558 F.3d 869

9th Cir.

2009

JOSEPH CARVER, Plaintiff-Appellant, v. JOSEPH LEHMAN; KIMBERLY ACKER; VICTORIA ROBERTS; SIX TO BE NAMED DEFENDANTS, Defendants-Appellees.

No. 06-35176

UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT

March 3, 2009

558 F.3d 869

Before: Stephen Reinhardt, Richard C. Tallman, and Milan D. Smith, Jr., Circuit Judges.

D.C. No. CV-04-05570-RBL. Appeal from the United States District Court for the Western District of Washington. Ronald B. Leighton, District Judge, Presiding. Argued and Submitted April 17, 2007—San Francisco, California. Filed December 22, 2008. Amended March 3, 2009.

COUNSEL

Tyler A. Baker, Todd Gregorian, and Heather N. Mewes, Fenwick & West, LLP, Mountain View, California, for the plaintiff-appellant.

Rob McKenna, Sara J. Olson, and Gregory J. Rosen, Office of the Washington Attorney General, Criminal Justice Division, Olympia, Washington, for the defendants-appellees.

ORDER

The opinion and concurrence filed on December 22, 2008, and appearing at 550 F.3d 889 (9th Cir. 2008) are hereby amended. The amended opinion and concurrence are filed concurrently with this order.

The petition for rehearing en banc filed January 20, 2009, remains pending.

No further petitions for rehearing or for rehearing en banc may be filed.

OPINION

MILAN D. SMITH, JR., Circuit Judge:

This case presents the question whether a Washington state law providing for convicted sex offenders’ early release into community custody creates a liberty interest that is protected under the Due Process Clause of the Fourteenth Amendment. We hold that it does not. We therefore affirm the decision of the district court denying Carver relief in this civil rights action.

Factual and Procedural Background

In August 1999, Joseph Dale Carver, then 20 years old, pled guilty to child molestation in the third degree. This conviction followed two prior convictions for child molestation in the first degree and a conviction for third-degree assault. Carver committed his first sex offense at age 14. He was sentenced to fifty-four months of confinement in the custody of the Washington State Department of Corrections (“DOC“) for his 1999 conviction. The brief record on appeal indicates Carver committed fifteen disciplinary infractions while incarcerated, including sexual harassment of a prison staff member.

Washington Revised Code § 9.94A.728(1)(b)(ii)(B)(I) prohibits early release for those convicted of sex offenses. However, section 9.94A.728(2)(a) provides that sex offenders may become eligible for transfer to community custody in lieu of early release.¹ Carver was sentenced to a consecutive thirty-six month period of community custody to begin on his adjusted release date.² Carver‘s behavior as a prisoner resulted in an adjusted release date of January 13, 2003.

Before an inmate is eligible for transfer to community custody, he must submit an acceptable “release plan.” WASH. REV. CODE § 9.94A.728(2)(c). Carver submitted his proposed plan in March 2002. It was denied in April 2002, pursuant to a DOC policy then in effect which provided for the categorical denial of release plans of offenders, like Carver, whom the DOC determined “appear[ed] to meet the definition of a sexually violent predator and [who had] been referred for Civil Commitment . . . .” DOC Policy Directive 350.200 (May 4, 2001).³ As a result of the denial of his proposed release plan, Carver served his full term of confinement.

In September 2004, Carver filed a civil rights suit under 42 U.S.C. § 1983, asserting that DOC officials denied him early release into community custody without affording him due process of law under the Fourteenth Amendment.⁴ The district court, adopting the report and recommendation of the magistrate judge, granted the DOC officials’ motion for summary judgment on two principal grounds: first, that Washington law does not create a liberty interest in early release into community custody and, therefore, Carver did not have a due process right protected by the Fourteenth Amendment; and second, that even if such a right existed, Defendant Lehman was entitled to qualified immunity. Carver timely appealed.

Jurisdiction and Standard of Review

We have jurisdiction to review the district court‘s determination pursuant to 28 U.S.C. § 1291, and we review de novo its grant of summary judgment and finding of qualified immunity. See Mabe v. San Bernardino County, Dep‘t of Pub. Soc. Servs., 237 F.3d 1101, 1106 (9th Cir. 2001); Galen v. County of Los Angeles, 477 F.3d 652, 658 (9th Cir. 2007).

Discussion

[1] The Due Process Clause of the Fourteenth Amendment provides that no state shall “deprive any person of life, liberty, or property, without due process of law. . . .” U.S. Const. amend. XIV, § 1. Our analysis of due process claims proceeds

in two steps. “[T]he first asks whether there exists a liberty or property interest which has been interfered with by the State; the second examines whether the procedures attendant upon that deprivation were constitutionally sufficient.” Ky. Dep‘t of Corr. v. Thompson, 490 U.S. 454, 460 (1989) (citation omitted).

[2] “A liberty interest may arise from either of two sources: the due process clause itself or state law.” Toussaint v. McCarthy, 801 F.2d 1080, 1089 (9th Cir. 1986). Carver concedes that the Due Process Clause does not create a liberty interest in an inmate‘s “conditional[ ] release[ ] before the expiration of a valid sentence.” Greenholtz v. Inmates of the Neb. Penal & Corr. Complex, 442 U.S. 1, 7 (1979). Rather, he argues that Washington‘s statutory scheme governing early release into community custody “uses mandatory language, ‘creat[ing] a presumption that . . . release will be granted’ . . . unless certain designated findings are made, and thereby gives rise to a constitutional liberty interest.” McQuillion v. Duncan, 306 F.3d 895, 901 (9th Cir. 2002) (quoting Greenholtz, 442 U.S. at 12; citing Bd. of Pardons v. Allen, 482 U.S. 369, 377-78 (1987)). As in prior cases, our task here is to apply the well-established mandatory language rule governing state-created liberty interests set forth by the Supreme Court in Greenholtz and Allen to the Washington sex offender statutory scheme at issue.⁵ See, e.g., Sass v. Cal. Bd. of Prison

Terms, 461 F.3d 1123, 1127-28 (9th Cir. 2006) (holding that California law creates a liberty interest in parole); Biggs v. Terhune, 334 F.3d 910, 914 (9th Cir. 2003) (same); McQuillion, 306 F.3d at 901-902 (same); Bermudez v. Duenas, 936 F.2d 1064, 1065-66 (9th Cir. 1991) (holding that Guam law creates a liberty interest in parole); Baumann v. Ariz. Dep‘t of Corr., 754 F.2d 841, 843-45 (9th Cir. 1989) (holding that Arizona law does not create a liberty interest in custodial release); Balla v. Idaho State Bd. of Corr., 869 F.2d 461, 469-70 (9th Cir. 1989) (holding that Idaho law does not create a liberty interest in parole).

