KEARSE, Circuit Judge:
Defendants United States Parole Commission and its chairman and members (collectively the “Commission”) appeal from a judgment of the United States District Court for the District of Connecticut, T.F. Gilroy Daly, Chief Judge, permanently enjoining the Commission to comply with procedures consistent with those specified by this Court in Drayton v. McCall, 584 F.2d 1208 (2d Cir.), (“Drayton”), rev’g, 445 F.Supp. 305 (D.Conn.1978), for parole rescission proceedings. 636 F.Supp. 101. On appeal, the Commission contends that decisions of the Supreme Court and this Court since Drayton have eroded the validity of Drayton’s holdings (1) that an inmate whose early effective release date has been set has a liberty interest protectable under the Due Process Clause of the Fifth Amendment to the Constitution, and (2) that due process requires that, to the extent consistent with institutional safety, such an inmate be given the rights to be represented by counsel, to call witnesses, to confront and cross-examine adverse witnesses, and to have a de novo hearing concerning certain factual issues, before his early release date may be rescinded. We conclude that post-Drayton decisions of the Supreme Court and this Court have not undermined Drayton’s analysis, and, accordingly, we affirm the judgment of the district court.
I. BACKGROUND
Under the federal parole system as it presently exists, see 18 U.S.C. §§ 4201-4218 (1982) (repealed effective Nov. 1, 1987, Pub.L. No. 98-473, tit. II, §§ 218(a)(4), 235(a)(1), 98 Stat. 2027, 2031 (1984), as amended by Pub.L. No. 99-217, § 4, 99 Stat. 1728 (1985)), the Commission conducts an initial parole hearing for most prisoners shortly after their incarceration, 18 U.S.C. § 4208(a), following which the Commission may set an “effective date of parole,” i.e., a date not more than six months away, 28 C.F.R. §§ 2.1(h), 2.12(b) (1986), or set a presumptive parole date as much as 15 years away, id. § 2.12(b), or schedule a hearing for reconsideration after 15 years, id. If an “effective date of parole” (which we also refer to as an “early release date”) is not set at the initial hearing, it may be set at a subsequent hearing. 18 U.S.C. § 4208(h); 28 C.F.R. § 2.14 (1986).
Plaintiffs Theodore Green and Daniel Porter commenced this suit, on behalf of themselves and others similarly situated, as federal prisoners incarcerated within the District of Connecticut whose early release dates had been set by the Commission but who had not yet been released from prison (“parole grantees”). At issue in this case are the hearing procedures to be followed when, after an early release date has been set but prior to the arrival of that date, the Commission is notified that the parole grantee has been guilty of new misconduct or receives other new unfavorable information and wishes to consider whether to rescind the early release date.
[286] A. Drayton and its District Court Predecessors
In 1972, the Supreme Court ruled that a parolee, i.e., one already released from prison on parole, had certain due process rights that must be respected before his parole could be revoked. These included the right to written notice of claimed parole violations, disclosure of the evidence against him, and a hearing before a neutral body such as a traditional parole board at which he could be heard, present witnesses, and confront and cross-examine adverse witnesses. See Morrissey v. Brewer, 408 U.S. 471, 92 S.Ct. 2593, 33 L.Ed.2d 484 (1972). The Court extended these requirements to probation revocation proceedings and added a requirement that, within the responsible agency’s discretion, the parolee was entitled to representation by counsel. Gagnon v. Scarpelli, 411 U.S. 778, 93 S.Ct. 1756, 36 L.Ed.2d 656 (1973).
