Philip Wharton v. Carl Danberg

PRECEDENTIAL

UNITED STATES COURT OF APPEALS

FOR THE THIRD CIRCUIT

_____________

No. 16-1988

_____________

PHILIP A. WHARTON; JOSEPH ROUNDTREE;

JAMES MADDOX; LAMAR CORREA,

Appellants

v.

CARL C. DANBERG; CATHY ESCHERICH;

REBECCA MCBRIDE; ROBERT COUPE

______________

APPEAL FROM THE UNITED STATES DISTRICT

COURT

FOR THE DISTRICT OF DELAWARE

(D.C. Civ. Action No. 1-12-cv-01240)

District Judge: Honorable Leonard P. Stark

______________

Argued November 3, 2016

______________

Before: JORDAN, GREENAWAY, JR., and RENDELL,

Circuit Judges.

(Opinion Filed: April 19, 2017)

Stephen A. Hampton, Esq. [ARGUED]

Grady & Hampton

6 North Bradford Street

Dover, DE 19904

Counsel for Appellants

Michael F. McTaggart, Esq. [ARGUED]

Delaware Department of Justice

820 North French Street

Carvel Office Building, 6th Floor

Wilmington, DE 19801

Counsel for Appellees

Richard H. Morse, Esq.

American Civil Liberties Union

100 West 10th Street, Suite 706

Wilmington, DE 19801

Counsel for Amicus Curiae

______________

OPINION

______________

GREENAWAY, JR., Circuit Judge.

This putative class action alleges that the Delaware

correctional system routinely fails to release inmates in a

timely manner, holding them for days or weeks beyond when

they should be set free. Appellants, a group of inmates who

were over-detained, have sued top correctional officials—

2

specifically, former Delaware Department of Corrections

(“DDOC”) Commissioner Carl Danberg, current DDOC

Commissioner Robert Coupe,1 and Rebecca McBride, the

current Director of the DDOC Central Offender Records

division (“COR”)—seeking both damages and structural

reform of COR. The District Court granted summary

judgment in favor of Danberg, McBride, and Coupe

(Appellees). We will affirm.

I. FACTS

In 2008, the Delaware correctional system was facing

scandal for its handling of inmate releases. One inmate,

Jermaine Lamar Wilson, committed suicide in his cell on the

day he was supposed to be—but was not—released.2 Dozens

of other inmates had either been released too early or too late.

National experts, cited in contemporaneous press reports,

expressed surprise about how many Delaware inmates were

improperly released. As gubernatorial candidates from both

parties attacked the state correctional system, there was high-

level support for reform.

That reform took shape in the establishment of a new

Central Offender Records office within the Delaware

Department of Corrections. Previously, staff at each prison

handled releases individually. COR was meant to centralize,

1

The District Court allowed the substitution of Coupe for

Danberg as a defendant for purposes of prospective relief

only.

2

Wilson v. Taylor, 597 F. Supp. 2d 451, 457-58 (D. Del.

2009).

3

standardize, and generally improve the state’s processing of

inmate releases. The creation of COR, led by then-DDOC

Commissioner Carl Danberg, was a substantial bureaucratic

undertaking, requiring the department to coordinate with

legislators, the judiciary, and its unionized employees.

This litigation, however, contests whether

centralization has brought improvement. Appellants allege

that Delaware’s problems with over-detentions have, if

anything, gotten worse since 2008.

Under the new system, after an inmate is ordered to be

released (because he or she posted bail, because their bail was

changed from secured to unsecured, or because they

completed their sentence, to offer a few examples), the court

is supposed to fax an order to COR. COR then checks

whether there is a reason to continue holding the individual—

for example, an outstanding warrant—and if there is not,

sends instructions to the facility where they are being held for

that individual’s release.

Undoubtedly, there were bumps along the way to a

centralized system. In 2008, Danberg himself admitted that

the creation of COR had caused confusion during the

transition itself. Led by Appellees, DDOC has attempted to

improve COR’s functioning since its inception. COR has an

official goal of processing all releases within 24 hours.