[3] Washington law mandates that an individual convicted of a sex offense be sentenced to a term of community custody that “shall begin either upon completion of the term of confinement or at such time as the offender is transferred to community custody in lieu of earned release.” WASH. REV. CODE § 9.94A.710(1). Unlike other inmates, then, a convicted sex offender who accrues “earned release time . . . for good behavior and good performance” is not entitled to early release; rather, he is eligible for discretionary transfer into community custody at an earlier date if his proposed placement is appropriate. Id. § 9.94A.728(1), (2)(a). The law requires the DOC to develop a suitable program to effectuate the transfer to community custody of such inmates. See id. § 9.94A.728(1). As part of that program, the DOC must “require the offender to propose a release plan that includes an approved residence and living arrangement.” Id. § 9.94A.728(2)(c). The law then describes how the DOC, in exercising its broad discretion, should evaluate such release plans:

The department may deny transfer to community custody status in lieu of earned release time pursuant to subsection (1) of this section if the department determines an offender‘s release plan, including proposed residence location and living arrangements, [1] may violate the conditions of the sentence or conditions of supervision, [2] place the offender at risk to violate the conditions of the sentence, [3] place the offender at risk to reoffend, or [4] present a risk to victim safety or community safety. The department‘s authority under this section is independent of any court-ordered condition of sentence or statutory provision regarding conditions for community custody or community placement . . . .

Id. § 9.94A.728(2)(d) (emphasis added).

[4] In order to comply with the statute, the DOC promulgated Policy Directive 350.200. Under the version of this policy in force when Carver submitted his release plan, the DOC instructed that release plans of sex offenders be assessed to determine “the degree of risk for victims and potential victims of similar age or circumstances” and to ensure that, subject to certain exceptions, “[s]ex offenders will not return to a residence where minor victim(s) or other children of similar age are present in the residence.” DOC Policy Directive 350.200 (May 4, 2001). This Policy Directive specified that a residence proposed by an offender within a release plan could be denied if the proposed location would place the offender in violation of court-imposed conditions; at the likely risk to reoffend; or in close proximity to the minor victim(s), schools, child care centers, playgrounds, or other facilities where children of similar age and circumstances surrounding the conviction are present and who may be put at substantial risk of harm by the offender residing at that location.⁶ The DOC‘s policy also provided for the categorical denial of release plans “if the End of Sentence Review Committee has determined that the offender appears to meet the definition of a sexually violent predator and s/he has been referred for Civil Commitment . . . .”⁷ This final provision, under which Carver‘s release plan was denied, was subsequently eliminated after the Washington Court of Appeals held that it violated the statutory requirement that all sex offenders “may become eligible” for community custody. See Dutcher, 60 P.3d at 638-40.

Carver argues that this statutory scheme creates a protected liberty interest because it requires the DOC to transfer an inmate to community custody in lieu of earned release “unless any one of the . . . specifically designated reasons are

found[,]” thereby “creat[ing] a presumption that . . . release [into community custody] will be granted, and that this in turn creates a legitimate expectation of release absent the requisite finding that one of the justifications for [denial] exists.” Greenholtz, 442 U.S. at 11-12; see also Allen, 482 U.S. at 377-78. We disagree. In order to create a constitutionally protected liberty interest, a statute must contain “‘explicitly mandatory language,’ i.e., specific directives to the decisionmaker that if the regulations’ substantive predicates are present, a particular outcome must follow.” Thompson, 490 U.S. at 463 (quoting Hewitt v. Helms, 459 U.S. 460, 472 (1983)). There is no “explicitly mandatory language” in section 9.94A.728(2) creating a substantive right to transfer to community custody. The statute, using classically permissive language, states that a “person convicted of a sex offense . . . may become eligible, in accordance with a program developed by the department, for transfer to community custody status in lieu of earned release time.” WASH. REV. CODE § 9.94A.728(2)(a). The only “explicitly mandatory language” in section 9.94A.728(2) concerns a procedural right to an individualized determination based on the merits of a proposed release plan.⁸ That language cannot create a “liberty interest” within the meaning of the Fourteenth Amendment because “expectation of receiving process is not, without more, a liberty interest protected by the Due Process Clause” of the Fourteenth Amendment. Olim v.

Wakinekona, 461 U.S. 238, 250-51 n.12 (1983); see also In re Cashaw, 866 P.2d 8, 12 (Wash. 1994) (“The United States Supreme Court and the Ninth Circuit have clearly held that procedural laws do not create liberty interests; only substantive laws can create these interests.“).⁹

[5] Pursuant to that procedural mandate, the DOC has no “discretion to decide whether or when to consider an offender for transfer to community custody,” In re Liptrap, 60 P.3d 1227, 1232 (Wash. Ct. App. 2005) (emphasis added). But Washington law places no substantive limitation on how the DOC is to make that determination. As noted above, section 9.94A.728(2)(d) enumerates four criteria for evaluating the transfer plan. The statute instructs that the DOC ”may deny transfer to community custody if” one or more of those criteria are met. Id. (emphasis added). Far from setting forth “substantive predicates” under which the DOC must grant transfer, the statute is silent regarding even precatory criteria for granting transfer to community custody, specifying only when the DOC “may“—but need not¹⁰ — “deny.”

No particular words are necessary to create a liberty interest.¹¹ If the section 9.94A.728(2)(d) criteria for denying transfer constituted an exhaustive list of reasons for denial, the language would be effectively mandatory, giving rise to a presumption of transfer. Nothing in the statute, however,

indicates that those four criteria are the sole reasons for which the DOC may deny transfer.¹² To the contrary: immediately following the enumeration, the statute goes on to state that the DOC‘s “authority under this section is independent of any court-ordered condition of sentence or statutory provision regarding conditions for community custody or community placement.” WASH. REV. CODE § 9.94A.728(2)(d). The statute‘s manifest purpose, therefore, is to preserve to the DOC the discretion to deny transfer in the event that it makes one of the four determinations, notwithstanding what other legal sources might otherwise require.