Thereafter, in a series of habeas corpus petitions filed in the District of Connecticut, a number of parole grantees, contending that they were entitled to the same protections as parolees, alleged that their due process rights were being violated by the Commission’s parole rescission procedures. In the first such case, Williams v. United States, 383 F.Supp. 402 (D.Conn.1974), the district court held that the due process rights accorded in parole revocation hearings were also required in parole rescission hearings; these included advance notice of the hearing, the assistance of counsel, and the opportunity to call witnesses and to confront and cross-examine adverse witnesses. Id. at 405. In at least five subsequent cases, the court found that the Commission had denied each petitioner the due process rights outlined in Williams. See, e.g., Green v. Nelson, 442 F.Supp. 1047, 1058-60 (D.Conn.1977) (finding that the Commission had provided Theodore Green, one of the named plaintiffs in the present action, with a constitutionally inadequate parole rescission hearing and ordering his release on parole), and cases cited therein; Porter v. Wilkinson, No. B-77-245 (D.Conn. Mar. 1, 1978) (granting relief to Daniel Porter, the other named plaintiff in this case, on the same grounds). None of these decisions was appealed by the Commission.
In 1978, when the district court in Dray-ton held that the Commission had once again violated a parole grantee’s due process rights by conducting a parole rescission hearing without the requisite procedural protections, 445 F.Supp. at 311, the Commission appealed. As discussed in greater detail in Part II below, this Court approved the district court’s holding that a parole grantee has a protectable liberty interest entitling him to procedural due process in parole rescission hearings, 584 F.2d at 1214-18, but concluded that not all of the procedures required for parole revocation hearings were required for parole rescission hearings. Id. at 1218-19. Instead, in order to balance the parole grantee’s interest against the appropriate governmental interests, id. at 1219, we ruled that the rights to have the evidence disclosed, to testify, and to present evidence need not be accorded if to do so would threaten institutional safety; and we ruled that the right to confront and cross-examine adverse witnesses should be recognized only as to such witnesses as are willing to testify, id. at 1220.
B. The Decision Below
In 1978, prior to our decision in Drayton, Green and Porter filed the present class action seeking, inter alia, an injunction requiring the Commission to follow the procedures found constitutionally mandated in the district court’s prior cases. The district court entered a preliminary injunction ordering the Commission to provide constitutionally adequate procedures as required in its Williams line of decisions.
Final judgment was entered in March 1986, permanently enjoining the Commission to follow procedures consistent with those set forth in Drayton. The court rejected the Commission’s contention that injunctive relief was unnecessary because the Commission had changed its written policies governing parole rescission hearings, finding that many of the Commission’s written policies were not always implemented. The court found, as one exam-[287] pie, that in connection with 21 of 44 parole rescission hearings held within the district between February 1983 and February 1985, the Commission had failed to send the parole grantee the advance notice of hearing mandated by both the preliminary injunction and Commission regulation, 28 C.F.R. § 2.34(a). Because of this and other evidence that the Commission was not fully complying with the ordered procedural requirements, the court permanently enjoined the Commission to give the parole grantee, inter alia, (1) written notice of action scheduling the hearing, describing the charges, and stating the grantee’s procedural rights; (2) disclosure of documents pertaining to the charges, unless this would threaten institutional or personal safety; (3) a de novo hearing before a neutral and detached hearing body; (4) the right to counsel; (5) an opportunity to be heard in person and to call witnesses, unless this would threaten institutional safety; (6) the right to confront and cross-examine adverse witnesses, unless this would threaten institutional safety; and (7) a written statement of the results of the hearing.
This appeal by the Commission followed.
II. DISCUSSION
On appeal, the Commission urges that we reverse the judgment of the district court, contending principally that postDrayton developments in the law establish that (1) a parole grantee has no protectable liberty interest, and (2) even if the parole grantee does have such an interest, its protection does not require implementation of several of the procedures ordered by the district court. We reject both contentions.
A. The Existence of a Protectable Liberty Interest
The Commission argues that certain Supreme Court decisions, principally that in Greenholtz v. Inmates of Nebraska Penal & Correctional Complex, 442 U.S. 1, 99 S.Ct. 2100, 60 L.Ed.2d 668 (1979) (“Greenholtz”), and decisions of this Court in the wake of Greenholtz, have in effect overruled Drayton, and that in light of these more recent authorities, we must conclude that a parole grantee has no protectable liberty interest. We disagree.
