EUGENE TRAYNOR, PETITIONER V. THOMAS K. TURNAGE, ADMINISTRATOR OF VETERANS' AFFAIRS AND VETERANS' ADMINISTRATION JAMES P. MCKELVEY, PETITIONER V. THOMAS K. TURNAGE, ADMINISTRATOR OF VETERANS' AFFAIRS AND VETERANS' ADMINISTRATION No. 86-622 and 86-737 In the Supreme Court of the United States October Term, 1987 On Writs of Certiori to the United States Courts of Appeals for the Second Circuit and the District of Columbia Circuit Brief for the Respondents TABLE OF CONTENTS Opinions below Jurisdiction Statutory and regulatory provisions involved Question presented Statement Summary of argument Argument: I. Section 211(a) precludes judicial review of VA decisions on veterans' benefit claims, including decisions resting on VA policy or regulations and decisions involving points of law under other statutes A. The language and legislative history of Section 211(a) demonstrate that Congress intended to have veterans' benefits claims decided in an informal, non-adversarial process and without judicial review B. The legislative purpose of Section 211(a) supports preclusion of review in these cases II. The VA's criteria for applying the "willful misconduct" standard of the veterans' benefits laws is a reasonable means of implementing those laws and is not inconsistent with the Rehabilitation Act A. VA regulations have historically considered some forms of alcoholism to be "willful misconduct" barring disability pensions B. The 1977 amendment of the G.I. Bill incorporated the VA's existing "willful misconduct" test C. The 1978 Rehabilitation Act amendment did not alter the effect of the 1977 GI Bill amendment 1. The Rehabilitation Act does not prohibit reasonable distinctions among different types of handicaps 2. There is a reasonable basis for treating alcoholism differently from other handicaps for purposes of veterans' benefits programs Conclusion Appendix OPINIONS BELOW The opinion of the court of appeals in No. 86-622 (Pet. App. 1a-38a) is reported at 791 F.2d 226. The opinion of the district court in No. 86-622 (Pet. App. 39a-82a) is reported at 606 F. Supp. 391. The opinion of the court of appeals in No. 86-737 (Pet. App. 1a-31a) is reported at 792 F.2d 194. The opinion of the district court in No. 86-737 (Pet. App. 32a-47a) is reported at 596 F. Supp. 1317. JURISDICTION The judgment of the court of appeals in No. 86-622 was entered on May 16, 1986. A petition for rehearing was denied on July 15, 1986 (Pet. App. 86a-87a). The petition for a writ of certiorari was filed on October 14, 1986 (a Tuesday following a legal holiday), and was granted on March 9, 1987. The jurisdiction of this Court rests on 28 U.S.C. 1254(1). The judgment of the court of appeals in No. 86-737 was entered on May 30, 1986. A petition for rehearing was denied on August 7, 1986 (Pet. App. 49a). The petition for a writ of certiorari was filed on November 5, 1986, and was granted on March 9, 1987. The jurisdiction of this Court rests on 28 U.S.C. 1254(1). STATUTORY AND REGULATORY PROVISIONS INVOLVED The relevant provisions of 38 U.S.C. 211(a), of Section 203 of the G.I. Bill Improvement Act of 1977, Pub. L. No. 95-202, Tit. II, 91 Stat. 1439, 38 U.S.C. (Supp. II 1978) 1662, of 38 C.F.R. 3.301(c)(2), and of Veterans Administration Manual M21-1, are set out at App., infra, 1a-2a. QUESTIONS PRESENTED 1. Whether 38 U.S.C. 211(a) precludes judicial review of a decision by the Veterans Administration denying a veteran's application for educational benefits and request to extend the statutory period within which the veteran may receive educational benefits. 2. Whether, if we assume that in these cases judicial review is not barred, the denial of benefits violated the Rehabilitation Act, 29 U.S.C. 794. STATEMENT 1. Congress has, for many years, enacted legislation providing benefits to disabled veterans except where the disability resulted from the veteran's willful misconduct. See, e.g., 38 U.S.C. 310, 410, 521 (disability pensions). The same exclusion from benefits also applies to the program involved in this case -- educational benefits for veterans. The educational benefits statute authorizes the payment of benefits within ten years following the veteran's last discharge or release from active duty; however, the strict ten-year limit on educational benefits may be extended for those veterans who were unable to use their benefits during that period "because of a physical or mental disability which was not the result of * * * (their) own willful misconduct." Pub. L. No. 95-202, Tit. II, Section 203(a)(1), 91 Stat. 1439, 38 U.S.C. 1662(a)(1). In the cases presently before the Court, petitioners are veterans who did not utilize the full educational benefits available to them during their respective ten-year periods. In each case, petitioner sought to extend his period of eligibility, contending