Observing delays in the processing of releases, Appellees

have increased staffing levels. They created a new six-month

orientation period for new hires at COR. In 2010, COR

adopted a new computer system, called the Delaware

Automated Correction System (“DACS”) which is meant to

foster better tracking of release dates and the “triage” of

4

records to prioritize releases.3 JA 15. Finally, COR created a

new priority unit for releases likely to be fast and easy, such

as those whose bail is changed to unsecured or those whose

charges are dismissed. According to McBride, she and her

co-defendants are “always looking at ways to be more

efficient.” JA 12.

Even with these interventions, however, Appellants

suggest that COR is badly broken, causing or allowing the

over-detention of as many as thousands of inmates a year.4

To support their theory of over-detention, Appellants

submitted a disparate and somewhat disjointed assortment of

affidavits from several witnesses whose work brings them in

close contact with the correctional system. These affidavits,

described below, reported huge numbers of over-detentions,

albeit in an impressionistic fashion based on the affiants’ own

personal observations and estimates.

First, a former records clerk at COR named Brenda

5

Bell estimated that 10 to 20 percent of release orders

3

One COR employee, however, averred that the new

computer system “caused more delays.” JA 163.

4

The parties dispute in the briefing exactly how “over-

detention” should be defined and in particular whether over-

detentions of a certain length of time should qualify. This

issue is not material to this appeal.

5

Bell worked at COR for roughly one year between 2011 and

2012. She worked as a Records Clerk and Records Specialist

during her employment, positions that involved work on

release orders.

5

received by COR were not processed and sent to a

correctional facility within 24 hours, and that 20 to 30

inmates per week ended up spending more than two days

waiting for COR to send their release order to their facility.

Second, a bail bondsperson, Bruny Mercado,6

calculated that about 35 percent of people for whom she had

posted bond were held for more than 24 hours after bond was

posted, and 25 percent of people were held for more than 48

hours afterward. Mercado also said that she had seen no

improvement at COR in its four years of existence.

Third, a Delaware public defender, Sandra Dean,7

averred that over-detention was a “consistent problem” for

her clients. JA 179. She reported that she had her secretary

call COR every day to inquire about clients whose release had

been ordered by the courts and that she followed up

personally with COR for clients who were not released after

three days. Notably, although Dean only served as a public

defender until 2010, she claimed that the over-detention

problem worsened at the end of that period.

6

Mercado has owned and operated her own bail bond

company since 2002 and posts bail for approximately 25 to

30 people in Delaware per month. She and her employees

observe the release process after they post bond for their

clients and communicate with COR during that process.

7

Dean worked as an attorney for the Delaware Office of the

Public Defender from 1991 to 2010. When her clients were

over-detained, she worked to secure their release.

6

Given that COR processes between 16,000 and 18,000

releases per year, these affidavits allege as many as 6,300

over-detentions a year. The affidavits also allege that COR

was informed of the problem, both by Dean’s regular

inquiries and by Mercado, who stated that she had spoken

personally with McBride about the over-detention problems.

On the other hand, hard, reliable data about the number

of over-detentions occurring each year is more or less missing

from the record. Appellants’ affidavits put forth various

estimates of the over-detention problem, but no precise

quantification or authoritative analysis. They offer a limited

ability to understand how the problem has changed over time.

In contrast, Appellees do not even attempt to provide a

systematic accounting of over-detentions from their own

archives. Rather, they base their count of over-detentions on

Appellants’ ability to identify specific over-detained inmates.

The record does include various tables purporting to

show the number of over-detentions each month, which

totaled to two each year of the relevant period except for FY

’10, when there were 18 over-detentions. But neither party

treats those tables as reliable. Given the absence of

information about the source of that data, we likewise decline

to treat these tables as reliable.8

8

At oral argument, Appellants claimed that those tables count

only the over-detentions specifically brought to McBride’s

attention. This is not evident from the record, but if it were

so, that would not provide any reliable metric for the actual

number of over-detentions.

