dissenting:
In my view, this case was not fit for disposition on summary judgment because there are genuine issues as to the material facts. Hence, the case should be remanded to the District Court for further factfind-ing. Accordingly, I dissent.
The appellants, thirty-one females convicted in District of Columbia (“D.C.” or the “District”) courts and incarcerated in federal institutions, present two challenges to the District of Columbia Good Time Credits Act of 1986, D.C.Code §§ 24-428 to 24-434 (1989) (“Act”). They claim, first, that the Act violates the equal protection clause by excluding female long-term offenders from the substantial benefit of good time credits because of their gender without an “exceedingly persuasive justification,” Mississippi Univ. for Women v. Hogan, 458 U.S. 718, 724, 102 S.Ct. 3331, 3336, 73 L.Ed.2d 1090 (1982). Next, they contend that it is irrational to dispense good time benefits randomly, or by gender, according to the building in which D.C. offenders are incarcerated. In short, the court is asked in the instant case to assess whether the Act, effected in concert with the assignment policy for D.C. offenders, is rational and gender neutral. The parties are, however, in total disagreement over the facts and policies underlying the assignment scheme; and, because the case was decided pursuant to summary judgment, the District Court did not make findings concerning these points.
Where, as here, a District Court decides a case on cross-motions for summary judgment, the court of appeals must remand when it finds “that final decision should have awaited a more complete factual development” — even if the parties purport to stipulate on appeal that the case was ripe for decision on summary judgment. Menard v. Mitchell, 430 F.2d 486, 488 (D.C.Cir.1970); see also Washington Post Co. v. United States Dep’t of Health & Human Servs., 865 F.2d 320, 326 (D.C.Cir.1989). “Before the court can apply the law, it must have an adequate factual basis for doing so.” 10A C. Wright, A. Miller & M. *201Kane, Federal Practioe and Prooedure § 2725, at 85 (1983).1
Facts are material if their existence or nonexistence tends to resolve any issue properly raised by the parties. “[I]t is the substantive law’s identification of which facts are critical and which facts are irrelevant that governs.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248, 106 S.Ct. 2505, 2510, 91 L.Ed.2d 202 (1986). With respect to the appellants’ challenges to the rationality of the good time credits provision, effected in conjunction with the offender assignment process, the court must defer to the legislative classification at issue if it is rationally related to a constitutionally permissible state interest. See e.g., McGinnis v. Royster, 410 U.S. 263, 270, 93 S.Ct. 1055, 1059-60, 35 L.Ed.2d 282 (1973); Dunn v. Blumstein, 405 U.S. 330, 337, 92 S.Ct. 995, 1000, 31 L.Ed.2d 274 (1972). In applying this test, the Supreme Court has made it clear that “[t]he State may not rely on a classification whose relationship to an asserted goal is so attenuated as to render the distinction arbitrary or irrational.” City of Cleburne v. Cleburne Living Center, Inc., 473 U.S. 432, 446, 105 S.Ct. 3249, 3258, 87 L.Ed.2d 313 (1985); see also United States R.R. Retirement Bd. v. Fritz, 449 U.S. 166, 177, 101 S.Ct. 453, 460, 66 L.Ed.2d 368 (1980) (legislature may not achieve its purpose “in a patently arbitrary or irrational way”). But before we may determine whether the classification at issue here “bears some fair relationship to a legitimate public purpose,” Plyler v. Doe, 457 U.S. 202, 216, 102 S.Ct. 2382, 2394, 72 L.Ed.2d 786 (1982), we must at least comprehend the challenged state action and the relevant characteristics of the groups to which it does and does not apply. Here, the record does not so permit.