In Drayton, this Court explored the circumstances under which the Commission is permitted to rescind an early release date it has set for a parole grantee. Looking at the Commission’s own regulations, we found the Commission’s rescission authority limited to “two narrowly circumscribed” sets of factual circumstances:
First, the Commission may reconsider its grant of parole when the grantee has been found guilty of institutional misconduct____ Second, reconsideration is authorized when new information adverse to the prisoner and unrelated to prison misconduct is discovered, 28 C.F.R. § 2.34(b) [1977], such as a prison[er]’s willful concealment or misrepresentation of information. Id. § 2.30 [1977]____
584 F.2d at 1215. We concluded that since the Commission, by these regulations, had limited its own rescission authority to these two carefully defined categories, a prisoner who had been granted an early release date had a justifiable expectation of achieving his liberty on that date. We noted that the limitations on the Commission’s authority to rescind an early release date distinguished Drayton’s case from cases such as Meachum v. Fano, 427 U.S. 215, 96 S.Ct. 2532, 49 L.Ed.2d 451 (1976), and Montanye v. Haymes, 427 U.S. 236, 96 S.Ct. 2543, 49 L.Ed.2d 466 (1976), which had held that state prisoners had no protectable liberty interest in remaining in their original or preferred places of confinement in light of the “unbridled discretion of state prison officials,” Drayton, 584 F.2d at 1214, to transfer prisoners. Accordingly, we held that a parole grantee has a protectable liberty interest that entitles him to due process in the Commission’s parole rescission hearings.
Though the regulations in effect when Drayton was decided have since been modified, the pertinent changes have not been substantial. The Commission’s rescission authority remains limited, as 28 C.F.R. § 2.34(a) (1986) provides that the Commission may reconsider the parole grantee’s early release date if the parole grantee has [288] been found to have violated institutional rules or is alleged to have committed a new criminal act, and 28 C.F.R. § 2.28(f) (1986) authorizes reconsideration when the Commission has received “new and significant adverse information.” Thus, unless Dray-ton has been undermined by later cases, its ruling that a parole grantee is entitled to due process in a rescission hearing remains the law of this Circuit.
We see nothing in Greenholtz or the other cases relied on by the Commission to undermine Drayton’s liberty interest analysis. Greenholtz dealt with state prisoners who had not been given release dates; thus, the Supreme Court’s focus was on (1) those prisoners’ rights to have the parole board set parole dates for them, and (2) the procedures to which they were entitled in connection with the board’s initial decision. The questions on which the Court granted the certiorari petition of the state parole officials were
whether the Due Process Clause of the Fourteenth Amendment applies to discretionary parole-release determinations made by the Nebraska Board of Parole, and, if so, whether the procedures the Board currently provides meet constitutional requirements.
442 U.S. at 3, 99 S.Ct. at 2102. The question ultimately decided by the Court was whether three new procedures ordered by the Eighth Circuit, which we discuss in Part II.B. below, were constitutionally required. Id. at 14-16 & n. 6, 99 S.Ct. at 2107-08 & n. 6.
