RYAN, J., delivered the opinion of the court, in which DAUGHTREY, J., joined. WELLFORD, J. (pp. 566-67), delivered a separate concurring opinion.
OPINION
RYAN, Circuit Judge.
After over 20 years of litigation, the defendants in this prisoners’ rights case sought to terminate federal court jurisdiction over Michigan’s efforts to achieve parity between male and female inmates in educational, vocational, apprenticeship, and work-pass opportunities, as well as access to courts. Following a thorough evidentia-ry hearing, the district court found that the defendants’ remedial efforts had succeeded in achieving parity. We conclude that the district court’s findings are not clearly erroneous and, therefore, terminate federal court jurisdiction over these matters.
I.
This case involves the effort of female prison inmates in Michigan to obtain educational, vocational, apprenticeship, and work-pass opportunities, as well as access to courts, comparable to those opportunities available to male inmates. The litigation is now over 20 years old and its course has been chronicled in numerous opinions from both the district court and this court. For a thorough history of the case, we invite the reader’s attention to our opinion in Glover v. Johnson, 138 F.3d 229 (6th Cir.1998). We now recite only those matters necessary to an understanding of the issues before us.
A.
In 1977, two separate groups of plaintiffs brought class actions on behalf of all female prison inmates in Michigan, pursuant to 42 U.S.C. § 1983. The plaintiffs claimed that the Michigan Department of Corrections (MDOC) and various MDOC officials violated the Equal Protection Clause of the Fourteenth Amendment of the federal Constitution with regard to educational and vocational programming and violated their First Amendment right of access to the courts. The district court consolidated the two cases and ruled in favor of the plaintiffs in 1979. Since then, the district court has continued to exercise jurisdiction over the case, and perforce, over a significant aspect of the Michigan corrections system in order to monitor compliance with its subsequent remedial orders.
[559]*559In December 1993, after almost 15 years of district court oversight, the defendants moved to terminate the district court’s jurisdiction pursuant to Fed.R.Civ.P. 60(b)(5) or, in the alternative, Fed.R.Civ.P. 60(b)(6). The district court denied the motion in 1995, finding that the defendants had failed to comply substantially with its remedial orders and various other remedial plans. Glover v. Johnson, 879 F.Supp. 752 (E.D.Mich.1995).
In March 1998, we vacated the district court’s judgment and remanded the matter with the following order:
Within 120 days following issuance of this opinion, the district court shall conduct hearings and receive evidence, including stipulations by the parties, in order to determine with particularity the educational, vocational, apprenticeship, and work-pass opportunities presently being provided (1) to male inmates and (2) to female inmates in the Michigan correctional system. The district court will then make particularized findings of fact and conclusions of law determining whether the male and female inmates are presently being provided sufficiently comparable education, vocational, apprenticeship, and work-pass opportunities as to satisfy the requirements of the Equal Protection Clause of the Fourteenth Amendment. In undertaking this task, the district court must take into account the present conditions of custody and population size at various institutions; any differences in educational and vocational interests between male and female inmates; available educational and vocational training resources; and such other considerations as the district court may deem appropriate.
Glover, 138 F.3d at 243.
In our 1998 opinion, we recognized that the district court had already set forth the defendants’ obligation to achieve parity in the treatment of male and female inmates. In 1979, the district court defined “parity of treatment” to mean:
Defendants here are bound to provide women inmates with treatment and facilities that are substantially equivalent to those provided the men[,] i.e., equivalent in substance if not in form unless their actions ... bear a fair and substantial relationship to achievement of the State’s correctional objectives.
Glover v. Johnson, 478 F.Supp. 1075, 1079 (E.D.Mich.1979), quoted in Glover, 138 F.3d at 241. That 1979 judgment has never been appealed.
We also directed the district court to determine whether female -inmates were denied their First Amendment right of access to courts. We retained jurisdiction with regard to all these matters. Glover, 138 F.3d at 254.