For example, the appellants concede that a large portion of male D.C.Code offenders serve their sentences in institutions operated by the District of Columbia. See Brief of Appellant at 10. However, appellants claim that many male prisoners needing special protection or presenting other unspecified concerns are assigned to federal prisons, and that the federal government maintains a constant number of male D.C. offenders in federal prison. See id. (citing Pitts v. Thornburgh, 866 F.2d 1450, 1457 (D.C.Cir.1989)); Supplemental Brief of Appellant at 11 & n. 15.2 The appellee District of Columbia rejects the latter contention, see Supplemental Brief of Appellees, District of Columbia at 4; and the United States contends that the federal system has accepted only males in need of special treatment and, on three “crisis” occasions, other male offenders in “substantial blocks.” See Supplemental Brief for Fed*202eral Appellees at 4 n. 2. It is unclear what percentage of male prisoners incarcerated in federal institutions is composed of “special concern” prisoners, and whether assignment to federal institutions — as opposed to assignment to D.C., state or county institutions — bears any relationship to the length of the sentence or type of offense.
The majority notes that the good time credits provision implicates two governmental purposes: (1) achievement of the traditional rehabilitative, deterrence and retributional goals of incarceration and (2) avoidance of unlawful prison crowding. The statute, however, reflects a third purpose of “promoting] a uniform system of awarding good time credits.” 3 If the type of offense for which a male is incarcerated determines whether males are sent to federal rather than state or county institutions outside the District, then application of D.C. good time credits to one group, but not to the other, may serve the first of the cited governmental interests. But there is no obvious indication that the statute and the related assignment policy serve the second and third purposes of the statute. Indeed, as the appellants suggest, the assignment scheme may be at odds with these two purposes. There is no way we can reasonably address these contentions without some findings on the details of the assignment and prisoner transfer schemes.
As is noted in Part III of the majority opinion, the record is devoid of information concerning the availability of space for male D.C. offenders in the federal facilities and the policy concerning transfer of males between federal and D.C. facilities. This information is particularly important in light of this court’s recent holding enjoining the District of Columbia “from refusing to accept into the D.C. Department of Corrections facilities newly sentenced adult male prisoners whom the Attorney General assigns to the Department, provided that the designation would not cause the District to violate court-ordered population caps or the constitutional rights of prisoners.” United States v. District of Columbia, 897 F.2d 1152, 1156 (D.C.Cir.1990).
Similarly, while it is undisputed that males incarcerated in state and county prisons do receive good time credits under the Act, neither the bases for assignment nor the factors affecting the duration of the period served within the District for such offenders is evident from the record. Thus, we are in no position to speculate that early release of prisoners in state and county jails, who generally return to the District for part of their sentence, may be expected to relieve the District of overcrowding. Indeed, the United States indicates that females incarcerated in federal institutions serve out a fixed six-month period in the D.C. facilities just prior to release. See Supplemental Brief for Federal Appellees at 5 n. 3 (citing Pitts v. Thornburgh, 866 F.2d 1450, 1452 (D.C.Cir.1989)).
Finally, I disagree with the majority’s characterization of the appellants’ challenge in this case. I understand the appellants to challenge the rationality of awarding good time credits on the basis of the building in which offenders are incarcerated. See Supplemental Brief of Appellant at 2. Any assessment of this claim implicates all of the asserted governmental interests, and the scope of the Act’s application to nonfemales may not be divorced from that assessment. Cf. City of Cleburne, 473 U.S. at 449-50, 105 S.Ct. at 3259-60 (considering rationality of classification in light of similarities and differences between the plaintiff group subjected to the zoning restriction and others not so restricted). Absent information about a large portion of the class affected by the good time credits provision — namely, males in federal, state and county facilities, some *203of whom receive its benefits and others who do not — the court cannot begin to evaluate whether the provision results in “uniform” sentencing or whether it will relieve crowding in the D.C. correctional facilities.
In short, although the parties appear to agree that most females receiving long-term sentences are confined in federal institutions and hence are denied the early-release benefits of the Act,4 it is unclear whether the group of males denied the benefits is comprised of some randomly selected group of offenders with sentences greater than a year, or of a group selected on some basis related to culpability. It is also unclear whether application of the credits to males in state and county institutions serves any of the legitimate governmental goals. Such determinations are material to an assessment of the rationality of the good time credits provision.