The Court began by analyzing the nature of the interest at stake. It stated that, in order to have a protectable liberty interest, a prisoner must have more than a hope or a unilateral expectation of release. “ ‘He must, instead, have a legitimate claim of entitlement to it.’ ” Id. at 7, 99 S.Ct. at 2104 (quoting Board of Regents v. Roth, 408 U.S. 564, 577, 92 S.Ct. 2701, 2709, 33 L.Ed.2d 548 (1972)). Comparing the legitimate expectations of the prisoners before it with those of persons already released on parole, the Court found important differences. First, it saw “a crucial distinction between being deprived of a liberty one has, as in parole, and being denied a conditional liberty that one desires.” Id. at 9, 99 S.Ct. at 2105. Second, it saw an important difference in the nature of the decision that must be made in each case. Noting that “ ‘[t]he first step in a revocation decision ... involves a wholly retrospective factual question,’ ” id. (quoting Morrissey v. Brewer, 408 U.S. at 479, 92 S.Ct. at 2599), it found significant contrast in the initial parole release decision, which is “more subtle and depends on an amalgam of elements, some of which are factual but many of which are purely subjective appraisals by the Board members based upon their experience with the difficult and sensitive task of evaluating the advisability of parole release.” 442 U.S. at 9-10, 99 S.Ct. at 2105. Thus, the Greenholtz Court viewed the initial decision of whether to set a parole date not as essentially fact-bound but rather as “turnpng] on a ‘discretionary assessment of a multiplicity of imponderables, entailing primarily what a man is and what he may become rather than simply what he has done.’ ” Id. at 10, 99 S.Ct. at 2105 (quoting Kadish, The Advocate and the Expert — Counsel in the Peno-Correctional Process, 45 Minn.L.Rev. 803, 813 (1961)).
The Court concluded that to the extent that a prisoner sought to have the parole board establish a parole date, “the general interest asserted here is no more substantial than the inmate’s hope that he will not be transferred to another prison, a hope which is not protected by due process.” 442 U.S. at 11, 99 S.Ct. at 2105 (citing Meachum v. Fano and Montanye v. Haymes). Nonetheless, the Court “accepted]” the Nebraska prisoners’ view that the language and structure of the Nebraska statute gave them a more substantial interest in having a parole date set than would be created by most state statutes. Id. at 12, 99 S.Ct. at 2106. Noting that it had held in Wolff v. McDonnell, 418 U.S. 539, 94 S.Ct. 2963, 41 L.Ed.2d 935 (1974), that a “good-time credit” statute created a liberty interest protected by due process against arbitrary loss because the only prerequisite to the earning of such credits was good behavior, the Greenholtz Court ac[289] cepted the argument that the Nebraska parole statute similarly gave the prisoners a legitimate expectancy in the board’s setting a parole date because the statute provided that the board “shall” order the prisoner’s release “unless” it be of the opinion that his release should be deferred. Though the statute allowed such a deferral for any of a number of discretionary reasons, the Court noted that there had been no interpretation by the Nebraska courts of the scope of the interest the statute was intended to afford the inmates, and it found acceptable the prisoners’ contention that the “shall ... unless” language was sufficient to create a liberty interest that was entitled to some measure of protection. 442 U.S. at 11-12, 99 S.Ct. at 2106. See also Board of Pardons v. Allen, — U.S. -, 107 S.Ct. 2415, 96 L.Ed.2d 303 (1987) (protectable liberty interest created by Montana statute requiring that a prisoner “shall” be released when certain conditions are met).
Given both the factual differences between Greenholtz and Drayton and the Greenholtz Court’s analytical framework, we think Greenholtz supports, rather than undermines, Drayton’s conclusion that parole grantees have a protectable liberty interest. First, though the liberty interest of a federal parole grantee is not as substantial as that of a parolee, it is far more substantial than that of the Greenholtz prisoners. We perceive a continuum that includes the liberty interests attributable to, in descending order, the parolee, the parole grantee, and the inmate without a parole date (“nongrantee”). Since the Greenholtz Court, and more recently the Court in Board of Pardons v. Allen, accepted the proposition that inmates for whom no parole date had been set could have some protectable interest, we are hard pressed to believe that that Court would not also find that a protectable interest is possessed by an inmate whose release date has already been set and is less than six months away.
