Kenneth Foster v. Sharee Booker

595 F.3d 353, 2010 U.S. App. LEXIS 3217, 2010 WL 546495

• Court: Court of Appeals for the Sixth Circuit
• Decided: February 18, 2010
• Docket: 08-1371, 08-1372, 08-1626
• Status: Published

Opinion

OPINION

ROGERS, Circuit Judge.

Defendants, the Michigan Parole Board and the Director of the Michigan Department of Corrections, appeal the district court’s grant of summary judgment in favor of plaintiffs on their ex post facto claim. Plaintiffs, inmates in the custody of the Michigan Department of Corrections sentenced to life imprisonment with the possibility of parole for offenses committed before October 1, 1992, brought this § 1983 action to challenge the Michigan Parole Board’s application of post-1992 changes to Michigan’s parole laws to plaintiffs’ parole review. The district court, in determining that application of the changes to plaintiffs violated the Ex Post Facto Clause, relied on statistics showing average annual rates of parole and average years served at parole from 1942 to 2004, as well as evidence that the post-1992 reconstituted Parole Board had curtailed the exercise of its discretion in a wholesale fashion. The district court subsequently entered a permanent injunction and awarded plaintiffs’ costs and attorney fees.

To the extent that plaintiffs face a risk of increased punishment under the post- 1992 parole laws as compared to the laws in effect when they committed their offenses, we cannot conclude that changes to the parole laws caused such an increase. Rather, the new Board’s low rates of parole may be attributable largely to the legitimate exercise of discretion in granting fewer paroles. Even assuming that changes in the number of paroles did not result from a stricter exercise of discretion, it is not clear that plaintiffs have shown a sufficient risk of increased punishment to prevail on their ex post facto claim. Thus plaintiffs were not entitled to judgment as a matter of law, and the summary judgment and grant of injunctive relief must be reversed.

I.

Plaintiffs filed this class action lawsuit against the Michigan Parole Board and the Director of the Michigan Department of Corrections pursuant to 42 U.S.C. § 1983. Plaintiffs allege that changes to Michigan’s parole laws in 1992 and 1999, as implemented and applied retroactively to their parole review, violate the Ex Post Facto and Due Process Clauses of the United States Constitution. The district court granted plaintiffs’ motion for class certification. The parties later agreed to define the class as

[a]ll parolable lifers in the custody of the Michigan Department of Corrections who committed crimes (for which they received a parolable life sentence) before October 1, 1992, and whose parole the “new” parole board has denied, passed over, expressed no interest in pursuing, or otherwise rejected or deferred. Excluded from this definition are so-called “drug lifers” who were convicted of distribution or possession of controlled substances, regardless of whether the crime was one originally subject to parolable life or one converted to parolable life at a later time.

“Parolable lifer” is a term used by the parties and the district court to refer to a “prisoner sentenced to imprisonment for life” for an offense other than first degree murder, first degree criminal sexual conduct, or a few other specific categories of offenses. See Mich. Comp. Laws § 791.234(6)-(7). Thus the plaintiff class generally includes inmates sentenced to life with the possibility of parole, for pre1992 non-drug crimes.

An inmate sentenced to parolable life for a crime committed before October 1, 1992, comes within the Board’s jurisdiction after he or she has served ten calendar years of the life sentence. Mich. Comp. Laws § 791.234(7)(a). Once an inmate comes within the Board’s jurisdiction, the Board may parole the inmate at any time, see id., although “release on parole is discretionary with the parole board,” id. § 791.234(11). At all times relevant to this lawsuit, the statutory discretion has been limited by the requirement that a “prisoner shall not be given liberty on parole until the board has reasonable assurance, after consideration of all of the facts and circumstances, including the prisoner’s mental and social attitude, that the prisoner will not become a menace to society or to the public safety.” Id. § 791.233(1)(a). If the Board decides not to parole an inmate after he or she has served ten years, then the Board reviews the inmate for parole periodically thereafter, until the inmate “is paroled, discharged, or deceased.” Id. § 791.234(8)(b).

If the Board decides to move forward with the parole process, either at the ten-year mark or following an inmate’s subsequent review, then the Board holds a public hearing. Id. § 791.234(8)(c). The Board must give notice of the hearing to the inmate’s sentencing judge (or that judge’s successor in office). Id. If the sentencing judge “files written objections to the granting of the parole within 30 days of receipt of the notice of hearing,” then the process comes to an end. Id. Absent an objection, however, the Board ultimately decides whether to grant parole after holding the public hearing. See id. § 791.234(8)(a)-(d).

Plaintiffs’ ex post facto and due process claims are based on the cumulative effect of statutory changes to the structure and composition of the Board in 1992 and changes to parole procedures in 1992 and 1999. From the Board’s establishment in 1953 until 1992, Board members were “within the state civil service.” 1953 Mich. Pub. Acts No. 232 at p. 413 (§ 32); see also 1982 Mich. Pub. Acts No. 314 at p. 1356 (§ 32(1)). According to one former Director of the Michigan Department of Corrections, serving as a Board member “was like any other civil service job: a member could stay until he or she wanted to leave or was removed for cause.” In 1992, however, the Michigan Legislature repealed the provision that had established the civil service board, 1992 Mich. Pub. Acts No. 181 at p. 1125 (§ 32(3)), and created instead a board of ten members “who shall not be within the state civil service,” Mich. Comp. Laws § 791.231 a(1) (emphasis added); see 1992 Mich. Pub. Acts No. 181 at p. 1124 (§ 31a(1)). This represented a three-member increase in the size of the Board. Compare 1982 Mich. Pub. Acts No. 314 at p. 1356 (§ 32(1)) (“There is established in the department a parole board consisting of 7 members .... ”), with Mich. Comp. Laws § 791.231a(1) (“Beginning October 1, 1992, there is established in the department, a parole board consisting of 10 members .... ”). As of 1992, Board members are limited to four-year terms, although a member may seek reappointment at the end of his or her term. Mich. Comp. Laws § 791.231a(2). The law now also requires that “[a]t least 4 members of the parole board shall be persons who, at the time of their appointment, have never been employed by or appointed to a position in the department of corrections.” Id.