7

According to Appellants, two specific problems at

COR, in addition to the general failure of the centralization

effort, account for the continuing trouble with over-

detentions. First, they claim that COR is under-staffed

generally and particularly short-handed on nights and

weekends. COR employees and former employees testified

that the division would be able to process releases more

quickly and avoid over-detentions if it had more staff or was

open more hours. Coverage is worse on weekends: although

fewer releases arrive at COR on weekends, staffing levels are

more-than-proportionally thinner. COR also typically closes

its offices between 10:00 p.m. and 5:00 a.m. and over

holidays, which Appellants allege leads to worse delays

during those periods (Delaware’s Justice of the Peace Courts,

which send releases to COR, remain open 24/7). During

those periods, however, COR supervisors are on call around

the clock to handle any problems that might arise. Appellee

McBride testified that she has never received any complaints

about coverage while COR offices were closed, and there is

no record evidence that over-detentions cluster around

holidays or are otherwise affected by these closures.

Second, Appellants argue that COR is unresponsive to

inmates and those acting on their behalf. Prisoners cannot

contact COR directly, except by mail. If they want more

immediate communication with COR—and time is of the

essence for an inmate detained past his release date—they

must request that prison staff email COR. Prisoners’ families,

friends, and bail bondsmen can call COR directly, but

generally complain that COR is frequently unhelpful or

indifferent, when it can be reached at all. Outside input

allegedly falls on deaf ears. But inmates and their associates

are the individuals best placed to know that they have been

8

over-detained—they have access to information and strong

incentives to monitor COR—and could play an important

oversight function if allowed.

Indeed, the record shows that when inquiries did reach

COR, they helped spur COR to fix problems. Sandra Dean,

the public defender, noted her practice of routinely contacting

COR about over-detained clients. She stated that additional

pressure, such as threatening to contact a judge or initiate a

contempt proceeding, helped more. Conversely, Dean

observed that those least able to speak for themselves when

over-detained, like the handicapped or those who did not

speak English, were at greater risk. Likewise, after

Appellants complained of being over-detained, prison

officials urged them to contact COR because that was the best

way to speed their release. COR’s alleged failure to

communicate potentially deprives COR of an alert system and

allows errors to fester. That said, McBride testified that

family members, courts, and correctional officers are able to

reach her directly with complaints about over-detention and

that she responds to those complaints with an immediate

investigation into the inmate’s situation.

II. PROCEDURAL BACKGROUND

Plaintiffs filed their complaint on October 1, 2012. At

the close of discovery, Plaintiffs moved for class certification

and Defendants for summary judgment.

The District Court’s opinion denied the motion for

class certification and granted summary judgment.

Defendants prevailed in toto. Class certification was denied

on commonality grounds because some members of the

proposed class were over-detained due to delays in the court

9

system, rather than delays at COR. The Court found that

there was no “common contention” the truth of which could

“resolve an issue that is central to the validity of each one of

the claims in one stroke.” Wal-Mart Stores, Inc. v. Dukes,

564 U.S. 338, 350 (2011). The Court also found that all

claims against Defendants in their official capacities were

barred by sovereign immunity; it declined to reach the

question of qualified immunity; and it granted summary

judgment on Appellants’ state law claims.

The Court framed its analysis of the core federal

constitutional claims by using the “more-specific-provision

rule.” Specifically, the Court determined that the rule meant

that any substantive due process claims should be addressed

only under the more specific Eighth Amendment analysis of

cruel and unusual punishment. See Betts v. New Castle Youth

Dev. Ctr., 621 F.3d 249, 261 (3d Cir. 2010). In that Eighth

Amendment analysis, the District Court held that Plaintiffs

failed to demonstrate a genuine dispute of material fact on

two required elements of their claim—deliberate indifference

by Defendants to the risk of over-detentions and a causal

connection between Defendants’ acts and Plaintiffs’ over-

detentions.

In this respect, the Court found particularly

determinative: 1) press coverage praising defendant Danberg

for his creation of COR and his efforts to fix the over-

detention problem, 2) defendant McBride’s familiarity with

COR procedures and her work to improve them, and 3)

defendant Coupe’s formation of a special unit to speed up

daily bail releases.

Plaintiffs filed a Rule 59(e) motion to amend the

judgment, which was denied because Plaintiffs simply

10

rehashed the arguments posed on summary judgment.

Plaintiffs then appealed all of the federal claims asserted.