Further, the Greenholtz Court’s analysis of the types of factors considered in reaching decisions affecting various levels of parole interest supports Drayton. The nature of the decision to be made with respect to the parole grantee is more similar to a parole revocation decision than to an initial parole decision, for the conditions impeding the parole grantee’s achievement of actual liberty are, in the first instance, specific and fact-bound, not, as is true of the nongrantee, dependent on considerations lying principally within the Commission’s discretion. Thus, rescission of the already granted early release date may not occur unless the grantee has been found guilty of institutional misconduct, or has committed a new criminal act, or is the subject of new adverse information received by the Commission. Exploration of these potentially impeding conditions involves retrospective determinations of factual matters, unlike most of Greenholtz’s “purely subjective appraisals by the Board members based upon their experience with the difficult and sensitive task of evaluating the advisability of parole release.” 442 U.S. at 10, 99 S.Ct. at 2105; see also id. at 14, 99 S.Ct. at 2107; id. at 15-16, 99 S.Ct. at 2107-08 (Nebraska board’s decision against early release is not like “a guilt determination”).
• In sum, both the concreteness of the parole grantee’s liberty expectation and the objective nature of the findings that must be made before that expectation may be eliminated are characteristics that, under Greenholtz, must be viewed as supporting the existence of a protectable liberty interest. We conclude, therefore, that Greenholtz does not undermine the liberty interest conclusions of Drayton.
We find no greater merit in the Commission’s invocation of the Supreme Court’s decisions in Jago v. Van Curen, 454 U.S. 14, 102 S.Ct. 31, 70 L.Ed.2d 13 (1981) (per curiam), and Connecticut Board of Pardons v. Dumschat, 452 U.S. 458, 101 S.Ct. 2460, 69 L.Ed.2d 158 (1981) {“Dumschat”), in support of its contention that Drayton has been overtaken. In Jago v. Van Curen, the Court ruled that an Ohio state prisoner who had been granted an early parole date had no protectable liberty interest. But the basis for that ruling was the fact that the Ohio law unambiguously left [290] any early release “wholly within the discretion of the” Ohio parole authorities and allowed that decision to be exercised at any time until the prisoner’s actual release. 454 U.S. at 16, 102 S.Ct. at 83. The wholly discretionary nature of the Ohio parole board’s authority to rescind an early parole date can hardly be equated with the quite limited authority of the Commission to rescind an early release date previously granted to a federal prisoner.
Dumschat is no more compelling. There, state prisoners argued that, based on the statistical frequency with which the Connecticut Board of Pardons had in the past commuted life sentences, they had a protectable interest in similar treatment. The Dumschat Court rejected this argument, noting that the Connecticut statute had “no definitions, no criteria, and no mandated ‘shalls’.” 452 U.S. at 466, 101 S.Ct. at 2465. The Court refused to establish any sort of “unwritten common law” that would give rise to a protectable interest. See id. at 465, 101 S.Ct. at 2464.
The Second Circuit decisions relied on by the Commission also fail to reveal any basis for disturbing the liberty interest analysis of Drayton. Both Boothe v. Hammock, 605 F.2d 661 (2d Cir.1979), and Berard v. State of Vermont Parole Board, 730 F.2d 71 (2d Cir.1984), involved (1) prisoners who had had no release date set, (2) statutes that did not use the Nebraska “shall ... unless” structure found persuasive in Greenholtz, and (3) decisions that were wholly discretionary. Pugliese v. Nelson, 617 F.2d 916 (2d Cir.1980), involved the parole board’s authority with regard to inmate classification, another wholly discretionary matter.
We have considered all of the Commission’s arguments in support of its challenge to Drayton’s conclusion that a parole grantee has a liberty interest that is protectable by due process and have found them unpersuasive. Drayton’s liberty interest analysis remains valid in light of the intervening case law.
We turn now to the Commission’s challenges to Drayton’s conclusions as to what procedures are required for the protection of the parole grantee’s liberty interest.