Following discovery, the district court held eight days of evidentiary hearings in January and February 1999, and then issued a comprehensive opinion. Glover v. Johnson, 35 F.Supp.2d 1010, 1012 n. 4 (E.D.Mich.1999). The court first described the facilities comprising the Michigan prison system as of October 1998. MDOC houses approximately 38,000 male inmates and 1,800 female inmates. Male and female inmates are housed in separate facilities. Over 35,000 of the male inmates live in 37 major male facilities, while close to 3,000 live in 12 male camp facilities. Approximately 1,400 female inmates are housed in the two major female facilities, Florence Crane (“Crane”) and Scott Correctional (“Scott”), while approximately 400 are housed in the single female camp facility, Camp Branch.
The district court found that the evidence demonstrated sufficient parity in the educational, vocational, apprenticeship, and work-pass opportunities afforded female and male inmates. Id. at 1017. The court’s opinion did not address the access to courts issue because the parties “agreed to settle that issue based on the Sixth Circuit’s resolution of the Knop v. Johnson and Hadix v. Johnson appeals, as well as [the district court’s] unappealed rulings [560]*560and orders in those cases.” Id. at 1012. While deferring to this court’s retention of jurisdiction in the case, the district court found that termination of its jurisdiction would be proper.
B.
Because we did retain jurisdiction over this matter, we invited the parties to submit supplemental briefings and heard oral argument on whether the district court erred in reaching its findings of parity. The plaintiffs raise the following four issues in challenging the district court’s judgment: (1) whether the district court erred in applying the “reasonable relationship” standard under Turner v. Safley, 482 U.S. 78, 107 S.Ct. 2254, 96 L.Ed.2d 64 (1987) or, in the alternative, whether the court failed to properly apply the four-factor Turner test; (2) whether the court improperly shifted the burden to the plaintiffs to establish current equal protection violations; (3) whether the district court’s factual findings of parity in educational, vocational, and apprenticeship programming were clearly erroneous; and (4) whether termination of the district court’s jurisdiction over the access to courts issue is premature.
Because the plaintiffs conceded before the district court that MDOC had achieved parity in work-pass opportunities, we do not consider that issue here. See Glover, 35 F.Supp.2d at 1024.
II.
We review the district court’s findings of fact for clear error and its conclusions of law de novo. Board of County Comm’rs of Shelby County v. Burson, 121 F.3d 244, 247 (6th Cir.1997), cert. denied, 524 U.S. 1114, 118 S.Ct. 1047, 140 L.Ed.2d 111 (1998). With respect to the district court’s finding of parity in the equal protection context, we review for clear error. See Bannum, Inc. v. City of Louisville, Ky., 958 F.2d 1354, 1359 (6th Cir.1992); Sullivan v. City of Cleveland Heights, 869 F.2d 961, 963 (6th Cir.1989) (per curiam). “ ‘A finding is “clearly erroneous” when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and .firm conviction that a mistake has been committed.’ ” Sullivan, 869 F.2d at 963 (quoting United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S.Ct. 525, 92 L.Ed. 746 (1948)).
Level of Scrutiny and Burden of Proof
We first address the plaintiffs’ argument that the district court applied the wrong level of scrutiny in evaluating whether female inmates’ educational, vocational, and apprenticeship opportunities satisfy the requirements of the Equal Protection Clause. This is an issue of law, which we review de novo.
When it originally considered the plaintiffs’ equal protection claims in the late 1970s, the district court applied heightened scrutiny pursuant to Craig v. Boren, 429 U.S. 190, 97 S.Ct. 451, 50 L.Ed.2d 397 (1976). The district court interpreted our 1998 remand opinion to require it “to measure the results of defendants’ remedial efforts against the standard of parity that the Equal Protection Clause presently requires.” Glover, 35 F.Supp.2d at 1012 n. 2 (some emphasis added). The court concluded that Turner, 482 U.S. 78, 107 S.Ct. 2254, 96 L.Ed.2d 64 — which mandates a lower level of scrutiny for at least some constitutional claims in the prison setting — supersedes Craig v. Boren and should now apply in this case. Glover, 35 F.Supp.2d at 1013. Turner holds that a regulation that “impinges on inmates’ constitutional rights ... is valid if it is reasonably related to legitimate penological interests.” Turner, 482 U.S. at 89, 107 S.Ct. 2254. The following four factors are relevant to the reasonableness determination under Turner: (1) whether there is a rational connection between the regulation and a legitimate governmental interest; (2) whether the inmates have alternative [561]*561means for exercising their constitutional rights; (3) whether accommodation of the constitutional rights will have a significant impact on guards, other inmates, and prison resources; and (4) whether there are alternative approaches that would accommodate the inmates’ constitutional rights with little impact on valid penological interests. Id. at 89-91,107 S.Ct. 2254.