that he was disabled during part of the delimiting period because of alcoholism. The Veterans Administration (VA) denied extensions to both petitioners in accordance with its longstanding interpretation of the circumstances in which alcoholism would be regarded as "willful misconduct" within the meaning of the benefits statute. The applicable VA regulation (38 C.F.R. 3.301(c)(2)) had been promulgated in 1972, prior to the enactment of the provision for extending the time limit within which disabled veterans could receive educational benefits. When the regulation was issued, it was addressed primarily to alcoholism as a basis for disability pensions and incorporated principles set forth in a 1964 VA administrative decision. 37 Fed. Reg. 20335-20336 (1972) (proposed regulation); 37 Fed. Reg. 24662 (1972) (final regulation). /1/ The 1964 administrative decision, drawing on VA rulings dating back to 1931, distinguished between "primary" alcoholism and alcoholism that is "secondary to and a manifestation of an acquired psychiatric disorder." Administrator's Decision No. 988 (Aug. 13, 1964) (J.A. 138, 142-143). Such "secondary" alcoholism is not considered willful misconduct (id. at 143). Nor does the 1964 VA decision regard as the kproduct of willful misconduct any organic disorder caused by chronic alcoholism, such as cirrhosis of the liver, gastric ulcer, peripheral neuropathy, vitamin deficiency, or chronic brain syndrome (id. at 144). "While it is proper to hold a person responsible for the direct and immediate results of indulgence in alcohol, it cannot be reasonably said that he expects and wills the disease and disabilities which sometimes appear as secondary effects" (ibid. (emphasis in original)). Consistently with this policy, the VA grants extensions of the delimiting period to disabled veterans whose alcoholism is the secondary product of a psychiatric disorder or whose alcoholism has caused an organic disorder. The agency's policy does not permit an extension to be granted to an alcoholic veteran who cannot show the existence of either the specified underlying or derivative disorder. It also does not grant an extension on account of a disability suffered, for example, in an automobile accident by a veteran who was driving under the influence of alcohol. Should the VA deny a veteran's request to extend his delimiting period for receiving educational assistance benefits, that veteran would still remain eligible to receive a VA educational loan covering the full-time studies the veteran was pursuing when his delimiting period ended (38 U.S.C. 1662(a)(2)(A)). 2. No. 86-622: Eugene Traynor was honorably discharged from the Army on August 27, 1969, after serving on active duty for 18 months. He entered college in 1977 and received veterans' education assistance benefits until those benefits were terminated when his ten-year period of eligibility expired on August 27, 1979. Traynor, who had used nine and one-half of the 24 months of benefits available to him (based on length of service), sought to have his period of eligibility for benefits extended. He contended that he had been unable to utilize his full benefits within ten years of discharge because he had suffered from alcoholism for 15 years ending in 1974. Pet. App. 3a-4a. During the administrative proceedings, Traynor asserted that the VA regulation stating the circumstances in which alcoholism constitutes willful misconduct is violative of the Rehabilitation Act, 29 U.S.C. 794. The Board of Veterans Appeals did not expressly adjudicate that statutory claim, noting that it was bound by VA regulations. The Board did, however, explain that the consistent VA policy (Pet. App. 117a) is: that alcoholism lcan and should be considered an illness for purposes of medical treatment and rehabilitation, and that the simple drinking of any alcoholic beverage is not in and of itself willful misconduct. On the other hand, if in the consumption of alcohol for the purpose of enjoying its intoxicating effect excessive indulgence leads to disability, such disability will be considered the result of the person's willful misconduct. Noting that "Congress has never enjoyed the luxury of having unlimited funds with which to provide for gratuitous Veterans Administration benefits," the Board explained that historically benefits have not been granted for a disability that results from willful misconduct (id. at 117a-118a). The Board observed that the veterans benefits programs have regarded alcoholism as potentially disqualifying misconduct ever since the earliest veterans regulations promulgated by President Roosevelt. The Board added that (id. at 118a-119a): Since then, a distinction has been maintained between fortuitously incurred disease or disability, for which gratuitous Veterans Administration benefits