B. The Process Due
The Commission argues that even if a parole grantee has a protectable liberty interest, existing Commission procedures are constitutionally adequate to protect that interest, and the court should not have ordered it, in addition, to (1) allow the parole grantee to be represented by counsel, (2) allow the parole grantee to call favorable witnesses, (3) allow confrontation and cross-examination of adverse witnesses, and (4) grant a de novo hearing into the disciplinary violations charged. These challenged procedures, tempered, as described in Part I.B. above, by respect for institutional concerns, are among those that the Drayton Court found necessary for the protection of the parole grantee’s liberty interest. To the extent that the Commission contends that Drayton was simply wrongly decided, we decline to entertain the argument, since our function is not to review Drayton de novo; if the Commission wished to overturn Drayton it should have sought review in 1978 by the Supreme Court. To the extent that the Commission argues instead that the Dray-ton procedures should no longer be required in light of the Supreme Court’s later decisions in Greenholtz and Jago v. Van Curen, we disagree for the reasons stated below.
The Commission’s reliance on Jago is flawed for the reason discussed in Part II.A. above. The Court found that there was no liberty interest because the Ohio parole board had complete discretion to rescind an early parole date; its authority, unlike that of the Commission, was not restricted. There being no cognizable liberty interest in Jago, no due process protections came into play.
Nor do we see anything in the holding, rationale, or tenor of Greenholtz that undermines Drayton’s rulings as to the process that is due a parole grantee, for the two cases differed not only as to what liberty interests were at stake, as discussed above, but also as to what procedures were [291] at issue. The Greenholtz Court did not purport to establish what procedures might or might not be required for the protection of even the very limited liberty interest at stake in Greenholtz; nor did it rule on or express a view as to any of the Drayton procedures at issue here.
The Greenholtz Court’s discussion of what procedures were required to protect the liberty interest of the Nebraska prisoners centered on three procedures, ordered by the lower courts, that were particularly challenged by the parole board, and its holding was that these three — (1) a formal hearing for every inmate, (2) the inclusion in every adverse decision of a statement of the evidence relied on, and (3) written advance notice listing the factors the board might consider — were not constitutionally required. See 442 U.S. at 14-16 & n. 6, 99 S.Ct. at 2107-08 & n. 6. The Court declined to express an opinion as to issues that had been decided adversely to the prisoners; these included two of the Drayton procedures involved in the present case, i.e., the right to call favorable witnesses and the right to cross-examine adverse witnesses. See id. at 16 n. 8, 99 S.Ct. at 2106 n. 8. Further, another of the Drayton procedures at issue here — the prisoner’s right to be represented by counsel of his choice — was not at issue in Greenholtz because state law already granted the prisoner that right. Id. at 5, 99 S.Ct. at 2105. Finally, the right to a de novo hearing could not have been at issue in Greenholtz since there had not been any prior decision with regard to these prisoners other than the convictions that sent them to prison in the first place. And though there were two stages of the Nebraska release date determination process, it is difficult to characterize the Eighth Circuit’s order of a “full formal hearing,” 442 U.S. at 6, 99 S.Ct. at 2103, as calling for a de novo determination even at the second stage, since that court stated that the prisoner was not to have the right to call witnesses in his own behalf or to cross-examine adverse witnesses, id. at 16 n. 8, 99 S.Ct. at 2108 n. 8. Accordingly, none of the Dray-ton procedures at issue in the present case was at issue in Greenholtz, and the Greenholtz holdings cannot be directly applied here.
Nor, in light of the Court’s rationale, does the Greenholtz ruling that the three procedures under scrutiny were not constitutionally required for parole nongrantees have negative implications for Drayton’s. analysis of what process is due parole grantees. The two concerns voiced by the Greenholtz Court were (1) that principles of federalism not lead the federal courts to impose such onerous procedural burdens on the states that the states would elect to abolish the parole system, and (2) that the parole evaluation proceedings not be turned into adversary ones that might impede the goal of inmate rehabilitation. The first does not require extended discussion, and neither undermines Drayton.