III. STANDARD OF REVIEW9

On appeal from a grant of summary judgment, the

Court of Appeals’ review is “plenary” and the court should

“apply the same test the district court should have utilized

initially.” Giles v. Kearney, 571 F.3d 318, 322 (3d Cir. 2009)

(citation omitted). Summary judgment should be granted

only when the record shows that “there is no genuine dispute

as to any material fact and that the moving party is entitled to

judgment as a matter of law.” Fed. R. Civ. P. 56(a). “[A]ll

justifiable inferences are to be drawn in [the nonmovant’s]

favor” but the “mere existence of some evidence in support of

the nonmovant is insufficient to deny a motion for summary

judgment; enough evidence must exist to enable a jury to

reasonably find for the nonmovant on the issue.” Giles, 571

F.3d at 322 (citing Anderson v. Liberty Lobby, Inc., 477 U.S.

242, 249, 255 (1986)).

In deciding whether to certify a class under Federal

Rule of Civil Procedure 23, a district court must make

“findings” and factual determinations. In re Hydrogen

Peroxide Antitrust Litig., 552 F.3d 305, 320 (3d Cir. 2008).

The burden of proof rests with the movant to “affirmatively

9

The District Court had jurisdiction over this civil rights

action pursuant to 28 U.S.C. §§ 1331 and 1343 and exercised

supplemental jurisdiction over state law claims pursuant to 28

U.S.C. § 1367. The District Court’s summary judgment

constituted a final decision and this Court has jurisdiction

pursuant to 28 U.S.C. § 1291.

11

demonstrate” certifiability by a preponderance of the

evidence. Wal-Mart, 564 U.S. at 350. The District Court’s

denial of class certification is reviewed for abuse of

discretion. Newton v. Merrill Lynch, Pierce, Fenner & Smith,

Inc., 259 F.3d 154, 165 (3d Cir. 2001).

IV. ANALYSIS

A. Eighth Amendment Legal Standards

Our standard for analyzing over-detention claims is

well-established. An inmate’s detention after his term of

imprisonment can, under certain circumstances, constitute

cruel and unusual punishment, in violation of the Eighth

Amendment. Montanez v. Thompson, 603 F.3d 243, 250 (3d

Cir. 2010). Continued incarceration beyond that point is

clearly punitive, and in many cases will serve no penological

justification at all. Sample v. Diecks, 885 F.2d 1099, 1108

(3d Cir. 1989). That said, we also recognize that “[t]he

administration of a system of punishment entails an

unavoidable risk of error” and that “[e]limination of the risk

of error in many instances would be either literally impossible

or unfeasible because prohibitively costly.” Id. The Eighth

Amendment does not, and could not, require the elimination

of all such risk of error.

Thus, we have established a three-part test for over-

detention claims. A plaintiff must show:

(1) a prison official had knowledge of the prisoner’s

problem and thus of the risk that unwarranted

punishment was being, or would be, inflicted; (2) the

official either failed to act or took only ineffectual

action under the circumstances, indicating that his

12

response to the problem was a product of deliberate

indifference to the prisoner’s plight; and (3) a causal

connection between the official’s response to the

problem and the unjustified detention.”

Montanez, 603 F.3d at 252.10

10

Our precedent also describes the test for supervisor liability

under the Eighth Amendment as a four-part test: “the plaintiff

must identify a specific policy or practice that the supervisor

failed to employ and show that: (1) the existing policy or

practice created an unreasonable risk of the Eighth

Amendment injury; (2) the supervisor was aware that the

unreasonable risk was created; (3) the supervisor was

indifferent to that risk; and (4) the injury resulted from the

policy or practice.” Beers-Capitol v. Whetzel, 256 F.3d 120,

134 (3d Cir. 2001). But see Barkes v. First Corr. Med., Inc.,

766 F.3d 307, 341 (3d Cir. 2014) (Hardiman, J., dissenting),

rev'd on other grounds sub nom. Taylor v. Barkes, 135 S. Ct.

2042 (2015) (questioning validity of four-part test). For

purposes of this litigation, the two formulations of the Eighth

Amendment standard are functionally equivalent, each

broadly requiring risk, knowledge, deliberate indifference and

causation. We find the over-detention-specific description of

our standard better structures our analysis in this case. In any

event, because Plaintiffs fail to show deliberate indifference,

as explained herein, they could not survive summary

judgment under either standard.