may be afforded, and other nonfortuitous disabilities incurred at the hands of the claimant himself/herself. Alcoholism is not singled out for special consideration; other disabilities may be considered the result of willful misconduct, under appropriate circumstances. Whether the illness i(n) question is alcoholism or some other disability, the Veterans Administration evaluates the circumstances of each individual in determining willful misconduct. Finding no error in its prior determination that the facts of this case warranted a finding of willful misconduct, the Board denied Traynor's request for benefits beyond his delimiting date. Traynor then filed suit in the United States District Court for the Southern District of New York. He alleged that the VA decision violated the Rehabilitation Act, the Due Process Clause and the Equal Protection component of the Fifth Amendment. The district court held that "(s)ince (the complaint) requires us to examine constitutional and statutory questions and not merely issues of VA policy, we conclude, in accordance with the Supreme Court's holding in Johnson (v. Robison, 415 U.S. 361 (1974)), that we are not precluded from exercising our jurisdiction in this matter by 38 U.S.C. Section 211(a)." Pet. App. 58a-59a. On the merits, the district court rejected the constitutional challenge (id. at 59a-64a), but held that the VA decision violated the Rehabilitation Act. The court held that alcoholism is a handicap covered by the Rehabilitation Act (id. at 69a-72a), and that the denial of benefits constitutes discrimination against alcoholics forbidden by that Act. The court of appeals for the Second Circuit reversed. The panel majority held that 38 U.S.C. 211(a) bars judicial review of the Rehabilitation Act issue. The court stated (Pet. App. 16a-17a) that although "many veterans have in the service of our country suffered injuries that qualify them as 'handicapped individual(s)' for purposes of Section 504 of the (Rehabilitation Act) * * * , Congress did not delineate any exception to section 211(a) for 'handicapped' veterans when it passed section 504." Thus, the court explained, there is no basis for concluding that Congress intended "to grant to 'handicapped' veterans the judicial review traditionally denied all other veterans." Pet. App. 17a. Judge Kearse dissented on the jurisdictional issue. /2/ She suggested that Section 211(a) does not bar judicial review because the Rehabilitation Act neither provides benefits to veterans nor is it administered by the VA (Pet. App. 32a). In addition, Judge Kearse deemed Section 211(a) to be inapplicable because there was no decision of the Administrator on the Rehabilitation Act issue, the Board of Veterans Appeals having "refused, on the ground of lack of authority, to decide whether the challenged regulations violated the Rehabilitation Act" (Pet. App. 36a). 3. No. 86-737: Petitioner McKelvey was honorably discharged from the Army in September 1966 after serving on active duty for three years (Pet. App. 4a). From 1966 to 1971 he was employed as a salesman for a surgical supply corporation (C.A. App. 65-66, 89-90). During the next four years he was hospitalized at various times for alcoholism and associated conditions. He received educational benefits from the VA briefly in 1973 and 1974 (C.A. App. 76-77). When he applied for additional benefits in 1978, more than 10 years after his discharge, the Board of Veterans Appeals denied his request to extend his period of eligibility and rejected his application for benefits. The Board found, after a hearing, that there was "'no evidence that an acquired psychiatric disease preceded (McKelvey's) alcoholism'" (Pet. App. 5a (citation omitted)). McKelvey filed suit in the United States District Court for the District of Columbia. He claimed that the denial of benefits was based on a misconstruction of the "willful misconduct" language of the veterans benefits statute. He contended also that the VA decision constituted discrimination against the handicapped in violation of the Rehabilitation Act, an argument he had not presented in the administrative proceedings. The district court held that it had jurisdiction to consider McKelvey's claims, stating that Section 211(a) "does not prevent judicial review of challenges to the VA's authority to promulgate regulations" (Pet. App. 36a). On the merits, the district court held that the VA had properly interpreted the "willful misconduct" standard of the veterans' benefits statute. The court noted that when Congress enacted the educational benefits extentions, the VA interpretation of "willful misconduct" already existed (in connection with earlier provisions on disability compensation), and that Congress specifically expressed an intent that the same interpretation be used (id. at 40a, quoting