Plainly the Court’s concern for federalism has no bearing on the current validity of Drayton, since Drayton and the present case involve the federal system itself, not state systems. Further, the Greenholtz Court’s caveat that disregard of federalism might have the undesirable effect of leading the states to eliminate their parole systems is ironic given the fact that the federal parole system is about to be eliminated.
Nor does the Greenholtz warning against turning the Nebraska process into one that is essentially adversarial, or one that focuses on the prisoner’s guilt of the offense that led to his incarceration, undermine Drayton’s analysis, given the parole grantee’s more concrete liberty interest and the indefeasibility of that interest in the absence of a determination of new facts adverse to him. Unlike the parole grantee, the Greenholtz prisoner with no parole date has not been accused of any wrongdoing other than commission of the offense that led to his incarceration. With respect to these nongrantees, there is little fact-finding for the parole board to do, 442 U.S. at 9-10,15 n. 7, 99 S.Ct. at 2104-05, 2107 n. 7, and therefore the “[procedures designed to elicit specific facts, such as those required in Morrissey, Gagnon, and Wolff, are not necessarily appropriate to a Nebraska parole determination,” 442 U.S. at 14, 99 S.Ct. at 2107. Thus, the Greenholtz [292] admonition against introducing adversarial procedures was made against the background of the Court’s repeated observations that the Nebraska parole decision was
necessarily subjective in part and predictive in part. Like most parole statutes, [the Nebraska statute] vests very broad discretion in the Board.
442 U.S. at 13, 99 S.Ct. at 2107; see also id. at 16, 99 S.Ct. at 2108:
The Nebraska statute contemplates, and experience has shown, that the parole-release decision is, as we noted earlier, essentially an experienced prediction based on a host of variables.
These admonitions have little import for Drayton, for as we have discussed, the federal parole grantee has already passed the subjective stage, has had his early release date set, and must be accused of specific wrongdoing or confronted with specific adverse information in order for the Commission to consider rescission of that date. Rather, the circumstances of the parole grantee are more similar to those of the prisoners in Wolff v. McDonnell, 418 U.S. 539, 94 S.Ct. 2963, 41 L.Ed.2d 935, whose earned good-time credits could be taken away only if they were guilty of new misconduct; the context of the inquiry in such a matter is innately adversarial. The Drayton Court, in fashioning the procedures required for the fact-specific inquiry into the new charges against the federal parole grantee, relied on Wolff as its “major point of reference,” 584 F.2d at 1219. We regard the Greenholtz Court’s emphasis on the discretionary nature of the Nebraska parole decision and its contrasting view of what procedures are required in a fact-specific inquiry as in Wolff as confirming both the dissimilarity between .the Greenholtz prisoners and federal parole grantees and the appropriate treatment of the latter in a manner consistent with Wolff.
Finally, to the extent that the Greenholtz Court’s analysis implicitly reflected the Wolff Court’s concern, 418 U.S. at 561-63, 94 S.Ct. at 2977-78, that the courts not impose requirements that would likely disrupt institutional discipline, we note that the Drayton Court, in fashioning procedures to protect the rights of the parole grantees, took painstaking care to protect important institutional interests. Thus, the Drayton Court, “[sjtriking an accommodation by taking cognizance of the previous liberty interest of a parole grantee while seeking to ensure against future prison disorder,” 584 F.2d at 1219, limited the rights of the grantee to have the evidence disclosed, to testify, and to call favorable witnesses, noting that those rights need not be accorded if to do so would threaten institutional safety. For the same reasons, the Drayton Court limited the right to confront and cross-examine adverse witnesses to cases in which those witnesses were willing to testify. Id. at 1220.
To summarize, Greenholtz did not involve the procedures at issue here, or as strong a liberty interest as is at issue here, or a procedural context as inherently adversarial as that at issue here. Viewed in context, the Court’s general admonitions do not suggest that Drayton has been undermined, and the Court’s express recognition of the difference between Wolff and Greenholtz confirms our view of the dis-positive differences between Drayton and Greenholtz. We conclude that Greenholtz in no way undermines Drayton's analysis of what process the parole grantee is due. As the injunction fashioned by the district court was consistent with Drayton, it was proper.