To convert this non obstante permissive clause, meant to preserve discretion in certain cases, into an expressio unius provision that would limit discretion to all but those cases would be to invert the very purpose for which this statute was drafted.¹³ These statutes were enacted to protect the commu-

nity from recidivist offenders whose proclivity to sexually abuse children is well-documented and whose anti-social tendencies are notoriously difficult to remedy through traditional penological methods. Carver‘s record would cause any reasonably cautious state official to pause before authorizing even closely supervised release to the community. On the basis of his prior record alone—even before this latest conviction—a Washington superior court found probable cause that Carver was a sexually violent predator. A forensic psychological examination had concluded that Carver was, in his own words, “a civil commitment case“—that is, he was found to meet the criteria of a sexually violent predator. Section 9.94A.728(2)(d) reserves discretion for DOC officials precisely so they may deny release plans of prisoners like Carver who remain threats to the community.

[6] Our reading of section 9.94A.728(2)(d) is consistent with the case law of the Washington state courts construing it.¹⁴ In the text of In re Liptrap, the Washington Court of Appeals noted that the statute “stat[es] reasons why the department may deny a release plan,” 111 P.3d at 1232, but kept those “reasons” in the indefinite: they are just “reasons“—not the reasons, let alone the only reasons—for denial. Similarly, In re Dutcher held simply that “the statute . . . requires DOC to base its community custody eligibility decisions on the merits of the release plan,” 60 P.3d at 638, which is hardly to imply that section 9.94A.728(2)(d) provides the sole rubric by which those “merits” are to be evaluated.

[7] It is true that the cases cited above refer to a “limited liberty interest” held by prisoners in transfer to community custody. Liptrap, 111 P.3d at 1231; see also Dutcher, 60 P.3d at 636 (describing a “limited but protected liberty interest” in

transfer); Crowder, 985 P.2d at 945 (“The statutory right to earned early release creates a limited liberty interest requiring minimal due process.“). We must not be confused, however, by those decisions’ use of a Fourteenth Amendment term of art: those cases concerned only the procedural right to compliance with individualized consideration on the merits of prisoners’ release plans, secured by Washington Rule of Appellate Procedure 16.4(a). Liptrap, 111 P.3d at 1234; Dutcher, 60 P.3d at 638. To the extent they contain dicta using the same term, “liberty interest,” to refer to both a substantive right to transfer and a procedural right to consideration on the merits, the most logical reading is that both derive from Washington law, and are hence of the same sub-constitutional nature. Regardless, those dicta provide no justification for disregarding the plain language of the statute.

Washington appellate courts have been careful to distinguish between the state habeas relief available through personal restraint petitions for violations of state law, and personal restraint petitions to redress violations of a constitutional magnitude. The decisions in Dutcher, Crowder, and Cashaw are examples of the former. In Cashaw, the Washington State Supreme Court explicitly rejected the lower court‘s holding that the Indeterminate Sentence Review Board‘s failure to follow mandatory parole procedures constituted an infringement of Fourteenth Amendment Rights. 866 P.2d at 12. Though it concluded that “the Board had violated its own procedural rules for parolability hearings,” it specifically held it was error to conclude “this violation was of constitutional magnitude.” Id. at 13. The cases that refer to a “limited liberty interest” do not support Judge Reinhardt‘s conclusion that state law regarding community custody creates a liberty interest arising under the Constitution.

Washington courts have implied only one limit on the substance of the DOC‘s exercise of discretion: its reasons for denial must be “legitimate.” Liptrap, 111 P.3d at 1234; Crowder, 985 P.2d at 946. But there is no indication that a reasonmay acquire “legitimacy” only by its enumeration in section 9.94A.728(2)(d). Indeed, every indication is to the contrary: In re Crowder, the first case to imply a requirement of “legitimate reasons” for denial of transfer to community custody, specified the petitioner‘s “own withdrawal of a suggested placement plan” as one of the “legitimate reasons” for denying him transfer, 985 P.2d at 946—a reason which certainly seems legitimate, but appears nowhere in section 9.94A.728(2)(d).¹⁵ No Washington case has provided a general definition for what makes a reason “legitimate,” which could be as broad as “any reason having a rational basis,” or even “any reason not otherwise proscribed by law.” This ill-defined “right” to transfer only in the absence of some “legitimate” reason to deny is hardly a sufficient “substantive predicate” to produce the “legitimate expectation of release” required of a protected liberty interest under Greenholtz. 442 U.S. at 12.

[8] The analysis employed by the Washington courts in parole and community release cases confirms our conclusion that section 9.94A.728(2)(d) does not create a liberty interest. DOC officials had legitimate concern that Carver, based on his multiple sex convictions and behavior in prison, had not presented an acceptable release plan to alleviate the concern for public safety. The Constitution requires no more.

Response to Preamble of Concurrence

We readily acknowledge the self-evident truth of Judge Reinhardt‘s observation that in our judicial system the outcome of important appellate cases can vary based on the composition of the judicial body or panel deciding those cases. It has been so ever since the founding of the Republic. We

respectfully disagree, however, with much of the balance of our concurring colleague‘s preambular observations about the processing of this case, and we feel a response is required.

Our colleague states that the “Constitution did not change between the time of the original panel‘s decision and the time of the new majority‘s opinion. All that changed is the composition of the three-judge panel.” Concurrence at 2511. This implies that the previous panel majority unearthed an unalloyed constitutional nugget waiting to be discovered within the primordial crust of the Fourteenth Amendment, but which must now be reinterred and disregarded as a result of the passing of one of our colleagues. This implication ignores Judge Reinhardt‘s candid admission that “[t]he constitutional question is a close one, and substantial arguments can be made for either position.” Concurrence at 2512. It further disregards both the rules of our court and the vicissitudes of life.

The respective corpora of the opinion and concurrence in this case discuss the disputed nature of the constitutional question. But whatever the merits of each side‘s constitutional analysis, we respectfully disagree with Judge Reinhardt‘s contention that the prior majority‘s opinion actually became a binding construction of the Constitution before Judge Ferguson‘s death, and that it is somehow unseemly to amend a published opinion when the reason for the change is the death of a member of the prior majority. No opinion of this circuit becomes final until the mandate issues, and the opinion issued by the prior majority was only part way through its finalization process.¹⁶ Until the mandate has issued, opinions can be,

and regularly are, amended or withdrawn, by the merits panel at the request of the parties pursuant to a petition for panel rehearing, in response to an internal memorandum from another member of the court who believes that some part of the published opinion is in error, or sua sponte by the panel itself. For example, in the ninety days between July 11, 2008, and October 9, 2008, at least ten published opinions were withdrawn¹⁷ and at least ten opinions were amended¹⁸ in our circuit. Thus, the prior majority‘s holding in this case may or may not have survived until the mandate issued, but it was certainly not yet enshrined as a binding construction of the Constitution when Judge Ferguson died. Even when an opinion deals with a constitutional issue, the referenced collaborative process strengthens, not weakens, the final quality of those opinions, thereby better enabling them to stand the test

of time, and engender the respect of thoughtful citizens for both the opinion, and the court that produced it.