Like the structure and composition of the Board, the procedure for paroling inmates sentenced to parolable life has also changed over time. Until 1982, inmates sentenced to parolable life could expect an initial interview with the Board after having served seven years, with subsequent interviews “at no greater than 36-month intervals following the initial interview,” i.e., a 7+3 + 3 interview structure. MDOC Policy Directive, Parole Board Interview and Decision Criteria, PD-DWA-45.05 (effective Oct. 15, 1982); see MDOC Policy Directive, Parole Board Interview and Decision Criteria, PD-DWA-45.05 (effective Nov. 15, 1980); MDOC Policy Directive, Parole Board Interview and Decision Criteria, PD-DWA-45.05 (effective June 18, 1979). In 1982, the Michigan Legislature amended the law to require the initial interview at the four-year mark with subsequent interviews “biennially thereafter,” i.e., a 4+2+2 interview structure. 1982 Mich. Pub. Acts No. 314 at p. 1356 (§ 34(4)(a)). The Legislature changed the law again a decade later such that, as of 1992, a Board member is not statutorily required to interview an inmate sentenced to parolable life before the inmate comes within the Board’s jurisdiction. Rather, the initial interview is required only after the inmate has served ten years. Mich. Comp. Laws § 791.234(8)(a); see 1992 Mich. Pub. Acts No. 181 at p. 1127 (§ 34(4)(a)). Moreover, as of 1992, the inmate could expect to be reinterviewed as infrequently as every five years, i.e., a 10+5 + 5 interview structure, not every two or three years as had been the previ ous practice. 1992 Mich. Pub. Acts No. 181 at p. 1127 (§ 34(4)(a)).

In 1999, the Legislature eliminated the statutory reinterview requirement altogether. As a result, as of 1999, a Board member need only interview an inmate sentenced to parolable life after the inmate has served ten years. See Mich. Comp. Laws § 791.234(8)(a); 1999 Mich. Pub. Acts No. 191 at p. 1076 (§ 34(6)(a)). Interviews take place “thereafter as determined by the” Board. Mich. Comp. Laws § 791.234(8)(a). Rather than requiring regular reinterview of an inmate, the statute now requires the Board to review the inmate’s paper file at five-year intervals. Id. § 791.234(8)(b); see 1999 Mich. Pub. Acts No. 191 at p. 1076 (§ 34(6)(b)). At oral argument, however, the Board’s counsel stated that, in practice, the Board exercises its discretion to reinterview an inmate every ten years following the initial ten-year interview.

Legislative changes in 1999 also curtailed an inmate’s right to judicial review of the denial of parole. The Legislature had made an inmate’s right to appeal explicit in 1982 by providing that the “action of the parole board in granting or denying a parole shall be appealable to the circuit court.” 1982 Mich. Pub. Acts No. 314 at p. 1357 (§ 34(5)). Before 1982, the law had provided only that the Board’s action of releasing an inmate was “not ... renewable if in compliance with law.” 1953 Mich. Pub. Acts No. 232 at p. 414 (§ 34); see 1958 Mich. Pub. Acts No. 211 at p. 281 (§ 34); 1978 Mich. Pub. Acts No. 81 at p. 219 (§ 34(5)). As a result of the 1999 amendments, only the prosecutor or the victim of an inmate’s crime has a statutory right to appeal the Board’s decision to grant parole. Mich. Comp. Laws § 791.234(11); see 1999 Mich. Pub. Acts No. 191 at p. 1077 (§ 34(9)).

Plaintiffs also allege that in the 1990s the Board stopped providing written reasons to explain its lack of interest in moving an inmate forward to a public hearing. See MDOC Policy Directive, Parole Board Interview and Decision Criteria, PD-DWA-45.05 (effective Feb. 10, 1986) (setting forth Board’s practice of “preparing] a written summary of [Board’s] action on all parole denial cases”). This change appears to have been within the statutory discretion of the Board. Since 1982, Michigan law has required that “[w]hen the parole board makes a final determination not to release a prisoner, the prisoner shall be provided with a written explanation of the reason for denial.” Mich. Comp. Laws § 791.235(12); see 1982 Mich. Pub. Acts No. 314 at p. 1358 (§ 35(10)). The Michigan Court of Appeals in 2001, by interpreting “final determination” to mean determinations that had progressed through all the steps in the parole eligibility process, refused to require a written explanation at the “no interest” stage. Gilmore v. Parole Bd., 247 Mich.App. 205, 635 N.W.2d 345, 357-58 (2001). While plaintiffs argue that a “subtle change” in the 1999 statutory amendments, dealing with the requirement of a public hearing, 1 provided the basis for the Board’s policy change in this regard, the Gilmore court did not mention, much less rely upon, that statutory change.