13

Up to now, our over-detention jurisprudence has

concerned individual plaintiffs challenging decisions specific

to themselves. In Sample v. Diecks, our first decision in this

line of cases, a prison records officer mistakenly determined

that an inmate still had time to serve on another sentence and

authorities therefore refused to release him. 885 F.2d at 1102.

Then came Moore v. Tartler, 986 F.2d 682 (3d Cir. 1993),

which concerned parole officers who initially misinterpreted a

judge’s sentencing order, resulting in a six-month delay in the

inmate’s release while officials conducted an investigation.

Most recently, Montanez v. Thompson likewise involved

various calculations of a particular inmate’s sentence. 603

F.3d at 246-48.

In contrast, Plaintiffs here allege systemic

shortcomings at COR. The problems are not prisoner-specific

misapplications of the law, but organizational policies and

practices. Nevertheless, the same standard applies, although

it must be applied with sensitivity to the change of context.

In particular, we have noted that the position of the defendant

in an over-detention suit must affect the second prong of our

test: deliberate indifference. “Among the circumstances

relevant to a determination of whether the requisite attitude

was present are the scope of the official’s duties and the role

he or she has played in the everyday life of the prison.”

Sample, 885 F.2d at 1110. An official is less likely to display

deliberate indifference if “there are procedures in place

calling for others to pursue the matter” and more likely to be

deliberately indifferent if given his or her role, a problem

14

“will not likely be resolved unless he or she addresses it or

refers it to others[.]” Id.11

This flexible standard for deliberate indifference,

which foresees suits against officials up and down the

organizational chart of the prison system, anticipates that

over-detention claims may raise structural challenges as well

as individual ones. Indeed, this is the necessary corollary of

our recognition that the “administration of a system of

punishment entails an unavoidable risk of error.” Sample,

885 F.2d at 1108. Operating a prison system is a major

bureaucratic undertaking. That fact compels us to offer

individual prison officials room for imperfection and

accidents. But it also teaches that preventing over-detentions

may require bureaucratic solutions from top management.

Litigation against top administrators, seeking structural

reforms of the agency as a whole, may be the only effective

way to reduce the overall risk of unconstitutional error.

Where appropriate, we must treat the correctional system as a

system.

Suits against high-level government officials must

satisfy the general requirements for supervisory liability. In

11

Here, Appellees are the officials tasked with resolving the

alleged problems. Appellants allege that structural features of

COR cause a systemic over-detention problem. Only top

administrators, not line staff processing individual releases,

can increase staffing levels, foster a more open culture of

communication or declare the creation of COR a success or

failure. The problems alleged in this litigation “will not likely

be resolved unless” top administrators like Appellees address

them. Sample, 885 F.2d at 1110.

15

particular, supervisors are liable only for their own acts; in

this case, they are liable only if they, “with deliberate

indifference to the consequences, established and maintained

a policy, practice or custom which directly caused [the]

constitutional harm.” A.M. ex rel. J.M.K. v. Luzerne Cty.

Juvenile Det. Ctr., 372 F.3d 572, 586 (3d Cir. 2004) (quoting

Stoneking v. Bradford Area Sch. Dist., 882 F.2d 720, 725 (3d

Cir. 1989)) (alteration in original). This standard for

supervisory liability largely overlaps with the over-detention

standard—both require a showing of deliberate indifference

and causation—but centers the inquiry around a policy or

practice.

B. Eighth Amendment Analysis

We agree with the District Court that Appellants

established a genuine dispute of material fact as to the first

prong of the over-detention standard: knowledge of a risk of

unwarranted punishment. Fundamentally, COR was created

because DDOC was aware of what it believed to be an

unacceptable level of over-detention in Delaware.12

Moreover, there is record evidence that those near to the

correctional system warned Appellees of continuing over-

detention problems post-2008, including the public defender,

Sandra Dean, and the bail bondsperson, Bruny Mercado.

While the record does not allow for the exact calculation of

over-detention levels or year-by-year trends in over-detention,

a jury could reasonably find the overall level of over-

detention to be quite substantial. Indeed, McBride’s own

12

Whether the pre-2008 system in fact failed to meet

constitutional standards is outside of our purview in this

matter.