Judge Reinhardt also suggests that the only appropriate way to have changed the previous panel‘s decision was to initiate an en banc process. En banc review is appropriate under our rules only when a panel decision (a) conflicts with decisions from another circuit, (b) conflicts with prior decisions of our own circuit, or (c) “substantially affects a rule of national application in which there is an overriding need for national uniformity.” 9th Cir. R. 35-1; see also Fed. R. App. P. 35(a)(1). As our colleague well knows, the previous panel‘s majority opinion does not technically qualify for en banc review under our rules,¹⁹ and it is very difficult to obtain en banc review of a case in our court.²⁰ Had we accepted his recommendation, the original opinion would undoubtedly have remained unchanged, the very result preferred by our colleague.

Conclusion

[9] Because Washington law does not create a liberty interest in transfer to community custody, we need not address the sufficiency of the procedures given (or denied) Carver. The judgment of the district court is AFFIRMED.

Each party shall bear its own costs on appeal.

REINHARDT, Circuit Judge, concurring in the judgment only:

Six months ago, the original panel in this case filed a majority opinion holding that Washington state law creates a liberty interest in an inmate‘s early release into community custody. We held that the prisoner‘s liberty interest is protected under the Due Process Clause of the Fourteenth Amendment and that, accordingly, when an inmate becomes eligible for a transfer to community custody, the prison authorities may deny his request for a transfer only for one of the reasons specified in the Washington statute — and only if he is afforded a minimal opportunity to present his side of the story before they do so. In short, we held that Washington prison authorities must follow Washington law and abide by the United States Constitution. Joining me in that opinion was Judge Warren J. Ferguson, who died before we could deny the petition for rehearing; dissenting was Judge Milan Smith. As a result of Judge Ferguson‘s death, it was necessary to replace him on this case with another member of this court drawn at random. There were no intervening decisions that changed the law between the time Judge Ferguson and I issued our opinion holding that a liberty interest exists that protects the prisoners’ rights at issue and the time that Judge Smith, joined by our colleague who replaced Judge Ferguson, issued a substitute opinion holding that no such liberty interest exists.

As stated above, it is indisputable that the law did not change and the Constitution did not change between the time of the original panel‘s decision and the time of the new majority‘s opinion. All that changed is the composition of the three-judge panel. To those who question whether the results in constitutional and other cases depend on the membership of the panel, or whether the replacement of even a single Supreme Court justice can change the fundamental nature of the rights of all Americans with respect to matters as basic as affirmative action, a woman‘s right of choice, and the nature of religious liberty, the result in the case currently before our panel is merely a minor illustration of how the judicial system currently operates. Solely because of fortuity, I am compelled to write in strong disagreement with the majority‘s constitutional analysis instead of simply reaffirming an opinion vindicating the constitutional rights of the petitioner and his fellow prisoners in the state of Washington.¹

In the case before us, it is not necessary for the new majority to undo the original majority‘s constitutional ruling, even if it disagrees with it. The constitutional question is a close one, and substantial arguments can be made for either position. Under these circumstances, the more important consideration, in my view, is maintaining the stability and legitimacy of the court‘s decisions. We have a procedure for correcting decisions that a majority of the court believes warrant reconsideration. That process is known as a en banc rehearing. It can be invoked if any single judge on the court, including either member of the majority, elects to make a call. Relying on this process would, in my view, be in the better interests of the court and the judicial system; increasing the extent to which judicial decisions depend on chance and subjectivity is not a wise alternative.

Let me make it clear that I have not suggested, nor do I believe, that Judge Smith and Judge Tallman do not have the authority to withdraw the opinion filed by the initial panel and published as the opinion of the court. Of course, they do.² Nor

have I suggested that the opinion was “final“: I merely stated what I strongly believe — that it is unwise for a court, once it has published an opinion on a constitutional question, to change its mind for so fortuitous and subjective a reason. Rather, I suggested that a procedure exists under which we could reach a different result through a more objective process in which the merits, not the composition of the panel, would provide the basis for our action. Proceeding in the latter manner would help to secure the legitimacy of court decisions and, necessarily, to maintain public confidence in the judicial system.

I. Liberty Interest

As Judge Ferguson and I previously held, Washington‘s statutory scheme creates a protected liberty interest because it requires the Department of Corrections (“DOC“) to transfer becomes aware on its own, of a “point of law or fact that . . . the court has overlooked or misapprehended . . . .” Fed. R. App. P. 40. I must note, however, that this practice of withdrawing published opinions is employed far more sparingly than Judge Smith suggests. Of the ten cases which he cites as “published opinions [that] were withdrawn,” Maj. Op. at 2509, only five other than the one before us fit that description. In two of those five, the panel remanded to an administrative agency in its revised decisions but ordered similar or broader relief than in the initial opinions. See Al-Mousa v. Mukasey, 2008 WL 4372965, at *2 (9th Cir. Sep. 22, 2008); Ctr. for Biological Diversity v. Nat‘l Highway Traffic Safety Admin., 538 F.3d 1172 (9th Cir. 2008). In a third, the panel dismissed a petition rather than remanding to an administrative agency after the government submitted the full record in place of the incomplete record it initially provided to the panel. Granados-Oseguera v. Mukasey, 546 F.3d 1011 (9th Cir. 2008). In a fourth case, the panel changed its reason for allowing the government to pursue a restitution order, preferring to rely on federal preemption. United States v. Gianelli, 543 F.3d 1178 (9th Cir. 2008). In only one case did the panel reverse a basic position it had taken. Johnson v. Riverside Healthcare Sys., 534 F.3d 1116 (9th Cir. 2008). In none of these cases was a constitutional question at issue, and in none was the withdrawal prompted by a matter unrelated to the law, such as the reconstitution of the panel. an inmate to community custody in lieu of earned release “unless any one of the . . . specifically designated reasons are found[,]” thereby “creat[ing] a presumption that . . . release [into community custody] will be granted, and . . . in turn creat[ing] a legitimate expectation of release absent the requisite finding that one of the justifications for [denial] exists.” Greenholtz v. Inmates of the Nebraska Penal and Corr. Complex, 442 U.S. 1, 11-12 (1979). See also Bd. of Pardons v. Allen, 482 U.S. 369, 377-78 (1987). The majority rejects this argument, finding no “‘explicitly mandatory language’ . . . creating a substantive right to transfer to community custody” under Washington law. Maj. Op. at 2502 (emphasis omitted). For the reasons set forth in our prior majority opinion and reiterated below, I believe that the current majority‘s reading is flawed.