S. Rep. 95-468, 95th Cong., 1st Sess. 69-70 (1977)). The district court reached a different conclusion on the Rehabilitation Act claim, holding that the VA interpretation constitutes discrimination against alcoholics in violation of Section 504 (Pet. App. 43a). The court of appeals for the District of Columbia Circuit reversed. The court held that while Section 211(a) does not preclude judicial review of the Rehabilitation Act claim, petitioner's substantive statutory claim has no merit. The court of appeals' decision on the jurisdictional issue rests on "the unusual, perhaps sui generis posture of this case" (Pet. App. 6a). The court focused on two particular facts: first, that a veteran is challenging the validity of a regulation under the Rehabilitation Act, a legal issue the Board of Veterans Appeals then regarded itself as lacking authority to decide, and second, that the VA had not otherwise made a determination on that issue prior to the filing of this lawsuit /3/ (id. at 7a). Since, in the court's view, Section 211(a) is applicable only when a claim has been "resolved by an actual 'decision of the Administrator'" (ibid., quoting Johnson v. Robison, 415 U.S. 361, 367 (1974)), it does not bar judicial review in these circumstances. The court emphasized "the narrowness of our holding" (86-737 Pet. App. 9a): (W)e do not anticipate another occasion to review a VA order on the basis that supports our review today. The VA has now determined it does have authority to decide on the effect and applicability of federal statutes other than veterans' legislation when the agency acts on benefits claims. We therefore expect that the VA will not again regard as outside the arsenal of law it applies any potentially relevant congressional enactment. On the merits, the court concluded that the VA could reasonably distinguish between those whose handicap was caused by their own willful misconduct, and those who are not responsible for their handicap. The VA's conclusion that alcoholics who cannot show an underlying psychiatric disorder are chargeable with willful misconduct reflects "general societal perceptions regarding personal responsibility" (Pet. App. 12a). Moreover, since "(a)lcoholism, unlike any other disability except drug addiction * * * , is self-inflicted * * * (,) (i)t is therefore feasible for alcoholism, as it is not for all other disabilities except drug addiction, to make a generalized determination that willfulness exists unless there is established the singular exculpation for self-infliction (psychiatric disorder) that the agency has chosen to acknowledge" (id. at 16a). In a separate opinion, Judge Ginsburg concurred in the court's holding that Section 211(a) does not bar judicial review in the unique circumstances of this case, and she dissented from the court's holding on the merits (Pet. App. 17a). Judge Scalia also wrote separately. He dissented from the court's holding that Section 211(a) is not applicable, stating that the "decision of the Administrator" which Section 211(a) immunizes from judicial review necessarily includes all issues within the competence of the agency to decide, "whether or not (the agency) specifically adverts to, or is even aware of them -- just as a court necessarily 'decides' all issues logically essential to the validity of its holding whether or not it explicitly addresses or considers them" (Pet. App. 30a). Any other view of Section 211(a), he wrote, would enable "the Administrator * * * to control the scope of judicial review of his determinations by simply designating which underlying issues he chooses not to decide" (Pet. App. 30a). Judge Scalia concurred in the court's decision on the merits, upholding the validity of the VA regulation. SUMMARY OF ARGUMENT 1. Section 211(a) bars judicial review in this case. The statute precludes review of "decisions of the Administrator on any question of law or fact under any law administered by the (VA) providing benefits." 38 U.S.C. 211(a). This case may be viewed as involving either a decision denying petitioners benefits, or a decision as to how the "willful misconduct" test of the veterans' benefits statute applies to alcoholics in light of the Rehabilitation Act. In either view, the case involves a "decision() of the Administrator on any question of law or fact under any (veterans' benefits) law." The legislative history of Section 211(a) supports this reading. The purpose of Section 211(a) was to avoid involving the courts in "day-to-day determination and interpretation of Veterans' Administration policy," particularly where that policy involves "technical considerations." Johnson v. Robison, 415 U.S. 361, 372, 373 (1974). The VA decisional process is tightly controlled by hundreds of regulations, many of them highly detailed and technical, appearing in nearly