In sum, the statutory amendments that provide the basis for plaintiffs’ ex post facto challenge (1) altered the structure and composition of the Board; (2) reduced the frequency of parole reviews after an initial ten-year interview; (3) substituted paper reviews for in-person interviews; (4) eliminated plaintiffs’ right to appeal a denial of parole; and (5) contained new language consistent with the Board’s practice of not giving written reasons for a statement of “no interest” in moving forward with parole. Plaintiffs allege that the Board’s retroactive application of the 1992 and 1999 changes to the parole laws creates a significant risk that plaintiffs will face greater punishment than they would have faced under the law in place at the time they committed their offenses.

The Board initially moved to dismiss plaintiffs’ complaint for failure to state a claim upon which relief can be granted. The district court granted in part and denied in part the Board’s motion. See Foster-Bey v. Rubitschun, No. 05-71318, 2005 WL 2010181 (E.D.Mich. Aug. 18, 2005). Most significant for purposes of this appeal, the district court granted the Board’s motion with regard to plaintiffs’ due process claim on the ground that plaintiffs do not have a constitutionally protected liberty interest in parole. Id. at *6. Moreover, to the extent that plaintiffs’ sentencing judges assumed that parole procedures would not change over time, this assumption was improper and, therefore, did not amount to a factual error in sentencing that violated due process. Id. The district court denied the Board’s motion with respect to plaintiffs’ ex post facto claim, id. at *7, and concluded that this court’s decision in Shabazz v. Gabry, 123 F.3d 909 (6th Cir.1997), did not preclude plaintiffs’ ex post facto challenge, Foster-Bey, 2005 WL 2010181, at *5. In Shabazz, a more comprehensive plaintiff class lost a facial challenge to the constitutionality of the 1992 change to a 10 + 5 + 5 interview structure. The district court determined that, because the Michigan Legislature had made additional changes to the parole laws in 1999 (after this court had decided Shabazz), the cumulative effect of the 1992 and 1999 changes to Michigan’s parole laws could not have been litigated in Shabazz. Id. at *5 & n. 10. According to the district court, a “state cannot continuously make minor changes in the parole process that, taken together, create a sufficient risk of an increased penalty; but, when looked at alone, would not violate the Ex Post Facto Clause.” Id. at *5 n. 10.

Following discovery, the parties cross-moved for summary judgment on plaintiffs’ ex post facto claim. On October 23, 2007, the district court granted plaintiffs’ motion and denied the Board’s motion. See Foster-Bey v. Rubitschun, No. 05-71318, 2008 WL 7020690 (E.D.Mich. Oct.23, 2008). The district court determined that the discretionary nature of the Board’s decision whether to grant parole did not preclude plaintiffs’ ex post facto challenge. Id. at *9. After considering the cumulative effect of the challenged amendments to the parole laws, id. at *10, the district court concluded that the “change in the make-up of the Michigan Parole Board, the Board’s understanding of why the change occurred and how it was to exercise its discretion, its redefining of the eligibility procedure for [inmates sentenced to parolable life], and changes to the timing and intervals of the interview and review process, when considered in total have significantly disadvantaged the class and constitute a violation of the Ex Post Facto Clause,” id. at *23.

More specifically, the district court found that the pre-1992, or “old,” Board had treated inmates serving parolable life sentences and inmates serving long indeterminate sentences the same to ensure that inmates who had committed similar crimes — and who had been similarly rehabilitated — would serve similar time in pris on. Id. at *13-14. The new Board, by-contrast, took the position that “life means life” and, accordingly, treated inmates sentenced to parolable life differently than inmates serving long indeterminate sentences. Id. at *14-16. In addition, members of the new Board focused on the seriousness of an inmate’s offense as the most relevant factor in deciding whether to grant parole. Id. at *16-17. The district court found that this “myopic view of the relevant factors” led to the “erosion of the substantive standard” for granting parole. Id. Furthermore, the district court found that, in practice, the new Board does not review the file of an inmate sentenced to parolable life any more frequently than every five years, no matter the inmate’s individual circumstances. Id. at *17-18. The district court also determined that “the value of the [in-person] interview,” all but eliminated from the parole process in 1999, “cannot be overstated.” Id. at *19.

The district court relied in large part on statistical evidence in concluding that, as a result of the 1992 and 1999 changes to the parole laws, plaintiffs faced a sufficient risk of increased punishment to prevail on their ex post facto claim. Id. at *19-23. According to the district court, “parole rates from 1942 until 1984 show that [inmates sentenced to parolable life] were paroled at a steady 5-15 percent rate, with the average time served steady at 15-18 years.” Id. at *21 (internal quotation marks omitted). The district court reported that the new Board’s parole rate was, by contrast, just 0.15%; meanwhile, the “average number of years served climbed from 19.2 during 1995-99, to 23.2 during 2000-04.” Id. The district court assigned credit for the “pipeline paroles” — paroles of inmates whom the old Board, near the end of its tenure, moved forward to public hearings but for whom the decision whether to grant parole ultimately rested with the new Board — to the old Board rather than the new Board. Id. at *20-21. As a result, the district court refused to credit the new Board with any paroles during the transition period from 1992 to 1994; thus, for purposes of the district court’s analysis, the new Board effectively began its tenure in 1995. Id. at *21. The district court also declined to consider the new Board’s post-2004 numbers, because plaintiffs filed their class action complaint in April 2005. Id.