The plaintiffs, citing United States v. Virginia, 518 U.S. 515, 116 S.Ct. 2264, 135 L.Ed.2d 735 (1996), contend that the district court should have applied heightened scrutiny rather than the Turner reasonable relation standard. We note, as did the district court, that Turner dealt with regulations that interfered with inmates’ First Amendment rights and the right to marry; there was no equal protection claim at issue. Because we have not addressed in a published opinion whether Turner applies to equal protection claims in the prison setting, the plaintiffs devoted significant attention to this issue in their supplemental brief and at oral argument. See Pearce v. Sapp, 182 F.3d 918, 1999 WL 503568, at *3 (6th Cir.1999) (per curiam) (unpublished disposition); cf. Pitts v. Thornburgh, 866 F.2d 1450, 1455 (D.C.Cir.1989).
However, this case does not require us to decide whether the Turner reasonable relation standard applies to equal protection claims in the prison setting. We note first that the “law of the case” doctrine may have precluded the district court from reexamining this issue, either because it had previously decided the same issue at an earlier stage in the litigation, Bowling v. Pfizer, Inc., 132 F.3d 1147, 1150 (6th Cir.1998), or because our mandate to the district court was “so narrow in scope as to preclude the district court from considering the issue a second time,” United States v. Thomas, 167 F.3d 299, 306 (6th Cir.1999). In any event, it is unnecessary for us to decide the appropriate level of scrutiny for equal protection claims in the prison setting given the district court’s findings of parity. Disparate treatment is the initial element of an equal protection claim. Bannum, 958 F.2d at 1359. Where a district court concludes, as it did here, that parity of treatment exists, it is unnecessary to engage in any further analysis, be it a reasonable relationship or an intermediate scrutiny analysis. Sullivan, 869 F.2d at 964 (per curiam). Moreover, we reject the plaintiffs’ contention that the district court’s reliance on Turner “permeate[d] the court’s conclusions with regard to parity of programming.” As we discuss in more detail below, the district court’s citations to Turner were immaterial.
We likewise reject the plaintiffs’ argument that the court improperly placed the burden on the plaintiffs to prove current equal protection violations. The plaintiffs contend that the burden should have been on the defendants since they were moving to terminate federal court jurisdiction. We agree with the plaintiffs’ position that the defendants should bear the burden at this stage in the proceedings, but find that the district court, in fact, imposed this burden on the defendants. The structure of the district court’s opinion makes clear that it found that the evidence submitted by the defendants demonstrated parity, while the plaintiffs’ attacks on this evidence were unpersuasive.
The District Court’s Factual Findings
Before we address the district court’s specific factual findings, we pause to consider the plaintiffs’ contention that the district court erred in refusing to admit or consider evidence concerning the quality of various programming. The plaintiffs insist that such evidence would show that even if the quantity of programming for female inmates is comparable to that for male inmates, the quality of the programming is substantially different. We decline to address this argument because the plaintiffs failed to properly preserve it for appeal. “As a general rule, error may not be predicated on a ruling [562]*562which excludes evidence unless an offer of proof is made.” Snyder v. Ag Trucking, Inc., 57 F.3d 484, 492 (6th Cir.1995); see Fed.R.Evid. 103(a). The plaintiffs failed to make such an offer. This failure to develop an evidentiary record deprives this court of a clear basis for evaluating the district court’s evidentiary ruling. See United States v. Luce, 713 F.2d 1236, 1241 (6th Cir.1983). We are even less inclined to excuse the failure here given that we had retained jurisdiction and the plaintiffs proceeded with the evidentiary hearings before the district court without seeking our review of the court’s evidentiary rulings.
1.