three hundred pages of the Code of Federal Regulations, 38 C.F.R. Pts. 3 and 4, as well as an even larger number of standards appearing in internal manuals. If challenges to regulations were subject to judicial review, "day-to-day determinations" involving "technical considerations" would be routinely brought into the courts. And many of these cases could be cast as complaints against handicap discrimination under the Rehabilitation Act, since a "disability" claimed to exist under the veterans' benefits law can often be described as a "handicap" under the Rehabilitation Act. Even if only a small percentage of administrative claims were to be litigated, a substantial addition to the federal court docket would result; in fiscal 1986, the Board of Veterans Appeals denied more than 28,000 claims. 2. Even if Section 211(a) were held not to bar judicial review in these cases, petitioners' challenges to the VA decisions would have to be rejected on their merits. The "willful misconduct" test of the veterans' benefits law was expressly intended by Congress to cover alcoholism and drug addiction and to endorse the VA's long-standing interpretation (which was cited in the Senate report). This language represents a deliberate congressional decision that these disabilities are unique in the sense that they frequently involve significant elements of volition. The VA regulations, by focusing the "willfulness" inquiry on whether an underlying psychiatric disorder exists, is a reasonable way of applying the "willful misconduct" test, and should be upheld. The Rehabilitation Act was not intended to alter or repeal the specific determination by Congress in the veterans' benefits law to treat alcoholism and drug addiction as involving significant elements of volition. The Rehabilitation Act was not intended to forbid differing treatment of different handicaps, particularly where differing treatment is required by some other statute and is supported by significant medical knowledge. Here, even those medical authorities who label alcoholism a "disease" concede that it is a disease that can and often does involve significant elements of volition. Indeed, many authorities stress that the successful treatment of an alcoholic requires that the patient assume personal responsibility for abstaining from drink. In addition, the authorities agree that societal attitudes -- which are reflected in laws and regulations -- can also influence the prevalence of alcoholism. Just as this Court has rejected the argument that the "disease" label absolves alcoholics from criminal responsibility for their conduct, Powell v. Texas, 392 U.S. 514 (1968), so too Congress and the VA have reasonably decided that alcoholics should bear a degree of personal responsibility in connection with a benefits program. ARGUMENT I. SECTION 211(a) PRECLUDES JUDICIAL REVIEW OF VA DECISIONS ON VETERANS' BENEFITS CLAIMS, INCLUDING DECISIONS RESTING ON VA POLICY OR REGULATIONS AND DECISIONS INVOLVING POINTS OF LAW UNDER OTHER STATUTES A. The Language and Legislative History of Section 211(a) Demonstrate That Congress Intended To Have Veterans' Benefits Claims Decided in an Informal, Nonadversarial Process and Without Judicial Review Congress created the Veterans Administration in 1930 and vested in the VA responsibility for administering the federal program for veterans' benefits. Walters v. National Ass'n of Radiation Survivors, 473 U.S. 305, 309 (1985). In the expectation that the system for disbursing veterans' benefits would be as "informal and nonadversarial as possible" (id. at 323), Congress did not "contemplate the adversary mode of dispute resolution utilized by courts in this country" (id. at 309). /4/ In accordance with that expectation, Congress also sought to "protect the Administrator from expensive and time-consuming litigation" by precluding judicial review of VA benefits decisions. Rose v. Rose, No. 85-1206 (May 18, 1987), slip op. 8; 38 U.S.C. 211(a). See Walters, 473 U.S. at 307; Johnson v. Robison, 415 U.S. 361, 370 (1974). Section 211(a) sets forth the prohibition of judicial review: (T)he decisions of the Administrator on any question of law or fact under any law administered by the Veterans' Administration providing benefits for veterans and their dependents or survivors shall be final and conclusive and no other official or any court of the United States shall have power or jurisdiction to review any such decision by an action in the nature of mandamus or otherwise. So plain and direct is this statutory proscription of judicial review that this Court has referred to Section 211(a) as the paradigm of the "unambiguous and comprehensive" language Congress employs when it intends to "bar judicial review altogether" (Lindahl v. OPM, 470 U.S. 768, 779-780 & n.13 (1985)). /5/ The legislative