Importantly, the district court excused the marked decrease in the old Board’s parole rates from 1985 to 1994 because the decrease was not due to a change in the parole laws. Id. at *23. Rather, the decrease was attributable to a number of other causes, including (1) the governor’s nine-time invocation of the Prison Overcrowding Emergency Powers Act, which, with each invocation, moved forward by ninety days the release dates for inmates with minimum sentences, thereby increasing the number of inmates eligible for parole and “ ‘burfying]’ the Board with new cases”; (2) an overall increase in the prison population and, in particular, an increase in the number of inmates sentenced to parolable life in the mid-1970s (these inmates were just coming within the Board’s jurisdiction by the mid-1980s); and (3) a 1982 statutory amendment that required more frequent parole interviews for inmates sentenced to parolable life — as of 1982, these inmates had a 4 + 2+2 interview structure. Id. at *21-23.

After granting plaintiffs’ motion for summary judgment, the district court solicited proposed remedial orders from both parties. On February 7, 2008, the district court entered a declaratory judgment “holding that the defendants have been and remain in violation of the ex post facto clause of the U.S. Constitution.” The district court also entered a permanent injunction with terms the court deemed the least restrictive it could impose to cure the Board’s ex post facto violation. The district court’s injunctive order required the Board to compile a list of the longest-serving first quartile of the plaintiff class “as soon as practicable” and to interview those plaintiffs within five months of the entry of the order. The injunctive order further instructed the Board to “apply the parole laws, policies, procedures, and standards that were applied by the old parole board in the decades before 1992,” to “the extent possible,” and listed a number of sources the Board should consult for guidance. The district court ordered the Board to report back with the results of the first-quartile interviews within six months and to have compiled a list of second-quartile plaintiffs by that time.

The Board has since submitted two progress reports for the district court’s review. On November 3, 2009, in response to the Board’s second progress report, the district court supplemented its original injunctive order by imposing additional reporting requirements. The district court’s supplemental order also required the Board to make an inmate’s complete psychological records available for review online by all Board members before they vote whether to move the inmate forward to a public hearing.

Subsequent to the entry of the initial remedial order, the district court awarded plaintiffs’ costs and attorney fees. The Board now appeals the district court’s grant of plaintiffs’ motion for summary judgment, entry of the declaratory judgment and permanent injunction, and award of plaintiffs’ costs and attorney fees. Plaintiffs cross-appeal the dismissal of their due process claim.

After filing notice of their appeal, the Board twice moved this court to stay the original injunctive order and all further district court proceedings during the pendency of the appeal. A motions panel of this court denied both motions. Following oral argument, however, and in response to the district court’s November 3, 2009, supplemental order, the Board again moved to stay the original injunctive order and all further district court proceedings. This panel granted the Board’s motion with respect to the November 3, 2009, supplemental order but denied the Board’s motion in all other respects. Thus the permanent injunction has remained in effect and the Board has released a number of plaintiffs on parole during the pendency of this appeal. 2 Indeed, as of November 2009, three of four quartiles of the plaintiff class had been or were currently going through the review process.

II.

A. Ex Post Facto Claim

Plaintiffs have not shown that they face a significant risk of increased punishment as a result of the challenged statutory changes to Michigan’s parole process rather than as a result of the new Board’s legitimate exercise of discretion in a way that results in fewer paroles. In any event, we are not confident that plaintiffs have even shown a significant risk of increased punishment under the post-1992 parole regime. Accordingly, plaintiffs were not entitled to judgment as a matter of law.

The Ex Post Facto Clause is “aimed at laws that ‘retroactively alter the definition of crimes or increase the punish ment for criminal acts.’ ” Cal. Dep’t of Corr. v. Morales, 514 U.S. 499, 504, 115 S.Ct. 1597, 131 L.Ed.2d 588 (1995) (citations omitted). When an inmate challenges an allegedly ex post facto parole law, this court “must examine the relevant law in effect at the time [the inmate’s] offense was committed and compare it with the retroactively-applied version of the law.” Shabazz, 123 F.3d at 912. The “focus of the ex post facto inquiry is not on whether a legislative change produces some ambiguous sort of ‘disadvantage,’ ... but on whether [the] change ... increases the penalty by which a crime is punishable.” Morales, 514 U.S. at 506 n. 3, 115 S.Ct. 1597. Where a legislative change is argued to increase the risk of affecting a prisoner’s punishment, the Supreme Court has made clear that the appropriate inquiry is whether the retroactively-applied version of the parole law “produces a sufficient risk of increasing the measure of punishment attached to the covered crimes.” Id. at 509, 115 S.Ct. 1597. When the retroactively-applied version of the law “does not by its own terms show a significant risk” of increased punishment, the inmate bringing the ex post facto challenge “must demonstrate, by evidence drawn from the [law]’s practical implementation by the agency charged with exercising discretion, that [the law’s] retroactive application will result in a longer period of incarceration than under the earlier” version of the law. Garner v. Jones, 529 U.S. 244, 255, 120 S.Ct. 1362, 146 L.Ed.2d 236 (2000).

To the extent that plaintiffs have shown they face a significant risk of increased punishment under the new parole regime, plaintiffs have not shown that this risk is attributable to statutory changes to the parole process and not to a change in the way the Board legitimately exercises its discretion. The decision whether to grant parole has always been within the Board’s discretion. See Mich. Comp. Laws § 791.234(11). Therefore, from the time plaintiffs committed their offenses, there was always the possibility the Board would exercise its discretion in a way that would result in fewer paroles and longer prison terms. As the Michigan Court of Appeals stated in People v. Hill, 267 Mich.App. 345, 705 N.W.2d 139, 143 (2005), “[t]here was always a ‘significant risk’ that [plaintiffs] would be made to serve [their] life sentence[s].”

