Obviously, the age of the child will dictate the frequency with which the photograph is updated. For example, an annual photograph for an infant might not be helpful.

(6) All PMC children shall be entitled to an attorney ad litem and a CASA volunteer, as well as any other representative appointed to TMC children that the Special Master determines is necessary for PMC children's safety and well-being.

(7) The Special Master shall recommend any provision beyond the Court's Goals that are deemed necessary to cure the State's constitutional violations outlined in this Opinion.

CVS Caseworkers:

(1) DFPS must track primary CVS caseworker caseloads on a child-only basis. The Special Master shall recommend whether tracking should be categorized separately for full-time and part-time primary CVS caseworkers, and how tracking should be categorized on a region and county-level. The State cannot include in the calculations secondary workers, workers in training, or fictive workers created out of overtime. The State is welcome to continue tracking caseloads by stages, but not in lieu of child-only tracking.

(2) DFPS must complete a Workload Study to determine the time required for caseworkers to adequately perform their tasks. DFPS will specify how long it takes to complete each task. The Special Master shall recommend how frequently DFPS must complete additional workload studies.

(3) The Special Master shall recommend the point at which caseloads are manageable for full-time and part-time CVS caseworkers, taking into account times of crises. What is manageable is to be understood as the level at which caseworkers are able to perform their basic functions and not compelled to quit at an unreasonable rate. In other words, a manageable caseload is the level at which children are free from an unreasonable risk of harm.

(4) DFPS must hire and maintain enough primary CVS caseworkers to ensure that caseloads are manageable in each county in the State.

(5) DFPS must significantly lower its primary CVS caseworker turnover rate.

(6) The Special Master shall evaluate and recommend whether secondary CPU and I See You workers should be maintained. If so, the Special Master shall recommend provisions to make them more effective at protecting foster children from an unreasonable risk of harm.

(7) The Special Master shall recommend other provisions deemed necessary to ensure that primary CVS caseworkers are able to protect foster children from an unreasonable risk of harm.

CCL Investigations, Inspections, and Licensing:

(1) DFPS must complete a Workload Study to determine the time required for investigators and inspectors to adequately perform their tasks. DFPS will specify how long it takes to complete each task. The Special Master shall recommend how frequently DFPS must complete additional workload studies.

(2) The Special Master shall recommend the point at which caseloads are manageable for investigators and inspectors. What is manageable is to be understood as the level at which investigators and inspectors are able to perform their basic functions. In other words, a manageable caseload is the level at which children are free from an unreasonable risk of harm.

(3) PMU must conduct case readings to assess RCCL investigations in a manner and at a frequency deemed appropriate by the Special Master. The Special Master will recommend the appropriate method to correct dispositions and order corrective actions when PMU identifies deficiencies.

(4) The Special Master shall recommend provisions to solve RCCL's unwillingness to institute corrective actions against violating facilities.

(5) DFPS shall track child-on-child abuse, and categorize it as such. The Special Master shall recommend the most appropriate fashion to track child-on-child abuse bearing in mind that the information should be easy to retrieve and should be used to inform all placements and treatments. The Special Master shall also recommend how to categorize the initiators of child-on-child abuse, sensitive to the consequences of labeling children as “perpetrators.” The Special Master shall also recommend if child-on-child abuse should simultaneously be categorized as neglectful supervision by the caregiver.

(6) The Special Master shall recommend other provisions deemed necessary to ensure that RCCL protects foster children from an unreasonable risk of harm.

Inadequate Placement Array:

(1) DFPS shall not allow unrelated children that are more than one service level apart (e.g., Moderate and Intense-level children) to be placed in the same room in any residential facility. The Special Master shall recommend if it is appropriate to allow unrelated children that are only one service level apart (e.g., Basic and Moderate-level children) to be placed in the same room in any residential facility.

(2) The Special Master shall recommend what age ranges of unrelated children are appropriate to be placed in the same room in any residential facility. The Court understands that larger age gaps may be more appropriate for younger children (e.g. a five-year-old and a ten-year-old in the same room) than for older children (e.g. a ten-year-old and a 15-year old in the same room).

(3) DFPS shall track how many placements in its array are designated as single-child homes (including biological and adopted children), and track how many foster children need single-child homes. DFPS shall explain its criteria for determining which children need single-child homes. DFPS shall ensure that all children who need single-child homes are placed in such homes, unless it is in the child's best interest to remain with siblings or be supervised at a congregate care facility. If a child who needs a single-child home is not placed in such a home, the child's primary CVS caseworker must explain why in the child's case files.

(4) DFPS shall conduct a formal statewide needs assessment to determine an adequate placement array, including by number, geographic distribution, and placement type. DFPS must procure a placement array that substantially matches this assessment, and place children appropriately. The Special Master shall recommend how frequently DFPS must complete additional statewide needs assessments. DFPS shall take whatever steps are necessary to ensure that it has available to it at all times an adequate placement array, including

by number, geographic distribution, and placement type, and that foster children are placed promptly in an appropriate facility.