history of Section 211(a) also strongly supports the conclusion that Congress intended to preclude judicial review in the circumstances of these cases. Section 211(a) was originally enacted as Section 5 of the Economy Act of 1933, ch. 3, 48 Stat. 9 (emphasis added), which provided: All decisions rendered by the Administrator of Veterans' Affairs under the provisions of this title, or the regulations issued pursuant thereto, shall be final and conclusive on all questions of law and fact, and no other official or court of the United States shall have jurisdiction to review by mandamuss or otherwise any such decision. Congress could hardly prohibit judicial review in more explicit terms (see Briscoe v. Bell, 432 U.S. 404, 409 (1977)). The original statutory language clearly applied to all decisions under the veterans' benefits laws, and to all questions of law involved in those decisions, including questions arising under other statutes. An applicant for veterans' benefits seeks a decision under the veterans' benefits laws; and a decision denying a claim for veterans' benefits is a decision under the veterans' benefits laws, whether or not the claim for benefits involves consideration of additional statutes. There is no indication that subsequent legislative changes in Section 211(a) were intended to change the original meaning. In 1940, the statute was amended to preclude judicial review of "the decisions of the Administrator of Veterans' Affairs on any question of law or fact concerning a claim for benefits or payments under this or any other Act administered by the Veterans' Administration." Act of Oct. 17, 1940, ch. 893, Section 11, 54 Stat. 1197 (emphasis added). The Senate Report that accompanied this amendment emphasized what the language of the amendment made obvious: that the statute "provides for the finality of decisions made by the Administrator of Veterans' Affairs on questions relating to claims under any of the laws administered by the Veterans' Administration." S. Rep. 2198, 76th Cong., 3d Sess. 11 (1940) (emphasis added). Thus, under the 1940 language, as well as the original 1933 language, Congress expressed its intent to preclude judicial review of all benefits decisions made by the VA Administrator, including decisions that involve questions of law arising under other statutes. After a minor amendment in 1957 not affecting the present case, /6/ the statute was amended in 1970 to preclude review of "decisions of the Administrator on any question of law or fact under any law administered by the (VA)." Act of Aug. 12, 1970, Pub. L. No. 91-376, Section 8(a), 84 Stat. 790, codified at 38 U.S.C. 211(a). The 1970 amendment represented Congress's response to several decisions by the United States Court of Appeals for the District of Columbia Circuit that had construed the preclusion of judicial review too narrowly. As this Court explained in Johnson v. Robison, 415 U.S. 361, 371, 373 (1974) (emphasis in original), the 1970 amendment was designed to restore the provision to its original unqualified meaning: Before (the 1970) amendment, the no-review clause made final "the decisions of the administrator on any question of law or fact concerning a claim for benefits or payments under (certain) law(s) administered by the Veterans' Administration" (emphasis added), 38 U.S.C. Section 211(a) (1964 ed.), 71 Stat. 92. In a series of decisions, e.g., Wellman v. Whittier, 104 U.S. App. D.C. 6, 259 F.2d 163 (1958); Thompson v. Gleason, 115 U.S. App. D.C. 201, 317 F.2d 901 (1962); and Tracy v. Gleason, 126 U.S. App. D.C. 415, 379 F.2d 469 (1967), the Court of Appeals for the District of Columbia Circuit interpreted the term "claim" as a limitation upon the reach of Section 211(a), and as a consequence held that judicial review of actions by the administrator subsequent to an original grant of benefits was not barred. * * * * * Thus, the 1970 amendment was enacted to overrule the interpretation of the Court of Appeals for the District of Columbia Circuit * * *. /7/ The Chairman of the House Committee on Veterans' Affairs argued that the District of Columbia Circuit's decisions gave "preferential treatment to a limited group of beneficiaries" by providing to them, and them alone, judicial review of VA benefits determinations. 116 Cong. Rec. 26490 (1970) (remarks of Congressman Teague). He recognized that such preferential treatment could be avoided either by making "court review * * * apply to all beneficiaries with equal force" or by making all VA benefits decisions nonreviewable (ibid.). In the 1970 amendment Congress sought to restore uniformity to the benefits process by opting for the latter approach. Chairman Teague explained that the statute "would seem to be perfectly clear in expressing the congressional intent that any and all decisions of the Administrator on questions of entitlement to veterans' benefits -- (with the exception of claims on insurance contracts) -- were to be final and not subject to judicial review." 