Throughout the time period relevant to this suit, the statutory scope of the Board’s range of discretion has remained the same. Parole is within the Board’s discretion, but a “prisoner shall not be given liberty on parole until the board has reasonable assurance, after consideration of all of the facts and circumstances, including the prisoner’s mental and social attitude, that the prisoner will not become a menace to society or to the public safety.” Mich. Comp. Laws § 791.233(1)(a). Despite the fact that the scope of the Board’s discretion has remained the same, plaintiffs argue that, in practice, the new Board applied a harsher standard than the old Board when deciding whether to grant parole. However, plaintiffs’ contentions do not make out an ex post facto violation.

If the Parole Board decided within its discretion to get tougher, that could hardly amount to an ex post facto violation as long as it was within the Parole Board’s discretion to get tougher. This would be true even if the tougher attitude resulted from a change in personnel on the Board, even if the Board members independently developed a tougher attitude, and even if the Board members partook of a general public attitude that parole decisions should be tougher. By analogy, there is no ex post facto violation if a lenient judge is replaced by a strict one in a particular jurisdiction, such that a certain crime, say shoplifting or drunk driving, now regularly gets a sentence well above what was previously the case. Without some legal change other than a difference in the proper exercise of discretion, the shoplifter or the drunk driver simply had no ex post faeto-protected interest in the more lenient sentence, as long as the more severe sentence was within the range available to the sentencing judge at the time of the crime.

Plaintiffs’ arguments that the new Board members acted in a tougher exercise of their retained discretion thus cut against finding an ex post facto violation. Plaintiffs point to evidence that changes to the parole laws in 1992 were motivated by the desire to increase public safety by granting fewer paroles. Indeed, a report issued by the Michigan Department of Corrections in 1997 suggests that, when the Legislature “overhaul[ed]” the Board and the parole process itself in 1992, the “intent ... was to make Michigan’s communities safer by making more criminals serve more time and keeping many more locked up for as long as possible.” Mich. Dep’t of Corr., Five Years After: An Analysis of the Michigan Parole Board Since 1992, at 2 (1997). The report also asserts that one of “the most important differences since the overhaul is a Parole Board that is much less willing to release criminals who complete their minimum sentences — and much less willing to release criminals at all, forcing many to serve their maximum sentences.” Id.

Plaintiffs contend that the new Board, which by statute must include at least four members with no experience in the Department of Corrections, Mich. Comp. Laws § 791.281a(2), placed greater emphasis on an inmate’s underlying offense and less emphasis on the inmate’s rehabilitation when deciding whether to grant parole. Plaintiffs submitted the sworn statements of several members of the new Board to support this contention. For example, Gary Gabry, a former county prosecutor who chaired the new Board from 1992 to 1996, recalls:

I often found myself trying to get the focus off the crime and onto the candidate’s recent record in prison. I pushed the board to focus more on the prisoner’s behavior, adjustment and future plans and not primarily the sentencing offense, but it nearly always fell on deaf ears with at least five members of the board.

Gabry believes that after the addition of “school teachers and other professionals” to the Board, “it was just much harder for a prisoner to get out on parole.” In Gabry’s opinion, “the drop-off in ... paroles ... was largely a reflection of the new type of people who were appointed to the parole board in 1992.”

Plaintiffs argue that the new Board’s low rates of parole “can only be attributed to its ‘life means life’ policy.” As evidence of the existence of this policy, plaintiffs quote public statements by and on behalf of the new Board. For example, when Stephen Marschke, who served as a member (1992-1996) and then as chair (1996-2002) of the new Board, submitted written testimony in support of proposed changes to the parole laws in 1999, he stated:

It has been a long[-]standing philosophy of the Michigan Parole Board that a life sentence means just that — life in prison.... It is the parole board’s belief that something exceptional must occur which would cause the parole board to request the sentencing judge or Governor to set aside a life sentence. Good behavior is expected and is not in and of itself grounds for parole.

Marschke Test, in Support of Proposed Legis. ¶ 2.

At bottom, plaintiffs’ evidence shows that the new Board exercised its discretion rigorously, resulting in fewer paroles than under the old Board. As the United States Supreme Court has explained in the clearest terms,

[T]o the extent there inheres in ex post facto doctrine some idea of actual or constructive notice to the criminal before commission of the offense of the penalty for the transgression, we can say with some assurance that where parole is concerned discretion, by its very definition, is subject to changes in the manner in which it is informed and then exercised. The idea of discretion is that it has the capacity, and the obligation, to change and adapt based on experience. New insights into the accuracy of predictions about the offense and the risk of recidivism consequent upon the offender’s release, along with a complex of other factors, will inform parole decisions.

Garner, 529 U.S. at 253, 120 S.Ct. 1362 (citations omitted). The most that can be said here is that, based on experience, the new Board’s discretion was informed and then exercised in a way that made it more difficult for plaintiffs to secure release on parole.

“[T]he Ex Post Facto Clause gives [an inmate] no cause to complain that the Board in place at the time of his offense has been replaced by a new, tough-on-crime Board that is much more parsimonious with parole____” Id. at 259, 120 S.Ct. 1362 (Scalia, J., concurring). Plaintiffs conceded at oral argument that there would be no ex post facto violation if the new Board had gradually changed the way it exercises discretion or if Board members had been replaced one by one over time. There is, however, no reason to distinguish between gradual change and wholesale change for purposes of this ex post facto analysis. Nor is there a constitutional requirement that changes in the exercise of discretion be made over time. Thus plaintiffs’ allegations pertaining to the composition of the new Board do not establish an ex post facto violation.