The only thing that is clear in this case is that we do not have the information necessary to understand the very scheme we are asked to review. While the Supreme Court has suggested that it might be the court’s role to hypothesize reasons to justify classifications drawn by the legislature, see, e.g., United States R.R. Retirement Bd. v. Fritz, 449 U.S. 166, 101 S.Ct. 458, 66 L.Ed.2d 368 (1980), it has never been our task to hypothesize facts underlying the application of a statutory scheme that we are asked to review. On the record before us, it cannot seriously be maintained that “there are no disputed issues of material fact,” see Jackson v. Thornburgh, 702 F.Supp. at 14. As the parties’ arguments to this court have shown, the District Court’s judgment on this point is in error. Hence, in my view, remand for further development of the factual record is required.
Conclusion
We know that there are large numbers of D.C.Code offenders who are serving time in the federal system. We also know that prisoners who are convicted of exactly the same crimes, even under like circumstances, may be assigned to either a federal or D.C. correctional facility. In addition, we know that these similarly-situated persons will receive different sentences solely on the basis of whether they serve their time in a federal or D.C. institution. Finally, we know that male D.C.Code offenders who serve their sentences in state or county facilities outside the District of Columbia do receive good time credits. This scheme surely does not serve a goal of “uniform sentencing,” and we cannot determine whether or how it serves the goals of rehabilitation/deterrence/retribution or avoidance of unlawful overcrowding. On its face, the good times credit provision, in concert with the disputed assignment system, seems irrational. At best, the question raised cannot be answered for lack of findings on material facts. The case should be remanded.
. Even in a case in which the trial court has rendered findings, the Supreme Court has made it clear that the case should be remanded to the trial court for further findings if there is any possibility of a factual void in the record. For example, in Lehman v. Trout, 465 U.S. 1056, 104 S.Ct. 1404, 79 L.Ed.2d 732 (1984), the Supreme Court vacated the judgment and remanded to the court of appeals with instructions to remand to the District Court "for findings of fact, based on new evidence if necessary,” concerning the evidentiary value of statistical evidence in an employment discrimination case. Id. at 1056, 104 S.Ct. at 1404. That case involved extensive factual determinations by the District Court in the trial context. As the dissent in Lehman noted, evidently the Supreme Court believed that “the Court of Appeals found the facts on ... two points, rather than permitting the District Court to do so in the first instance." 465 U.S. at 1059, 104 S.Ct. at 1406 (Stevens, J. dissenting). In Pullman-Standard v. Swint, 456 U.S. 273, 293, 102 S.Ct. 1781, 1792-93, 72 L.Ed.2d 66 (1982), the Court rejected as "incredible" an attempt by the Court of Appeals to make its own factual determinations with respect to issues on which the district court's findings were set aside for error of law.
. One of appellants’ claims is that the limited application of the Act to persons imprisoned in a District correctional facility is not rationally related to the goal of reducing overcrowding at D.C. prisons. On this point, appellants argue that, if the average sentence length of District offenders in federal prisons were reduced, the federal prison system could absorb a larger number of District prisoners. The District of Columbia contends that "[t]his theory ... is based on the false premise that the federal prison system holds a fixed number of District offenders so that federal prison spaces opened up by shorter sentences for District offenders will be filled by District offenders who otherwise would be sent to the District prison system.” Supplemental Brief of Appellees, District of Columbia at 4. We have no way to determine the validity of the "premise,” because the record is unclear on this point.
. The preamble to the Act explains that the purposes of the Act are:
[t]o promote a uniform system of awarding good time credits; to develop procedures for the administration of good time credits; to provide incentives for offenders to earn good time credits based on individual institutional adjustment, performance, and educational achievement; and to improve prison population control through the use of good time credits.
34 District of Columbia Register 484 (Jan. 23, 1987), reprinted in Brief of Appellant addendum at 19.
. Even this point, however, does not comport with the District Court’s observation that "there was uncontroverted evidence that many women sentenced to prison terms exceeding one year are, in fact, housed in the District of Columbia’s city jail, where they are able to obtain the benefits of the Act.” Jackson v. Thornburgh, 702 F.Supp. 9, 12 (D.D.C.1988).