The Washington statutory scheme uses language that effectively mandates the transfer to community custody of those inmates who have earned release time and who have not been found to meet one of the statutory reasons for denial of a release plan set forth in Washington Revised Code § 9.94A.728(2)(d). Section 9.94A.710(1) requires that sex offenders be sentenced to a term of community custody to begin either when the offender‘s term of confinement is complete or when he is transferred as a result of earned release time. Section 9.94A.728(1) likewise requires the DOC to develop and promulgate procedures by which a sex offender may become eligible for transfer to community custody in lieu of earned release time. The same section then sets forth the limited circumstances under which the Department “may deny” an inmate‘s proposed plan for transfer to community custody. WASH. REV. CODE § 9.94A.728(2)(d) (listing as legitimate reasons for denial, a DOC determination that the release plan may violate the conditions of the sentence or conditions of supervision, place the offender at risk to violate the conditions of the sentence, place the offender at risk to reoffend, or present a risk to victim or community safety). By placing substantive limitations on DOC‘s discretion to deny release plans and, in particular, by requiring that denial of such plans be based on the limited criteria contained in section 9.94A.728(2)(d), Washington has created a liberty interest in early release into community custody that is protected by the Due Process Clause of the Fourteenth Amendment.³ See Allen, 482 U.S. at 375-76 (clarifying that a state may grant “significant discretion to the decisionmaker” to apply “general or broad release criteria” without “depriv[ing] the prisoner of the liberty interest in parole[,]” so long as “release is

required after the [decisionmaker] determines (in its broad discretion) that the necessary prerequisites exist“). See also Baumann, 754 F.2d at 844 (noting that “[a] state may create a constitutionally protected liberty interest by establishing regulatory measures that impose substantive limitations on the exercise of official discretion“); Bergen v. Spaulding, 881 F.2d 719, 721 (9th Cir. 1989) (“A board charged with deciding a prisoner‘s early release may be delegated significant discretion in making its decision, and yet be constrained by legal standards in exercising that discretion . . . . that scheme may give rise to a liberty interest in early release.“) (internal citation omitted).

That the Washington statute governing transfer to community custody does not use the more traditional mandatory language formula used in other early release statutes, i.e. stating that the DOC “shall” grant release “unless” certain findings are made, does not require a contrary result. See Allen, 482 U.S. at 378 (rejecting the argument that a statute must contain the “shall/unless” formula in order to create a liberty interest). Although we have accorded significance to the use of the term “shall” in assessing whether early release statutes create a liberty interest, we have never held that use of this term is required in order to satisfy the mandatory language rule set forth in Greenholtz and Allen. Nor does the majority so hold today. Maj. Op. at 2503-04. Despite the absence of the word “shall,” I conclude, as the prior majority did, that the language of section 9.94A.728(2)(d) is mandatory. Section 9.94A.728(2)(d) provides that the department “may deny transfer to community custody . . . if” any one of four criteria is met. By establishing criteria under which release may be denied, section 9.94A.728(2)(d) creates the presumption that, absent the existence of one those criteria, release will be granted. The repetition of the criteria, albeit in more specific form, in section 72.09.340(3)(a) and in Policy Directive 350.200 confirms this understanding of the mandatory nature of the statute.

The “may deny . . . if” formula operates in precisely the same manner and has precisely the same effect as a “shall grant . . . unless” clause. Under the “may deny . . . if” formula, the provision sets forth the conditions under which the agency may deny release. Otherwise, it must grant it. This is distinguishable from state statutes that provide that a decisionmaker “may grant . . . if” certain criteria are met. Under the “may grant . . . if” formula, the agency may only grant release if the relevant criteria are met, but it is not required to do so.⁴ Therefore, under the “may deny . . . if” formula, as under a “shall grant . . . unless” clause, there is an expectation that release will be granted unless one of the specified conditions exists.

Judge Smith, who separately concurred in the original majority opinion, remains unconvinced that the language of the statutory scheme is mandatory. My colleague who replaced Judge Ferguson agrees with him. But my reading is supported by that of the Washington state courts. In interpreting the statutory scheme governing early release into community custody, the state Court of Appeals has on two occasions discussed the mandatory nature of the law. First, in In re Dutcher, an inmate similarly situated to Carver challenged the DOC‘s failure to review his plan for release into community custody pursuant to the Department‘s policy of categorically denying the plans of those offenders who appeared to be sexually violent predators and who were referred for civil commitment. 60 P.3d 635, 635-36 (Wash. Ct. App. 2002). The court held that the DOC‘s policy violated the statutory mandate, explaining that “the statute compels DOC to require offenders to develop a release plan, and requires DOC to base its community custody eligibility decisions on the merits of the release plan.” Id. at 638 (emphasis added).

Second, my conclusion that requiring a decision on the merits of a release plan permits the Department to deny such a plan only if it finds one of the statutory criteria listed in section 9.94A.728(2)(d) is bolstered by the state court‘s interpretation of the statute in In re Liptrap. In Liptrap, inmates challenged the DOC‘s policy of refusing to review release plans of sex offenders until a forensic psychological evaluation had been completed. 111 P.3d 1227, 1229 (Wash. Ct. App. 2005). In finding that the Department‘s policy violated inmates’ due process rights, the court explained that “[t]he provisions in subsections [9.94A.728(2)] (c) and (d), spell[ ] out what is required in a release plan and stat[e] reasons why the department may deny a release[.]” Id. at 1232 (emphasis added). Accordingly, the Liptrap court found that “the department [does not have] unlimited discretion to decide whether and when to consider an offender for transfer to community custody.” Id. Rather, the DOC‘s failure to “state[ ] a legitimate reason” for the denial of a release plan, the court concluded, “deprived [the inmates] of earned early release credits in violation of due process.” Id. at 1234 (emphasis added).⁵

This holding is consistent with my conclusion that Washington law requires that the DOC‘s denial of a release plan be based on a “legitimate reason,” and that such reasons are enumerated in section 9.94A.728(2)(d).⁶