With respect to plaintiffs’ loss of the right to appeal a denial of parole, any harm resulting from this change to the law is too speculative to contribute to the alleged risk of increased punishment in any significant way. From 1995 to 1999, inmates filed 3,800 appeals from Board decisions, but only about four percent of those cases were remanded to the Board for reconsideration, and only about six-tenths of a percent ultimately resulted in parole. Jackson v. Jamrog, 411 F.3d 615, 620 (6th Cir.2005) (quoting House Legislative Analysis, First Analysis, H.B. 4624 at 9 (Mich. Mar. 21, 2000)). As the Board points out, appeals by inmates sentenced to parolable life likely represented only a fraction of those percentages. Plaintiffs counter that the right to appeal is “important” because “just one successful case can change the law for all prisoners.” Although plaintiffs’ statement is arguably true, it does not change the facts. Because the loss of the right to appeal “creates only the most speculative and attenuated possibility of producing the prohibited effect of increasing the measure of punishment” attached to plaintiffs’ crimes, Morales, 514 U.S. at 509, 115 S.Ct. 1597, this change in the law does not contribute in a significant way to the risk of increased punishment alleged as the basis for the ex post facto claim.

Plaintiffs have not shown that the remainder of the statutory and regulatory changes to the parole process — i.e., the change to a 10 + 5 + 5 interview structure, the substitution of paper reviews for in-person interviews, and the loss of written reasons to explain a statement of “no interest” in moving forward with parole — as practically implemented and applied retroactively, create a sufficient risk of increased punishment to prevail on their ex post facto claim. First, this court decided previously that the change from a 7+3+3 or a 4 + 2 + 2 interview structure to a 10 + 5 + 5 interview structure does not on its face create a sufficient risk of increased punishment to establish an ex post facto violation. See Shabazz, 123 F.3d at 914-15. Second, and more clearly dispositive, to the extent that plaintiffs have shown a decrease in the average annual rate of parole or an increase in the average years served at parole since the new Board took office, this court cannot isolate the cause of those effects on the record in this case. In other words, there is no way for this court to determine whether any decrease in the parole rate or any increase in the average years served is due to the challenged statutory changes to the parole process or to the Board’s stricter exercise of its discretion. Indeed, plaintiffs’ ex post facto claim is explicitly premised on the cumulative effect of changes to parole policies and procedures in 1992 and 1999 — including changes in the way Board members exercise their discretion. It is not enough for plaintiffs to show that “a legislative change produces some ambiguous sort of ‘disadvantage.’ ” Morales, 514 U.S. at 506 n. 3, 115 S.Ct. 1597. And plaintiffs cannot distinguish the effect of the statutory changes to the interview process from the effect of the Board’s more rigorous, yet still legitimate, exercise of its discretion. Therefore, plaintiffs were not entitled to summary judgment on their ex post facto claim.

In any event, we are not even confident the statistics show a significant risk of increased punishment under the post-1992 parole laws as compared to the laws in effect when plaintiffs committed their offenses. As a preliminary matter, the parties disagree as to how this court should measure the effect of changes to the parole laws. Plaintiffs assert that the average annual rate of parole is the relevant measure, and they compare the new Board’s rates to the old Board’s rates in an effort to show that the statutory changes created a significant risk of increased punishment. In contrast to plaintiffs, the Board argues that the annual number of paroles is the proper measure and that plaintiffs’ ex post facto claim is unfounded because the new Board’s number of paroles (47) for the ten-year period from 1995 to 2004 was greater than the old Board’s number of paroles (28) for the ten-year period from 1983 to 1992.

In fact, the data suggest that the real difference between the tenures of the two Boards was in the number of inmates sentenced to parolable life who had served ten years and could thus be considered for parole. This number provides the denominator for plaintiffs’ parole rate calculations. According to the data plaintiffs provided in their brief on appeal, see Pls.’ Br. 35, 41-42, 3 this number has increased significantly since the early 1980s:

Average Annual Number of Inmates Sentenced to Parolable Life and Subject Date/Period to Consideration for Parole

2000-04_1454 4

New Board 1995-99_1091

1990-92_740

1985-89_474

1980-84_141.0

1975-79_63.2

1970-74_72,8

1965-69_ 84.6

1960-64_125.4

1955-59_169.0

1950-54_184,2

1945-49_221.0

Old Board 1942-44_199/7

Despite the increasing number of inmates sentenced to parolable life who could be paroled, according to the figures plaintiffs provided in their brief, Pis.’ Br. 41-42, 5 the new Board’s parole rates did not differ significantly from the old Board’s rates during the final years of its tenure:

Average Average Annual Rate Years Served Date/Period of Parole

2000-04 0,15%_23.2

New Board 1995-99 0.20%_19.2

1990-92_0.18% 6 _unknown

Old Board 1985-89_0fi%_15.1

These data suggest that the practical implementation of changes to the parole laws posb-1992 did not result in a parole rate significantly lower than the rate that had resulted from the practical implementation of the old parole laws by the mid-1980s. In other words, these numbers indicate that plaintiffs did not face a significant risk of increased punishment under the new parole laws as compared to the old parole laws, as those laws were implemented and applied in practice.