Educational Opportunities
The district court limited its evaluation to college programming opportunities available to male and female inmates — a limitation that the plaintiffs do not challenge. All inmates regardless of gender may obtain college education via correspondence courses at their own expense. Glover, 35 F.Supp.2d at 1017. Only three facilities in the entire MDOC system provide college programming at the state’s expense, all pursuant to court order. Those facilities include one male facility and both of the major female facilities, Crane and Scott. Id. Female inmates at the Camp Branch facility may transfer to Crane to attend college programming at the state’s expense. Id. at 1017-18. Thus, all female inmates in the Michigan prison system have the opportunity to participate in college programming at the state’s expense. In contrast, the only male inmates who may do so are in one major facility housing approximately 1,000 male inmates. See id. at 1016, 1018. The enrollment rate for female inmates was 20.12% in October 1998, far exceeding the 0.478% enrollment rate for male inmates. In addition, MDOC spent almost twice as much money on female inmate college programming in FY 1998 as it did on male inmate programming. The district court also found the range of degree programs offered to male and female inmates to be comparable. Id. at 1018.
The plaintiffs do not challenge any of the district court’s specific factual findings. Rather, they assert that the district court erred by limiting its inquiry to the course offerings as of October 1998 and refusing to consider degree completion rates and qualitative evidence concerning “the structure and coherence of degree programs.” Id. at 1018-19. The court’s focus on the current course offerings was appropriate given our mandate that it consider the conditions as they exist now, rather than in the past. Glover, 138 F.3d at 254. Moreover, as discussed above, the plaintiffs have not properly preserved the “qualitative” issues for our consideration.
The plaintiffs also suggest that the district court improperly relied on Turner in assessing educational programming opportunities. It is true that the district court cited Turner and our remand opinion in refusing to consider the broader qualitative issues raised by the plaintiffs. However, the district court also found that the evidence before it concerning access to educational programming, the number and types of programs available, enrollment rates, and budgetary expenditures was “more than sufficient for a ruling,” and, in fact, “decisively indicates that female in-matés presently have sufficiently comparable educational opportunities to those provided to male inmates.” Glover, 35 F.Supp.2d at 1019. Thus, the district court did not rely on Turner in reaching its ultimate finding of parity. We hold that the district court’s finding of parity between the educational opportunities offered to male and female inmates is not clearly erroneous.
2.
Vocational Opportunities
The district court found that 17 vocational programs were available to male inmates as of October 1998. Thirty-two of [563]*563the 37 major male correctional facilities offer at least one vocational program, most major male facilities offer one to three programs, and only seven such facilities offer four to six programs each. No major male facility offers more than six vocational programs, and none of the male camp facilities offers any vocational programming. Only 5.473% of the male inmates in major facilities are enrolled in vocational programs. Glover, 35 F.Supp.2d at 1019.
In contrast, female inmates may participate in seven types of vocational programs. Crane offers four programs and Scott offers six. Two of the six programs offered by the Scott facility' — auto mechanics and building trades/restoration — are actually taught at a male facility, and female inmates are permitted to travel to the male facility to attend the programs. Moreover, unlike the males in camp facilities, inmates in the female camp facility who wish to participate in vocational programming may transfer into a major female facility. Id. at 1020 n. 10. The enrollment rate in vocational programs for female inmates as of October 1998 was 7.198%. Id. at 1020. The court concluded that the vocational opportunities available to female inmates are substantially the same as those available to male inmates. In reaching this conclusion, the court noted specifically that the six programs offered most frequently to, and attended most frequently by, males — institutional maintenance, horticulture, building trades/restoration, food service/management, business education technology, and auto mechanics — are also offered to female inmates. This fact, in combination with the comparable enrollment rates, supports the district court’s conclusion. Id. at 1020-21.
While the plaintiffs do not seriously dispute any of the district court’s factual findings outlined above, they raise several challenges to-the district court’s approach to the parity analysis. First, the plaintiffs point out that vocational programming is available to female inmates on a part-time basis only, while males at some facilities have access to full-time vocational programming. We note that the district court did cite to Turner in reasoning that it was required to treat this administrative decision with deference. It also noted, however, that most of the male inmate vocational programming is also part-time and “that full-time programming is a rarity.” Id. 35 F.Supp.2d at 1021 n. 11. Thus, the court’s reliance on Turner merely bolstered its independent finding that this did not constitute a disparity of constitutional magnitude, and we do not believe the district court erred in this regard.