I conclude, then, as did the original panel majority, that the language of the Washington statutory scheme, as supported by Washington case law, mandates the transfer to community custody of those inmates who have earned release time and who have not been found to meet one of the statutory reasons for denial of a release, thereby creating a constitutionally-protected liberty interest in the transfer. This conclusion is further supported by Washington state court decisions finding a limited liberty interest in transfer to community custody in lieu of early release. The Washington Court of Appeals has consistently found a “limited liberty interest in early release into a community custody program . . . .” In re Crowder, 985 P.2d 944, 944-45 (Wash. Ct. App. 1999) (holding inmate had liberty interest in grant or denial of community custody placement upon earning of early release, but that the minimum level of due process required to protect this interest was provided). See also In re Dutcher, 60 P.3d at 636 (“An inmate‘s interest in his earned early release credits is a limited, but protected, liberty interest.“); In re Liptrap, 111 P.3d at 1231(same).⁷

In sum, I would hold once again that the Washington statutory scheme governing transfer to community custody of those inmates who have earned early release time creates a liberty interest that is protected under the Due Process Clause of the Fourteenth Amendment. Having found this liberty interest, I would, like the original panel, go a step beyond the current majority and consider “whether the procedures attendant upon th[e] deprivation [of Carver‘s liberty interest] were constitutionally sufficient[.]” Kentucky Dept. of Corrections v. Thompson, 490 U.S. 454, 460 (1989). Carver‘s release plan was denied under a categorical policy that provided him with no process at all; it simply rejected all plans of offenders who, like Carver, appeared to fall under the definition of a sexually violent predator and who were referred for civil commitment. Accordingly, the complete absence of procedures deprived Carver of his liberty interest in transfer to community custody without due process of law.

Finally, I would note that the majority‘s analysis of whether Carver presents a danger to the community is entirely beside the point. That is not a question for this court. Whether the prison officials followed the law is. Equally important, the fact that the statute creates a liberty interest does not mean that it does not also serve the purpose of protecting the community. The majority states that the purpose of the statute is to provide “discretion for DOC officials precisely so they may deny release plans of prisoners like Carver who remain threats to the community.” Maj. Op. at 2505. But in establishing a liberty interest, the Washington statute did not ignore legitimate concerns about the safety of the community. Indeed, among the four permissible reasons for denying transfer into community custody are whether such transfer would “place the offender at risk to reoffend, or present a risk to victim safety or community safety.” WASH. REV. CODE § 9.94A.728(2)(d). This threat to the community, however, must be established through a proper procedure in order to justify the deprivation of the inmate‘s liberty interest; it is not simply left to the unbridled discretion of the DOC. The point, then, is that inmates have a constitutionally-protected liberty interest in release to community custody such that when the DOC denies release, it must do so for one of the statutorily-enumerated reasons, and it must do so pursuant to proper procedures. Here, neither requirement was met. The prison officials simply ignored the statutory requirements and categorically denied all prisoners like Carver the release to which they may have been entitled. Thus, Carver and others were denied their due process rights. Let me make it absolutely clear, I do not contend that Carver was entitled to release. He may well not have been. On this record, however, we cannot know. Had the DOC followed the procedures provided by statute, and had it found that Carver‘s release “present[ed] a risk to community safety,” the prison authorities could well have retained him in custody. Following the law is not that difficult, and we are entitled to expect no less from our officials, prison or otherwise.

II. Qualified Immunity

Although unlike the current majority I conclude that Carver was deprived of a liberty interest, I would as did the original majority affirm the district court‘s finding of qualified immunity and therefore concur in the current majority‘s judgment.

Qualified immunity protects “government officials . . . from liability for civil damages insofar as their conduct does not violate clearly established statutory or constitutional rights of which a reasonable person would have known.” Harlow v. Fitzgerald, 457 U.S. 800, 818 (1982). The qualified immunity analysis proceeds in two parts. First, we consider whether “the facts alleged show that [Lehman‘s] conduct violated a constitutional right[.]” Galen, 477 F.3d at 658-59 (citing Saucier v. Katz, 533 U.S. 194, 200-02 (2001)). For the reasons discussed above, I would answer this question in the affirmative. Second, we ask if “the right [Lehman is] alleged to have violated [was] clearly established such that a reasonable [official] would have understood that he was violating that right[.]” Id. I conclude that the answer to this question is no.

In determining whether the right alleged to have been violated was clearly established, we must consider the right “in light of the specific context of the case, not as a broad general proposition[.]” Saucier, 533 U.S. at 201. “‘The contours of the right must be sufficiently clear that a reasonable official would understand that what he is doing violates that right.‘” Saucier, 533 U.S. at 202 (quoting Anderson v. Creighton, 483 U.S. 635, 640 (1987)). Here, because section 9.94A.728(2)(d) does not use the more common mandatory term “shall,” a reasonable correctional official might not have understood that the Washington statutory scheme created a liberty interest in early release into community custody. Certainly, the question was highly debatable at the time that Lehman was required to act. Carver‘s plan was denied before the Washington Court of Appeals had issued its decisions in Dutcher and Liptrap, which clarified that not only does a limited liberty interest exist under state law, but that the DOC‘s discretion to deny release into community custody is limited to rejection of a plan on the basis of the legitimate statutory criteria set forth in section 9.94A.728(2)(d).

Because I conclude, as did the original majority, that the right at issue here was not sufficiently clear at the time of the facts giving rise to this case such that a reasonable official would understand that denying a release plan without providing a legitimate statutory reason for that denial would violate due process, I would once again affirm the district court‘s grant of qualified immunity.

III. Conclusion

Washington state law creates a liberty interest in an inmate‘s early release into community custody that is protected under the Due Process Clause of the Fourteenth Amendment. Carver was denied his due process right by the state officials’ refusal to approve his release plan without reviewing it on its merits. At the time, however, the due process right arising from the existence of his liberty interest was not sufficiently clearly established to meet the Saucier standard. Because I would affirm the district court‘s determination that Lehman is entitled to qualified immunity, I therefore concur in the majority‘s judgment, but I respectfully disagree with its reasoning.