Plaintiffs provide a number of excuses for the old Board’s low rates during the last seven years of its tenure, but do not acknowledge that the new Board inherited some of the same problems that had plagued the old Board. Plaintiffs explain the old Board’s low rates by pointing to an overall increase in the prison population as a result of tougher criminal laws, reasoning that “[a]s the prison population mushroomed, the parole board began to fall behind in its work.” Plaintiffs also attribute the old Board’s low parole rates to (1) the governor’s repeated invocation of the Prison Overcrowding Emergency Powers Act, which, with each invocation, moved up by ninety days the release dates for inmates sentenced to a minimum term of years, thereby increasing the number of possible paroles (i.e., increasing the denominator), see Oakland County Prosecuting Attorney v. Mich. Dep’t of Corr., 411 Mich. 183, 305 N.W.2d 515, 517 & n. 6 (1981); (2) more frequent interviews with inmates sentenced to parolable life under the 4+2+2 interview structure after 1982; and (3) an increase in the number of inmates sentenced to parolable life in the mid-1970s who were just becoming eligible for parole in the mid-1980s but who, according to plaintiffs, were not yet ready for parole. William Hudson, who served as a member (1980-1985) and then as chair (1985-1991) of the old Board, remembers:

It is fair to say that the board was overwhelmed by the numbers at some point, and that we had to put our energy and resources into interviewing prisoners who were most likely to be paroled. Lifer interviews got pushed back, and even when we did lifer interviews, it was more to comply with the law, and not with an eye to moving anyone forward to parole, because we were so far behind in our work. In the best of circumstances we kept just marginally abreast of the regular parole cases, and no doubt in the mid-to-late 1980s and early 1990s the lifers suffered for it.

Although the new Board had the benefit of three additional members and a 10 + 5 + 5 interview structure, when it took office it faced the same overwhelming numbers the old Board had faced near the end of its tenure. The new Board did not enter the environment that had prevailed from the 1940s to the 1970s, when the average annual number of inmates sentenced to parolable life who were eligible for parole rarely exceeded 200 and annual parole rates were relatively high. Rather, the new Board stepped into the environment inhabited by the old Board in the mid-1980s and early 1990s, by which time the old Board’s parole rates had decreased significantly.

Plaintiffs have not shown that any decrease in the parole rate or any increase in the average years served under the new Board was not attributable to the sheer number of inmates within the new Board’s jurisdiction. And, again, the new Board’s parole rates did not differ significantly from the old Board’s rates during the final years of its tenure. This suggests that plaintiffs did not actually face a significant risk of increased punishment as a result of the new Board’s implementation of post-1992 changes to the parole laws.

While our rejection of plaintiffs’ ex post facto claim relies primarily on the impossibility of distinguishing the effect of the Board’s legitimate exercise of discretion from the effect of statutory changes to the parole process on the record before this court, our conclusion is bolstered by some doubt, outlined above, that the new Board’s parole rates were in fact significantly lower than the old Board’s rates during its final years in office.

We reverse the district court’s grant of summary judgment in plaintiffs’ favor and vacate the permanent injunction. Because plaintiffs are no longer the prevailing party, we also reverse the district court’s award of plaintiffs’ costs and attorney fees.

B. Due Process Claim

Plaintiffs’ cross-appeal in this case involves a due process claim raised in the complaint as an alternative to the ex post facto claim. The due process claim was dismissed at an early point in the litiga tion, primarily on the ground that Michigan prisoners have no constitutionally protected liberty interest in parole. The dismissal was proper, although the reasons for that conclusion depend on which of several theories serves as the basis for the due process claim.

The due process claim could be taken as a challenge to the sufficiency of parole procedures under general procedural due process principles applicable when liberty or property interests are deprived by the government. The complaint at ¶ 117 alleges that the changes to Michigan’s parole standards and policies “have deprived the plaintiffs of meaningful parole review.” A different type of due process theory would be analogous to ex post facto jurisprudence, protecting against punishment that could not have been anticipated when the crime was committed, but where ex post facto protections technically do not apply because the increased punishment does not result from a promulgated statute or regulation. A third theory would be that constitutional due process precludes the imposition of punishment greater than that actually imposed or contemplated by the sentencing judge.

The complaint does not state a claim under any of these theories. First, considered as a claim under general procedural due process principles with regard to the denial of parole, the district court properly dismissed on the ground that plaintiffs lack a constitutionally protected liberty interest in parole. See Foster-Bey, 2005 WL 2010181, at *6 (citing Sweeton v. Brown, 27 F.3d 1162, 1164-65 (6th Cir. 1994) (en banc)). Plaintiffs do not contest this, but rather argue that this is not the nature of their claim.

We do not read the complaint, however, as raising the second possible type of due process theory — that the Due Process Clause serves to extend ex post facto protections to situations where punishment is tougher than the perpetrator had reason to expect at the time of the crime, but not because of any promulgated law or regulation. The due process allegations in the complaint are not phrased in those terms. In any event, such a theory is not supported in the law. The district court, in denying the motion to dismiss the ex post facto claim, relied on the allegations that laws and regulations had caused the alleged changes in parole eligibility. The district court in doing so recognized that the ex post facto claim would otherwise have to be dismissed, relying on the following excerpt from our holding in Shabazz:

The Ex Post Facto Clause exists to protect citizens from retroactive increases in punishment. Changes in the administration and enforcement of statutes have little impact on these public expectations .... [T]he internal memoranda and policy directives ... are not published in the Michigan Administrative Code or presented to the public for comment. Accordingly, the Parole Board and the MDOC’s policies and directives did not likely influence public expectations as to parole or create a reliance interest in the public on a particular parole hearing schedule.

123 F.3d at 916. If unpublished policy statements do not implicate ex post facto concerns in the parole context, it follows of necessity that pure changes in the exercise of discretion do not do so. Further, there is no apparent reason that such limits on the applicability of the Ex Post Facto Clause should not govern an application of the Due Process Clause that would protect the identical underlying interest.