The plaintiffs also maintain that some female inmates are subject to unique restrictions on their participation in vocational programs. For example, the female inmates at Scott who wish to enroll in either the auto mechanics or building trades/restoration program at one of the male facilities are ineligible to do so if they have been found guilty of substance abuse during the preceding year or if they have been found guilty of misconduct during the preceding nine months. The district court reasoned that Turner mandated that it defer to this decision because these limits were necessary to maintain security while transporting inmates to and from the facility. Id. 35 F.Supp.2d at 1021 n. 12. Even if the district court’s reliance on Turner were misplaced, as the plaintiffs urge, we believe that such reliance was unnecessary and immaterial given the district court’s other factual findings. The district court properly observed that the Equal Protection Clause does not require identical treatment. In fact, the evidence demonstrates that there are significant disparities between the opportunities available to male inmates depending on the facilities in which they are housed. We cannot ignore these disparities when comparing the opportunities of male inmates to those of female inmates as a group. The building trades/restoration program is offered in less than one third (11 of 37) of the major male facilities, and the auto mechanics pro[564]*564gram is offered in only 10% (4 of 37) of the major male facilities. Id. at 1019. The plaintiffs have offered no evidence that males housed in the majority of other male facilities that do not offer either of these two programs have any opportunity whatsoever to travel to another facility to participate in one of the programs. Thus, female inmates as a whole are actually afforded a greater opportunity to participate in these two programs than male inmates, regardless of the eligibility restrictions.
The plaintiffs also argue that MDOC “restricts” female inmates by requiring those inmates categorized as the lowest level security risk, Level I, to agree to be categorized at the higher Level II security risk in order to qualify for vocational programming, a requirement not imposed on Level I male inmates. Id. at 1021 n. 12. The district court deferred “to the decision of the prison administrator as this particular limitation falls within the realm of legitimate security concerns protected by Turner.” Id. Again, regardless of the district court’s stated reliance on Turner, the “restriction” cited by the plaintiffs does not undermine the district court’s ultimate finding of parity between the vocational opportunities offered to male and female inmates. The plaintiffs’ comparison between Level I female inmates and Level I male inmates is simply improper. The Level I category for male inmates actually consists of two sub-categories: “secure” Level I and un-seeure, or camp, Level I. See id. at 1016. Such a distinction does not exist within the Level I female inmate population. Moreover, any slight disparity that may exist between Level I female inmates and a comparable population of Level I male inmates is mitigated by the great disparity in the size of the male and female populations system wide, the huge disparity in number of facilities housing each group, and the fact that all female inmates who meet MDOC eligibility requirements enjoy access to vocational training, while males housed in camp facilities have no such access. See id. at 1021.
We likewise reject the plaintiffs’ contention that the district court erred in finding parity because the range of vocational opportunities offered to male inmates is greater than that offered to female inmates. The district court acknowledged the literal difference in the range of vocational programs offered to male and female inmates but concluded that the enrollment rates, combined with evidence of substantially similar core offerings, supported a finding of parity. We agree. This finding is buttressed by the evidence of the substantial disparities in the program opportunities available to male inmates. Parity does not require identical treatment, especially in this context where each of the numerous prison facilities offers a somewhat different menu of programming regardless of the gender of its inmates.
While the district court cited to Turner in comparing the range of programs available to male and female inmates, it did so only as an alternative to its finding that MDOC had achieved parity in vocational programming. See id. Since we agree with the district court’s finding as to parity, we find the district court’s reference to Turner to be immaterial.
The plaintiffs next argue that the district court erred in calculating the number of vocational programs offered to female inmates, insisting that the court should not have considered three vocational programs that MDOC instituted between August and December 1998. The plaintiffs expressed some concern that the defendants would discontinue these programs. We hold that the district court properly considered these programs in comparing the offerings for male and female inmates. Once parity has been achieved, there is no legitimate basis for federal court monitoring of a state prison system. The plaintiffs’ distrust of the defendants’ motives is insufficient to justify continuing federal court jurisdiction.