116 Cong. Rec. 19734 (1970). In sum, the history of Section 211(a) demonstrates that prior to the 1970 amendment, the statutory preclusion of review plainly applied to all decisions of the Administrator under the veterans' benefits laws, even where questions of law were raised under other statutes. The 1970 amendment was not intended to change this result; instead, it was designed to reaffirm the original meaning by overruling a series of judicial decisions that Congress viewed as erroneously narrowing the statute. B. The Legislative Purpose of Section 211(a) Supports Preclusion of Review in These Cases This Court recently observed that the principal purposes of Section 211(a) are "to achieve uniformity in the administration of veterans' benefits and protect the Administrator from expensive and time-consuming litigation" (Rose v. Rose, No. 85-1206 (May 18, 1987), slip op. 8). Cognizant of these purposes, the Court has stated that Congress sought to avoid "involv(ing) the courts in day-to-day determination and interpretation of Veterans' Administration policy," particularly where that policy involves "technical considerations." Johnson v. Robison, 415 U.S. at 372, 373. /8/ These legislative purposes would plainly be advanced by precluding judicial review in these cases. Despite petitioners' efforts to portray this litigation as something other than a challenge to the VA's benefits decisions in their particular cases, each petitioner's complaint focuses on the circumstances of his individual benefits determination and each complaint requests the court to "(g)rant (petitioner's) appliction for an extension of his delimiting date" (J.A. 31; see id. at 129). Thus petitioners plainly seek to involve the courts in the "day-to-day determination and interpretation" of VA policy, a result that is directly contrary to Congress's objective. Petitioners offer three arguments in favor of creating an exception to Section 211(a) in these cases: that lawsuits challenging the legality of a policy or regulation should be permitted even if judicial review of individual benefit determinations is barred; that these cases involve decisions under the Rehabilitation Act rather than under a veterans' benefits statute; and that in these cases the VA did not conclusively decide the question arising under the Rehabilitation Act, so there is no "decision" to trigger the preclusion of review under Section 211(a). We address each contention in turn. 1. Petitioners' first theory would mean, at a minimum, that the VA's regulations as well as its manuals would be open to judicial review. /9/ The courts would then face precisely the dangers that Congress sought to avoid, because these regulations and manuals are filled with detailed, technical provisions, prescribing in elaborate detail how veterans' benefits claims are to be decided. The VA regulations comprise 272 pages in the 1986 edition of the Code of Federal Regulations (38 C.F.R. Pts. 3 and 4, at 126-398) and include detailed provisions on difficult and controversial subjects. For example, 11 pages address the question whether a disability is service-connected (38 C.F.R. 3.303-3.312), with detailed provisions relating to claims based on exposure to herbicides in Vietnam and exposure to ionizing radiation (38 C.F.R. 3.311a, 3.311b). /10/ Part 4 has 95 pages elaborating on the rating of particular conditions, including 11 pages on neurological, convulsive and mental disorders (38 C.F.R. 4.120-4.132), as well as other provisions on impairment of the musculoskeletal system (38 C.F.R. 4.40-4.73 -- 25 pages), visual impairment (38 C.F.R. 4.75-4.84a -- 8 pages), auditory impairment (38 C.F.R. 4.85-4.87a -- 4 pages), cardiovascular diseases (38 C.F.R. 4.100-4.104 -- 4 pages), and mental disorders (38 C.F.R. 4.125-4.132 -- 5 pages). The Code of Federal Regulations represents only the beginning of potentially litigable issues concerning veterans' disability claims should Section 211(a) be construed to have an implicit exception for VA regulations and policies. An even greater volume of policy pronouncements exists in manuals that amplify the rules set forth in the Code of Federal Regulations. These manuals constitute "instructions of the Administrator" that are binding on the Board of Veterans Appeals under 38 U.S.C. 4004. Under petitioners' construction of Section 211(a), the VA manuals would be subject to judicial review. See Pet. Br. 38-39. For example, the VA regulation on alcoholism that is contained in the Code of Federal Regulations does not discuss the relationship between alcoholism and psychiatric disorders; for guidance on this topic, one must turn to the VA Manual M21-1, ch. 50, Section 50.40a.