16

testimony shows an awareness of continuing challenges at

COR: She admitted at her deposition that she has always

perceived a problem with the timely processing of releases,

since the creation of COR. Her efforts to improve COR came

in response to a sense that there was a need for change in the

agency’s operations.

Appellants fail, however, to show a genuine dispute of

material fact as to the second prong: deliberate indifference.

As the District Court held, the record shows a variety of

efforts by Appellees to improve COR and address the over-

detention problem. For example, not only did McBride

increase staffing levels at COR, she offered uncontradicted

testimony that she did so specifically in response to delays in

processing. Likewise, her efforts to improve the agency’s

training system, to upgrade its technology, and to create

special units to more efficiently handle certain types of

release show, as she testified, that COR leadership was

“always looking at ways to be more efficient.” JA 12.

These facts weigh heavily against any reasonable

finding of deliberate indifference. In Moore v. Tartler, we

observed that deliberate indifference had been “demonstrated

in those cases where prison officials were put on notice and

then simply refused to investigate a prisoner’s claim of

sentence miscalculation.” 986 F.2d at 686 (citing Alexander

v. Perrill, 916 F.2d 1392, 1398 (9th Cir. 1990) and Haygood

v. Younger, 769 F.2d 1350 (9th Cir. 1985)). Because the

parole board officials’ investigation in Moore constituted

“affirmative steps” to resolve the issue, we could not find

deliberate indifference. Id. at 687. Here, Appellees have also

taken affirmative steps to address over-detentions in the

Delaware system and this makes a finding of deliberate

indifference difficult.

17

But the presence of such affirmative steps is not

necessarily dispositive. A further inquiry, not undertaken by

the District Court, is required in this case. In Moore, the

parole officers’ investigation was targeted to resolve the

inmate’s complaint in full. Although the investigation was

slow, the officials could safely conclude that upon its

completion they would have done all that they could do to

address Moore’s over-detention. Here, it is not self-evident

that the COR reforms were up to the task at hand.

“[I]neffectual action under the circumstances” can also

indicate deliberate indifference. Montanez, 603 F.3d at 252.

This is not to say that federal courts conduct

independent reviews of the wisdom of prison policy. The

purpose of addressing “ineffectual action” is not to render a

program evaluation. The ultimate subject of inquiry remains

deliberate indifference: a state of mind. We look to see

whether the gap between the officials’ actions or inaction and

the problem they were trying to solve was so large that those

actions display deliberate indifference. Imagine an inmate

who came to the prison infirmary with a cut and with kidney

failure and was given only a bandage. We would have no

trouble concluding that this could constitute deliberate

indifference. Cf. Rouse v. Plantier, 182 F.3d 192, 194-95,

198-99 (3d Cir. 1999) (suggesting that providing diabetic

inmates with only one insulin injection per day and less-than-

daily blood sugar monitoring, when they needed more, can

constitute deliberate indifference).

By the same token, supervisory efforts to minimize

over-detentions in one manner could, in principle, co-exist

with deliberate indifference to a festering over-detention

problem rooted in different agency practices or policies.

Given the allegations of rampant over-detention—affecting as

18

many as one-third of inmates—a jury could might reasonably

ponder whether something along these lines was occurring.

But an argument of this sort requires evidence to

survive summary judgment. While Appellants may have

shown a genuine dispute whether over-detentions remain a

large-scale problem in the Delaware correctional system,

there is no genuine dispute regarding whether Appellees have

tried to address the over-detention problem. Viewing the

facts in the light most favorable to Appellants, we could

conclude that over-detentions are rampant in Delaware and

that correctional officials are trying, albeit without great

success, to tackle that challenge. So far, this is not deliberate

indifference. Appellants need more to rescue their claim.

They would need to show that Appellees’ efforts to improve

COR so obviously miss the mark that pursuing those efforts

manifests disregard for the real problem and thereby amounts

to deliberate indifference. Such evidence is absent from the

record.

On summary judgment, the nonmoving party must

affirmatively “show where in the record there exists a genuine

dispute over a material fact.” Doe v. Abington Friends Sch.,

480 F.3d 252, 256 (3d Cir. 2007). “[S]peculation and

conjecture may not defeat a motion for summary judgment[.]”