[565]*565Finally, the plaintiffs point out that only male inmates can participate in supplemental vocational programming through the On-the-Job Training program (OJT). The court found some evidence that a “very small number” of male facilities provided such programming. Id. at 1022. While the court treated the evidence concerning the existence of OJT programming with some skepticism, it found that the supplemental programs, if they did exist, did not alter the balance of the evidence supporting parity. Id. Thus, contrary to the plaintiffs’ assertion, the district court did not entirely disregard the OJT programs. The district court’s findings in this regard were not clearly erroneous.
3.
Apprenticeship Opportunities
The apprenticeship program is comprised of three components: specific apprenticeship programs, prison industries, and the OJT programs. With regard to specific apprenticeship programs, the district court found that as of October 1998, male inmates at only one major facility could participate in 12 apprenticeships certified by the U.S. Department of Labor’s Bureau of Apprenticeship and Training (BAT). In contrast, female inmates at Crane could participate in five BAT-certified apprenticeships, and female inmates at Scott could participate in six BAT-certified apprenticeships. Between the two facilities, seven different types of apprenticeships are available. See id. at 1023. No camp facility for male or female inmates offers apprenticeships. However, female inmates at the Camp Branch are eligible to transfer to one of the two major female facilities to participate in apprenticeships; no comparable opportunity is offered to male inmates of camp facilities. Glover, 35 F.Supp.2d at 1022-23. As of October 1998, only eight male inmates were enrolled in apprenticeships at the major male facility, while 11 female inmates were enrolled in the major female facilities. Id. at 1023.
The plaintiffs insist that informal apprenticeships, in which 44 male inmates were enrolled at the end of 1998, exist at two additional major male facilities. Id. The district court found that the evidence of this fact was questionable, a conclusion that the plaintiffs contest. In any event, the district court found that even if those informal apprenticeships did exist, the opportunities for female inmates are sufficiently comparable to those for male inmates. In light of the evidence that all eligible female inmates could enroll in seven different kinds of apprenticeships and only a small portion of male inmates could enroll in the 12 programs offered at one major male facility, “female inmates as a group enjoy much greater access than male inmates as a group.” Id. We agree. While the district court cited to Turner to bolster this finding, its reliance on Turner was immaterial.
The district court also compared the offerings in prison industries, referred to as Michigan State Industries (MSI). Figures from September 1998 reveal comparable enrollment rates: 3.6% for female inmates versus 2.8% for male inmates. Id. at 1024. Noting that male inmates as a group enjoy opportunities in 20 MSI programs, compared to five programs for the female inmates, the district court nevertheless concluded that this difference was reasonable under Turner given the vast disparity in the population sizes between male and female inmates. Id. Again, we hold that the district court’s conclusion was not clearly erroneous and, in fact, its reliance on Turner was unnecessary. When viewing the male and female prison populations as a whole, the apprenticeship opportunities available to female inmates are comparable to those available to male inmates.
The district court finally compared the OJT offerings. Noting that the evidence as to the existence of such offerings was sketchy, at best, the court emphasized that the OJT program is provided only at a few of the 37 major male facilities. Id. There [566]*566was no OJT program available to female inmates. The court reasoned that “[g]iven that female inmates already have a robust and sufficiently comparable opportunity to participate in traditional apprenticeship and prison industry opportunities, this minor benefit to the males does not constitute a disparity in treatment under the Equal Protection Clause.” Id. This finding was not clearly erroneous.
C.
Access to Courts
The parties settled the access to courts issue based upon MDOC’s agreement to adopt the court orders and plans governing male inmates’ access to courts in Knox v. Johnson, 63 F.Supp.2d 751 (S.D.Tex.1999) and Hadix v. Johnson, 182 F.3d 400 (6th Cir.1999). The plaintiffs argue that the parties may disagree over the interpretation of the Knox and Hadix court orders and the proper implementation of those orders in female facilities. Given the potential for such disputes, the plaintiffs maintain that termination of the district court’s jurisdiction over access to courts is premature. We disagree. If, in the future, the parties dispute the interpretation or implementation of their agreement and cannot resolve their dispute, they can seek judicial intervention in the same manner as any other party alleging a breach of contract. The history of this case, unfortunate as it may be, does not entitle the plaintiffs to judicial oversight merely as a prophylactic measure.
III.
Having fully examined the record and considered the parties’ arguments, we AFFIRM the district court’s findings of fact and conclusions of law and we order the termination of all federal court jurisdiction of this litigation.