Notes

“Community custody is the intense monitoring of an offender in the community for a period of at least one year after release or transfer from confinement. Although it has other purposes, community custody continues in the nature of punishment, and is not equivalent to general release.” (footnote omitted). Offenders in community custody live in a residence pre-approved by the DOC and are subject to mandatory and discretionary conditions imposed by either the DOC or the sentencing court. See, e.g., WASH. REV. CODE §§ 9.94A.710(2)-(3), 9.94A.700(4)-(5) (listing mandatory and discretionary conditions imposed on offenders subject to community custody including reporting requirements, payment of supervision fees, and prohibitions against possession of controlled substances and consumption of alcohol). For the benefit of my colleagues in the majority, I should state that I, of course, do not believe that the original panel majority unearthed a “constitutional nugget,” Maj. Op. at 2508, alloyed or otherwise. This is simply a case in which Judge Ferguson and I tried our best to do our job, including the mundane task of seeing that prisoners, like all other persons, are afforded the rights to which they are entitled under the law. I recognize that these days there are many who do not share the view that prisoners’ rights deserve judicial protection, but neither Judge Ferguson nor I was aware of any constitutional or statutory provision to that effect, except perhaps for the Antiterrorism and Effective Death Penalty Act, 28 U.S.C. §§ 2241 et seq.

When an inmate is first transferred to the DOC, the DOC calculates three possible release dates for the inmate. First, the maximum release date is the date the inmate would finish serving the entire sentence imposed. Second, the earned early release date is the date the inmate would be released if he earned all available sentence reductions and does not lose time for misbehavior. Third, the adjusted release date is the projected date on which the inmate would be released if he loses no further good time or earned time credits. All would agree that an opinion may be properly withdrawn or amended when the panel is persuaded by a party or a colleague, or

As we explain infra, this policy was subsequently struck down by the Washington Court of Appeals in (holding that “DOC Policy 350.200 . . . violates the governing statutes“). At oral argument, Lehman urged that the presence of more specific criteria for denial of proposed residence locations in section 72.09.340(3)(a) undermines Carver‘s contention that the four criteria listed in section 9.94A.728(2)(d) provide the exclusive legitimate bases for denial of a release plan. Lehman is incorrect: the bases for denial of proposed residence locations contained in section 72.09.340(3)(a) are, in effect, specific variations of the same more general formulations set forth for denial of release plans in section 9.94A.728(2)(d). Section 72.09.340(3)(a) provides:

[T]he department shall not approve a residence location if the proposed residence: (I) Includes a minor victim or child of similar age or circumstance (as a previous victim who the department determines may be put at substantial risk of harm by the offender‘s residence in the household; or (ii) is within close proximity of the current residence of a minor victim, unless the whereabouts of the minor victim cannot be determined or unless such a restriction would impede family reunification efforts ordered by the court or directed by the department of social and health services. The department is further authorized to reject a residence location if the proposed residence is within close proximity to schools, child care centers, playgrounds, or other grounds or facilities where children of similar age or circumstance as a previous victim are present who the department determines may be put at substantial risk of harm by the sex offender‘s residence at that location.

These bases for denial of a proposed residence, which take into account whether the residence includes or is near a past or potential victim, are simply more specific examples of the final two criteria that serve as legitimate bases for denial under § 9.94A.728(2)(d), whether the “proposed residence location and living arrangements . . . place the offender at risk to reoffend, or present a risk to victim safety or community safety.” WASH. REV. CODE § 72.09.340(3)(a).

Carver‘s original complaint named as a defendant only Joseph Lehman, secretary of the DOC at the time that Carver‘s release plan was denied. In his amended complaint, Carver named two additional defendants, Kimberley Acker and Victoria Roberts, both DOC officials involved in making end of sentence review determinations. The district court adopted the magistrate judge‘s finding that Acker and Roberts were entitled to summary judgment because Carver failed to show that “these defendants played any part in enacting the policy that precluded [Carver] from being considered for release.” Carver does not appeal this portion of the judgment below. For this reason, courts have routinely held that the “may grant . . . if” formulation does not create a liberty interest. See, e.g., (New York parole statute); (South Dakota parole statute); (Missouri parole statute); (Arkansas parole statute); (Utah parole statute); (Mississippi parole statute); (New Mexico parole statute); (Texas parole statute); ; (New York parole statute); (Oklahoma parole statute); (Ohio parole statute).

In his response brief, Lehman argued that , should control our liberty interest inquiry. In Sandin, the Supreme Court considered a challenge to a prison regulation imposing disciplinary segregation for misconduct. . In holding that the regulation did not create a liberty interest, the Court did not apply the “mandatory language” framework of Greenholtz and Allen. . Instead, it focused on whether the challenged restraint arising from the regulation “impose[d] atypical and significant hardship on the inmate in relation to the ordinary incidents of prison life.” ; see also (applying Sandin to determine whether Ohio inmates have a liberty interest in avoiding placement in a “supermax” prison). As Lehman properly conceded in his supplemental submissions to the court, we have since held that Sandin‘s holding was limited to “the separate but related question of when due process liberty interests are created by internal prison regulations.” (emphasis added). See also (explaining that this court has “consistently rejected th[e] argument” that Sandin eliminated the “‘mandatory language’ approach of Greenholtz and Allen“). Accordingly, we continue to apply the “mandatory language” rule set forth in Greenholtz and Allen in order to determine whether Washington‘s statutory scheme creates a liberty interest in early release into community custody. The court specifically noted section 72.09.340‘s provision for denial of a release plan because the proposed residence is near young children as a “legitimate statutory reason for disapproving a release plan for a sex offender.” (citing WASH. REV. CODE § 72.09.340(3)). For the reasons set forth in footnote 3, supra, this provision is simply a specific example of the more general bases for denial set forth in section 9.94A.728(2)(d).

The last criterion derives from section 72.09.340(3)(a), which provides that the DOC is “authorized to reject a residence location if the proposed residence is within close proximity to schools, child care centers, playgrounds, or other grounds or facilities where children of similar age or circumstance as a previous victim are present who the department determines may be put at substantial risk of harm by the sex offender‘s residence at that location.” WASH. REV. CODE § 72.09.340(3)(a) (2006). This provision goes on to state that “[t]he department‘s authority under this section is independent of any court-ordered condition of sentence or statutory provision regarding conditions for community custody or community placement.” WASH. REV. CODE § 9.94A.728(2)(d). The majority argues that this sentence makes it clear that the function of section 9.94A.728(2)(d) is “to preserve to the DOC the discretion to deny transfer in the event that it makes one of the four determinations, notwithstanding what other legal sources might otherwise require.” Maj. Op. at 2504. I do not disagree. The fact that the provision permits the DOC to deny release even where other legal sources would allow for it is irrelevant, however, to the question at hand: regardless of other legal sources, does the statute itself place substantive limits on the DOC‘s exercise of discretion? As I explained supra, the provision‘s preservation to the DOC of discretion to deny transfer only “in the event that it makes one of the four determinations[,]” is precisely the type of substantive limitation that gives rise to the liberty interest asserted here.