Acumed LLC v. Advanced Surgical Servs., Inc., 561 F.3d 199,

228 (3d Cir. 2009) (citing Lexington Ins. Co. v. W. Pa. Hosp.,

423 F.3d 318, 332-33 (3d Cir. 2005)). But speculation is all

that Appellants can put forward to show that the DDOC

officials were deliberately indifferent notwithstanding their

efforts to improve the release process. The record offers no

reason to believe that Appellees’ chosen interventions were

callously misguided.

19

For example, Appellants offer an affidavit from a

former COR employee stating that the computerized DACS

system increased delay. Whether DACS was a successful

technological upgrade is not a dispute sufficient to go to trial.

It does not show a dispute that computerization was

misguided from the start. Moreover, McBride testified that

DACS has helped COR be more efficient and has sped up the

release process. She admitted that there were issues with the

system that needed to be resolved, but did not indicate that

COR would leave those issues unaddressed. DACS may have

been “ineffective” in the sense that it did not immediately

reduce over-detentions—but that is not the test. Rather,

Appellants must show that COR’s ineffectiveness amounted

to deliberate indifference.

Likewise, it may be the case that COR needs more

staff. There is evidence to that effect, sufficient to create a

factual dispute. But the record also shows that McBride

observed a need for more staff, worked with other officials to

calculate how many more staff were required, and secured

those positions from more senior authorities within state

government. Separately, COR also increased the number of

casual/seasonal staff used to pull and re-file records, again in

response to a perceived need and a review of agency

operations. To show deliberate indifference on the staffing

issue, Appellants would need evidence that would allow a

jury to conclude, for example, that COR knew its staffing

increases would be insufficient or that after it realized it still

needed more staff after the first round of hires, it did nothing

in response. No such evidence is in this record.

With regards to COR’s unresponsiveness to outside

communications, a different sort of evidence would be needed

to establish deliberate indifference. Appellants demonstrated

20

a dispute whether COR was open to inquiries from inmates

and whether increased openness would help reduce over-

detentions. As already noted, there is evidence to suggest that

inmates were urged to contact COR to expedite their release

and that where contact was made, such as by a public

defender, it really did spur action. But the record does not

show that COR took any particular action to improve this

problem. Although this may be more indicative of deliberate

indifference, what is absent here is any evidence showing that

COR should have addressed this problem with particular

alacrity as opposed to any other. Had the crux of the

evidence presented to the District Court been that closed

channels of communication caused particularly large numbers

of over-detentions; that an alternative system had been

presented to COR or was a best practice they should have

known to adopt; or that changes to COR’s communications

policy would have been easier or more efficacious than

COR’s other reform efforts, then Appellants may have been

in a different posture regarding summary judgment.13

Nor could a reasonable jury infer deliberate

indifference from the simple fact that over-detentions

increased in this period (if the jury found that they did).

There are surely many variables that affect the over-detention

13

Even less indicative of deliberate indifference is

Appellants’ assertion that COR failed to track certain

performance metrics, such as the number of lost files or the

number of inmate letters received. No record evidence is put

forward that could allow a finding that these metrics should

have been used rather than the alternative forms of tracking

and oversight employed at COR, much less that Appellees

were deliberately indifferent for failing to use them.

21

problem. In particular, there is substantial and uncontested

record evidence that many over-detentions originate in the

court system rather than at COR. It is entirely plausible that

the overall increase in over-detentions stemmed from changes

outside COR and that the reform efforts at COR, though

effective, were swamped by external forces. To survive

summary judgment, Appellants need more than speculation

connecting any increase in over-detentions with the COR

policies they deem ineffective.

A comparison with a similar, successful over-detention

suit is instructive. In Barnes v. District of Columbia, 793 F.

Supp. 2d 260 (D.D.C. 2011), the Court was presented with a

far richer evidentiary picture—and plaintiffs were able not

only to avoid summary judgment, but to win summary

judgment themselves on certain of their claims. There, as

here, the correctional officials had made efforts to reduce

over-detentions and there, as here, plaintiffs argued that those

efforts were ineffectual. But plaintiffs were able to carry their

burden. They hired a statistical expert to sift through

correctional records and provide reliable annual estimates of

how many people had been over-detained. Id. at 269-70.