The complaint appears instead to address a different concern — not that the perpetrator could not be aware of the severity of punishment for his act, but that a convicted defendant should not be pun ished more than the sentence provides. That this is the gravamen of plaintiffs’ due process claim is reflected in ¶ 115 of the complaint: “The changes to Michigan’s substantive parole standards and to the state’s parole laws and policies, described above, could not have been foreseen by the state court trial judges when they sentenced the plaintiffs.” This theory, however, is not supported in the law on the facts alleged in the complaint.

We assume for the sake of argument that the Due Process Clause would prohibit the imposition of punishment beyond limits explicitly imposed by the sentencing court, at least in the absence of some further infraction. But that is not what is alleged in the complaint. Instead, the complaint alleges the imposition of punishment within a range explicitly imposed, but beyond the subjective expectation of the sentencing judge. We are unaware of any authority that would make this a due process violation.

The closest cases cited by plaintiffs in this connection are different because they reflect post-sentencing changes that conflict with explicit limits in the judge’s sentence. The Seventh Circuit on direct appeal in United States v. Kerley, 838 F.2d 932, 941 (7th Cir.1988), remanded for re-sentencing, noting that the trial judge “may have sentenced Kerley to a longer term in prison than the judge realized,” but the crux of the ruling was that the trial judge had incorrectly stated at the sentencing hearing that the time served could be less than one year, when the sentencing statute mandated a sentence of at least one year. In Cutter v. United States, 241 F.Supp.2d 19, 22 (D.D.C.2003), the court relied on that fact that “the Court’s sentence was premised on an explicit understanding of where petitioner would serve her time.” Plaintiffs’ complaint does not allege that such explicit limits were contained in the sentences imposed on plaintiffs.

For these reasons, plaintiffs’ alternative due process claim fails, and the district court properly dismissed it.

III.

For the foregoing reasons, the district court’s grant of plaintiffs’ motion for summary judgment is reversed and this case is remanded to the district court with instructions to grant the Board’s motion for summary judgment. Accordingly, we vacate the permanent injunction and reverse the award of plaintiffs’ costs and attorney fees. We also affirm the district court’s dismissal of plaintiffs’ due process claim.

1 . As of 1999, the statute provides that a “decision to grant or deny parole to [a] prisoner [sentenced to parolable life] shall not be made until after a public hearing,” Mich. Comp. Laws § 791.234(8)(c); see 1999 Mich. Pub. Acts No. 191 at p. 1076 (§ 34(6)(c)), whereas the predecessor provision had stated that a "parole shall not be granted a prisoner so sentenced until after a public hearing,” 1992 Mich. Pub. Acts No. 181 at p. 1127 (§ 34(4)(b)).

2 . At oral argument, the Board's counsel stated unequivocally on behalf of the state defendants that if this court were to reverse the district court's grant of plaintiffs' motion for summaiy judgment, plaintiffs paroled pursuant to the district court’s original injunctive order would not have their parole revoked solely on the basis of this court’s reversal.

3 . The support for these numbers in the record appears to come from Pls.' Mot. for Summ. J. Ex. 20, Attach, to Decl. at 3 and Pls.’ Mot. for Summ. J. Ex. 52 at 9.

Both in the district court and on appeal, plaintiffs presented their data as five-year averages. Plaintiffs claim that "[fjive-year averages give a more accurate picture, offsetting year-to-year fluctuations.” Pls.’ Br. 35. Although we make no judgment with respect to this assertion, because our analysis is based on the data plaintiffs provided in their brief, and for which we find support in the record, we also analyze the data in five-year intervals. The interval spanning the early 1990s is shortened to exclude data from 1993 to 1994, during the transition from the old Board to the new Board.

4 . Our own calculations, based on the record, suggest that this number should be 1498.8. See Pls.' Mot. for Summ. J. Ex. 52 at 9.

5 . The support for these numbers in the record appears to come from Pls.’ Mot. for Summ. J. Ex. 52 at 9 and Defs.' Mot. for Summ. J. Ex. 3 at 5-8.

6 . Plaintiffs assert that the average annual rate of parole from 1990 to 1992 was 1.9%. Pis.' Br. 41. However, to arrive at this number, plaintiffs included in their calculations — as actual paroles — thirty-nine inmates sentenced to parolable life whom the old Board, near the end of its tenure, moved forward to public hearings and who were not blocked by judicial veto. Id. at 39-40. The decision whether to parole these inmates ultimately rested with the new Board after it took office. Indeed, the new Board paroled, at most, only fifteen of these thirty-nine inmates in 1993 and 1994. Defs.' Mot. for Summ. J. Ex. 3 at 6. Plaintiffs argue that the old Board should receive credit for paroling the thirty-nine inmates, even though not all were actually paroled, and even though none were actually paroled by the old Board, because "the old board rarely denied parole after moving forward on a case." Pls.' Br. at 40. Thus "all or nearly all would have been paroled if the old board had not been removed.” Id. (emphasis added).

Although the parties dispute whether the old Board or the new Board should receive credit for inmates paroled in 1993 and 1994, we need not decide the issue. This table, like the previous table, excludes data from these transition years.

We calculated 0.18% as the average annual rate of parole from 1990 to 1992 by summing the number of inmates sentenced to parolable life who were actually paroled from 1990 to 1992 — four, Defs.' Mot. for Summ. J. Ex. 3 at 6 — and then dividing by the number of years (three) to get the average annual number of paroles (1.33). We then divided that number by the average annual number of inmates sentenced to parolable life who could be considered for parole — 740, according to plaintiffs’ brief, Pls.’ Br. 41; see Pls.’ Mot. for Summ. J. Ex. 52 at 9.