Washington law defines “sexually violent predator” as “any person who has been convicted of or charged with a crime of sexual violence and who suffers from a mental abnormality or personality disorder which makes the person likely to engage in predatory acts of sexual violence if not confined in a secure facility.” WASH. REV. CODE § 71.09.020(16) (2006). An offender determined to be a sexually violent predator may be subject to civil commitment after his term of confinement. WASH. REV. CODE § 71.09.010 (2006). Carver‘s criminal record and poor behavior while incarcerated provide sufficient basis on which the DOC could reasonably determine that Carver appeared to meet the definition of a sexually violent predator and that it would be appropriate to refer him to be civilly committed for further custodial treatment after he completed his criminal sentence. The legislature of the State of Washington has enacted a civil commitment custodial treatment program run by the Department of Social & Health Services in which dangerous sexual predators may be safely held while efforts are made to treat their condition. See WASH. REV. CODE § 71.09.010. The majority contends that I read too much into the use of the word “liberty interest” in these cases. It first argues that the liberty interest recognized by the Washington Court of Appeals is merely procedural, rather than substantive, in nature. Maj. Op. at 2506 (“[T]hose cases concerned only the procedural right to compliance with individualized consideration on the merits of prisoners’ release plans . . . .“). To the contrary, the Washington court explained: “An inmate‘s interest in his earned early release credits is a limited, but protected, liberty interest. Likewise, the department‘s compliance with requirements of statutes affecting his release is a protected liberty interest.” See (quoting ) (emphasis added). As this language makes clear, the Washington court recognizes a liberty interest in both the substantive right to earned early release (here, in the form of transfer to community custody) and the distinct procedural right to have the DOC comply with the requirements of the statutes governing such release. The majority next suggests that “To the extent [the Washington decisions] contain dicta using the same term, ‘liberty interest,’ to refer to both a substantive right to transfer and a procedural right to consideration on the merits, the most logical reading is that both derive from Washington law, and are hence of the same sub-constitutional nature.” Maj. Op. at 2506. That the Washington court had the federal Due Process Clause—and not a sub-constitutional right—in mind when analyzing section 9.94A.728(2) is, again, evidenced by the plain language of its decision in Liptrap, where the court framed its discussion of the interest at stake in the familiar terms of federal due process jurisprudence. See (explaining that “[d]ue process protects against the deprivation of life, liberty, or property” and finding that “[a]n inmate‘s interest in his earned early release credits is a limited, but protected, liberty interest“).

Subsection (2)(c) reads: “The department shall, as a part of its program for release to the community in lieu of earned release, require the offender to propose a release plan that includes an approved residence and living arrangement.” As the Washington Court of Appeals in held, this language, combined with language in subsection (2)(d), mandates that the DOC make an individualized determination “based on the merits of a release plan.” . The DOC‘s failure to provide such an evaluation rendered unlawful its further detention of sex offenders without an individualized determination, a statutory wrong for which the Washington Rules of Appellate Procedure provide relief. See Wash. R. App. P. 16.4(a), (c)(2) (providing that “the appellate court will grant appropriate relief to a petitioner if . . . the petitioner‘s restraint is unlawful.“). Thus, Dutcher stands for the proposition that release plans may not be categorically denied. It does not require that release plans be categorically approved.

Cashaw, like Dutcher, instead grounded the inmate‘s right to challenge his restraint because of procedural error in Washington Rule of Appellate Procedure 16.4. .

Section 72.09.340(3) contains two narrower circumstances under which “the department shall not approve a residence location“: if the proposed residence includes a minor child who might be put at risk, or if it is close in proximity to the current residence of the prisoner‘s minor victim.

In Allen, the Supreme Court “reject[ed] the argument that a statute that mandates release ‘unless’ certain findings are made is different from a statute that mandates release ‘if,’ ‘when,’ or ‘subject to’ such findings being made.” .

If, for example, the statute stated that DOC “may” deny transfer “only if” certain criteria are met or “unless” they are not, that might sufficiently limit the discretion of the DOC to the point it creates an expectation of release. The distinction between “if” and “only if,” however, is not a mere quibble over vocabulary—it goes right to the heart of whether the criteria of section 9.94A.728(2)(d) are necessary or sufficient conditions for transfer, and therefore whether transfer is mandatory or entirely discretionary. “May . . . only if” would be effectively identical to “shall . . . unless“; “may . . . if” is not.

The canon of construction expressio unius est exclusio alterius stands for the proposition that, when the legislature provides a list of related items, it impliedly means to exclude other items not listed. See Norman J. Singer, 2A Sutherland Statutes and Statutory Construction § 47:23 (7th ed. 2007). Such a result will obtain, however, only in the absence of evidence to the contrary. “The maxim expressio unius est exclusio alterius is an aid to construction, not a rule of law. It can never override clear and contrary evidences of [legislative] intent.” ; see also (“[T]he rule of expressio unius est exclusio alterius d[oes] not necessarily apply without considering other factors which may persuade the court that legislative intent was the opposite of what the statutory construction rule would require.“). In this case, the statement that sex offenders only “may” receive transfer, combined with the final sentence of the paragraph, make clear the permissive intent of the statute.

“Whether a state statute provides such a protectable entitlement depends on the structure and language of the statute, as well as the state courts’ interpretation of the scope of the interest.” (emphasis added).

In re Crowder predated the enactment of section 9.94A.728(2)(d) by three years. There is no indication in subsequent case law, however, that codification of certain reasons in section 9.94A.728(2)(d) has somehow narrowed the range of “legitimacy.”

clarifies that “no expectation of finality can attach during the period in which either party may petition for rehearing.” (quoting ). Thus, until the mandate issues, an opinion is not fixed as “settled Ninth Circuit law,” and reliance on the opinion is a “gamble.” Id. Moreover, in recognition of the human condition, our rules provide that “[i]f a member of a three-judge panel becomes unavailable by reason of death, disability, or departure from the court and the case is under submission, the Clerk shall draw a replacement by lot.” Ninth Circuit General Orders 3.2g.

; ; ; ; ; ; ; ; ; .

The prior concurrence did not note, and there are no, conflicts with other circuits, nor direct conflicts with Supreme Court rulings. As discussed in this opinion, this is an issue of first impression for the Ninth Circuit. Finally, as discussed in the substance of this opinion, the issue in this case is one concerning Washington‘s statutory scheme governing early release into community custody. This is not an issue affecting national uniformity, only one specific to Washington state.

Of the 1097 Petitions for Rehearing En Banc that were filed in 2007, nineteen cases were taken en banc (1.7%).