What is more, they were able to estimate how many of those

over-detentions were attributable to specific policies. Id. at

271. This allowed the Court to determine that the District of

Columbia’s early efforts to reduce over-detentions were

utterly ineffectual, allowing a grant of summary judgment for

plaintiffs, and that the District’s later efforts were quite

effective, allowing a grant of summary judgment for

defendants (whether the District’s efforts during an

intermediate period showed deliberate indifference required

factfinding). Id. at 280-81. Plaintiffs in Barnes could also

demonstrate precisely how long processing a release should

22

take and how much longer it often took in practice. Id. at

278-79. This sort of data allowed plaintiffs to show

deliberate indifference to over-detentions, even in the face of

affirmative steps to improve matters. Appellants have not

shown deliberate indifference here. We therefore affirm the

District Court’s grant of summary judgment for Appellees on

all Eighth Amendment claims.

C. Fourteenth Amendment Analysis

For the same reasons, we affirm the District Court’s

grant of summary judgment on all Fourteenth Amendment

substantive due process claims. The District Court dismissed

Plaintiffs’ Fourteenth Amendment claims under the “more-

specific-provision rule.” That rule holds that “if a

constitutional claim is covered by a specific constitutional

provision, such as the Fourth or Eighth Amendment, the

claim must be analyzed under the standard appropriate to that

specific provision, not under the rubric of substantive due

process.” United States v. Lanier, 520 U.S. 259, 272 n.7

(1997); see also Betts, 621 F.3d at 261. The District Court

held that because the Third Circuit addresses over-detention

under the rubric of the Eighth Amendment, Plaintiffs could

not bring parallel claims under the Fourteenth Amendment’s

protection of substantive due process.

Appellants and amicus argue that the more-specific-

provision rule does not apply to all claims, because some

plaintiffs were pretrial detainees, who are not protected by the

Eighth Amendment.14 Our Court has always analyzed over-

14

They also argue that Plaintiffs’ claims arose under a

procedural due process framework rather than a substantive

due process framework. This argument is waived. Plaintiffs’

23

detention claims under the Eighth Amendment, unlike some

other courts. See Barnes, 793 F. Supp. 2d at 274-75

(“Overdetentions potentially violate the substantive

component of the Due Process Clause. . . .”). But we have

applied the Eighth Amendment because each of our over-

detention cases involved convicted and sentenced inmates.

Montanez, 603 F.3d 243; Moore, 986 F.2d 682; Sample, 885

F.2d 1099.

Our precedent is clear that while the detention of

sentenced inmates is governed by the Eighth Amendment, the

treatment of pretrial detainees is governed by the Due Process

Clause. Bell v. Wolfish, 441 U.S. 520, 535 n.16 (1979) (“Due

process requires that a pretrial detainee not be punished. A

sentenced inmate, on the other hand, may be punished,

although that punishment may not be ‘cruel and unusual’

under the Eighth Amendment.”); Boring v. Kozakiewicz, 833

F.2d 468, 471 (3d Cir. 1987) (“Pretrial detainees are not

within the ambit of the Eighth Amendment but are entitled to

the protections of the Due Process clause.”); Hubbard v.

Taylor, 399 F.3d 150, 164-67 (3d Cir. 2005) (same).

For pretrial detainees, therefore, there is no applicable

provision more specific than the Due Process Clause and the

more-specific-provision rule does not apply. A separate due

process analysis is required.

The protections of the Eighth Amendment and Due

Process Clauses are sometimes, but not always, the same.

Hubbard, 399 F.3d at 164-67. We need not delve into the

arguments below expressly identified their claims as being for

substantive due process.

24

differences between those two analyses in this context,

however. This is a suit against supervisory officials, for the

creation of policies and practices. Supervisory policy-and-

practice liability requires deliberate indifference. A.M. ex rel.

J.M.K., 372 F.3d at 586. Thus, for the same reasons as in our

Eighth Amendment analysis, we conclude that there is no

genuine dispute of material fact as to deliberate indifference

under the Fourteenth Amendment. We will affirm.15

15

Having affirmed the District Court’s grants of summary

judgment on the merits, we need not reach the other issues in

the Court’s opinion: qualified immunity, sovereign immunity,

and class certification. We also need not reach the causation

prong of the Eighth Amendment analysis.

25