(5) DFPS shall track how many children are in each residential facility, including biological and adopted children, as well as each facility's licensed capacity. DFPS shall make this information easy to retrieve.

(6) DFPS can continue to pursue Foster Care Redesign, but only if the Special Master recommends, and the Court agrees, that Redesign meets the statewide needs assessment. The Special Master can evaluate Foster Care Redesign.

(7) The Special Master shall recommend provisions to solve the problem of children being removed from placements where they are succeeding because their level of care has altered. The Special Master shall also recommend provisions to solve the perverse incentive of DFPS providing additional funds to caregivers for children at increased levels of care.

(8) The Special Master shall recommend other provisions deemed necessary to ensure that DFPS's placement array no longer causes an unreasonable risk of harm to foster children.

Foster Group Homes:

(1) The Special Master shall recommend if FGHs should continue to operate based on whether FHGs can be improved to the extent that they will not cause an unreasonable risk of harm to foster children.

(2) If the Special Master determines that FGHs should continue to operate, the Special Master shall recommend necessary provisions, including awake-night supervision, additional staff training, and more appropriate staff-to-child ratios.

VI. CONCLUSION

Texas's foster care system is broken, and it has been that way for decades. It is broken for all stakeholders, including DFPS employees who are tasked with impossible workloads. Most importantly, though, it is broken for Texas's PMC children, who almost uniformly leave State custody more damaged than when they entered.

The Court believes that Specia and his staff have the best intentions to run an effective foster care system. The Court is also mindful that Specia is the seventh commissioner since 2004, each of whom was surely ushered in with promises that this time it will be different. The reality is that DFPS has ignored 20 years' of reports, outlining problems and recommending solutions. DFPS has also ignored professional standards. All the while, Texas' s PMC children have been shuttled throughout a system where rape, abuse, psychotropic medication, and instability are the norm. The Court has no assurance that anything has changed. Of the two reforms to which DFPS pointed—CPS Transformation and Foster Care Redesign—one was instituted in the wake of this lawsuit, the other is an abject failure, and neither answers any of Plaintiffs' claims.

Plaintiffs have a Fourteenth Amendment substantive due process right to be free from an unreasonable risk of harm caused by the State. Texas currently violates that right.

SIGNED and ORDERED.

VII. GLOSSARY

CCL —Child Care Licensing. A division of DFPS.

COA —Council on Accreditation for Services to Children and Families. A nonprofit child advocacy organization that recommends professional standards for state child welfare systems. CPA —Child Placement Agency. A private agency contracted by DFPS to place foster children in homes.

CPS —Child Protective Services. A division of DFPS.

CPU —Centralized Placement Unit. A secondary caseworker who does not interact with children.

CVS —Conservatorship, i.e., foster care.

CWLA —Child Welfare League of America. The nation's oldest and largest child welfare organization, which recommends professional standards for state child welfare systems.

DFPS —Department of Family Protective Services. A Defendant, and the Texas State agency responsible for protecting the State's children, elderly, and disabled.

FGH —Foster Group Home. A childcare facility that provides care for 7 to 12 children for 24 hours a day.

GRO —General Residential Operation. A child-care facility that provides care for more than 12 children for 24 hours a day. GROs include RTCs, halfway houses, emergency shelters, and therapeutic camps, and may be a single building or a campus with multiple cottages.

IMPACT —Information Management for the Protection of Adults and Children in Texas. An automated system, included in case files, in which DFPS staff record casework related activities.

LFC —Licensed Foster Care. Refers to foster care operations that receive licensing and oversight from DFPS. LFCs include independent foster family homes, independent foster group homes, GROs, RTCs, maternity homes, child-placing agencies, CPA foster family homes, and CPA foster group homes.

PMC —Permanent Managing Conservatorship. A type of legal custody granted by the courts to DFPS. The legal status for children typically progresses to PMC from TMC, 12-18 months after the child enters foster care.

PMU —Performance Management Unit. A division of CCL that performs internal quality control.

RCCL —Residential Child Care Licensing. A division of CCL that regulates, licenses, and investigates residential foster care operations.

R/O —Ruled Out. An investigation disposition, meaning that a preponderance of evidence indicates that abuse, neglect or exploitation did not occur.

RTB —Reason to Believe. An investigation disposition, meaning that a preponderance of evidence indicates that abuse, neglect, or exploitation occurred.

RTC —Residential Treatment Center. A type of GRO for children with more serious physical and mental health needs.

TMC —Temporary Managing Conservatorship, a type of legal custody granted by the courts to DFPS. A child may remain in the State's TMC for 12 months, although a court can order a 6 month extension.

UTD —Unable to Determine. An investigation disposition, meaning that a determination could not be made because of an inability to gather enough facts. The investigator concludes that there is not a preponderance of evidence that abuse or neglect occurred; but it is not reasonable to conclude that abuse or neglect did not occur.

ATTACHMENT 1

(PX 811)

DFPS Hierarchy

ATTACHMENT 2

(PX 806)

Map of Regions