C. Inmates for Whom the Procedures Are Required
Plaintiffs appear to assume that the judgment in this case requires the Commission to follow the specified procedures not only for parole grantees but also for inmates for whom the Commission has set only a presumptive parole date as much as 15 years away (“presumptive parolees”), not an early release date. The judgment does not expressly require this and there are valid reasons why it should not do so.
First, the liberty interest of the presumptive parolee occupies a lower place in the inmate-liberty-interest continuum than does the interest of a parole grantee. Not only [293] is the presumptive parolee’s release expectation temporally more remote, it is also subject to greater control by the Commission. The release of the presumptive parolee is not conditioned just on the absence of any new unfavorable facts but is also “contingent upon an affirmative finding by the Commission” that his conduct has been good and that he has an acceptable plan for the period following his release. 28 C.F.R. § 2.12(d) (emphasis added). Thus, the right of the presumptive parolee is significantly more conditional than that of the parole grantee, and the focus of the additional condition may ordinarily be far less fact-specific than is the Commission’s focus when charges of specific misconduct or adverse information have been leveled against the parole grantee. The district court did not purport to explore these differences or to analyze whether the same procedures that Drayton ordered for parole grantees would be needed to protect the interests of presumptive parolees. The district court simply relied on Drayton as establishing required procedures.
In light of the district court’s reliance on Drayton and the absence of any analysis with regard to the presumptive parolee, we conclude that the court did not intend its injunction to reach Commission proceedings involving presumptive parolees in addition to parole grantees, for Drayton plainly focused only on those for whom early release dates had been set. It described such an inmate as “ha[ving] the taste of freedom in his mouth, the smell of freedom in the air, the touch of freedom within his grasp.” 584 F.2d at 1219. In contrast, under the regulations in effect when Drayton was decided, a presumptive parole date could be as much as four years away, see 28 C.F.R. § 2.12(c) (1977), and Drayton did not purport to deal with these more remote expectations. Certainly Drayton could not have had in mind today’s presumptive parolee, whose presumptive date of release under current regulations may be as much as 15 years away.
Lastly, it appears that all of the plaintiffs in this series of proceedings, from Williams onward, have been parole grantees, not merely presumptive parolees. Plainly the two named plaintiffs here, Green and Porter, were parole grantees; and in light of the greater liberty interest of parole grantees, plaintiffs’ adequacy to represent a class of inmates who were merely presumptive parolees is somewhat suspect.
Accordingly, we construe the judgment as requiring the Commission to provide the specified procedures for parole grantees and as not governing the rights of presumptive parolees, and we affirm it as thus construed.
D. The Propriety of Injunctive Relief
Finally, we address briefly an argument made by the Commission in the district court though not so clearly pursued here, namely that to the extent that the Commission has in place regulations that require procedures ordered by the district court, no injunction was necessary. We think the district court was well within the bounds of its discretion to order injunctive relief.
First, we note that over a span of some four years in the 1970’s at least seven suits were brought to compel the Commission to observe due process rights in connection with parole rescission hearings. The Commission lost each case in the district court. In the first six it did not appeal, but it also did not respect parole grantees’ rights in succeeding cases before it. Apparently it was the practice of the Commission’s hearing examiners, pursuant to Commission regulations, “to ignore these constitutional rulings.” Drayton, 584 F.2d at 1211.
Second, the district court found that even after a preliminary injunction had been issued in the present case, the Commission failed in many instances to follow the ordered procedures. In some instances the failure represented also a failure to follow procedures required by the Commission’s own regulations.
In light of the Commission’s past conduct, we conclude that injunctive relief was entirely appropriate.
[294] CONCLUSION
The judgment of the district court enjoining the Commission to follow the specified procedures for parole grantees is in all respects affirmed.