(1) (entitled Rating Procedure Relative to Specific Issues), Subchapter XII (entitled Mental Disorders). The VA Manual M21-1, which prescribes policies for disability adjudications, has 56 chapters. Of these, at least 25 chapters contain provisions that are potential targets of litigation. In addition, another VA Manual M22-2 governs adjudication of claims for educational benefits; Parts II, III and IV of Manual M22-2, which deal with substantive issues arising in educational benefits claims, contain 31 chapters with detailed provisions that are potential targets of litigation. With this volume of regulatory material waiting in the wings, it is apparent that the suggestion that judicial review be permitted for challenges to regulations and policies would profoundly distort the informal, nonadversarial scheme Congress envisioned. /11/ Petitioners' reading of Section 211(a) finds no support either in the statutory language or in any conceivable view of the legislative intent. The language of the statute precludes review of "decisions of the Administrator on any question of law." The statute does not, as petitioners would have it, limit preclusion of review to "decisions of the Administrator on any question of law (except for decisions embodied in a regulation or manual)." Nor is there any reason to suppose that Congress intended to except from the statutory bar on judicial review those decisions of law that are embodied in VA regulations. Such an exception would serve no discernible congressional purpose. The VA, like most administrative agencies, is free to develop substantive standards under its governing statute either by regulation, by case-by-case adjudication, or by a combination of the two. NAACP v. FPC, 425 U.S. 662, 668 (1976); SEC v. Chenery Corp., 332 U.S. 194, 202-203 (1947). While there may be instances in which reliance on adjudication rather than rulemaking would be an abuse of discretion, NLRB v. Bell Aerospace Co., 416 U.S. 267, 294 (1974), certainly in applying the "willful misconduct" standard of the veterans' benefits statute -- as well as the antidiscrimination provision of the Rehabilitation Act -- the VA has discretion to develop standards through the process of case-by-case adjudication rather than by regulation. In fact, the VA's interpretation of the statutory "willful misconduct" standard as applied to alcoholism originated in decisions on specific benefit claims (see page 3, supra; pages 28-29, infra). Under petitioners' interpretation, the federal courts could review a policy or legal standard announced in a regulation, but not in an individual decision on a claim for benefits. But they offer no cogent reason why Congress would create -- and no evidence that Congress intended to create -- a system barring judicial review where an interpretation is adopted in an adjudicative context (with no prior notice to the general public and with retroactive effect on the parties), but allowing judicial review if the same interpretation were adopted prospectively after public notice and comment. If petitioners' theory were the law, then the agency would have a strong incentive to maintain uniformity in its decisions and minimize its litigation costs by developing standards and policies through case-by-case administrative adjudication rather than regulation, thereby depriving the public of the advantages of the public notice and comment involved in rulemaking proceedings. Congress could not have intended such a bizarre result. /12/ 2. Petitioners' second contention is that judicial review is not barred in these cases because the VA's decisions were made under the Rehabilitation Act rather than under a veterans' benefits statute. This is the theory espoused in Judge Kearse's dissenting opinion (86-622 Pet. App. 25a-38a). On this view, Section 211(a) would be limited to instances in which the Administrator's decision was made exclusively under a veterans' benefits statute and involved no consideration of other statutes. Petitioners' theory does not flow naturally or comfortably from the language of Section 211(a). The VA decisions denying additional benefits to petitioners and denying petitioners' requests to extend their delimiting dates are, by any standard, decisions under a "law administered by the Veterans' Administration providing benefits." That the VA may have been asked by petitioner McKelvey (though not by petitioner Traynor) to consider the effect of the Rehabilitation Act on his claim under the educational benefits program does not cause the agency's decision to cease being a decision under a veterans' benefits law. It is that law, rather than the Rehabilitation Act, that authorizes the benefits being sought. And, as the legislative history shows (see pages 13-16, supra), Congress viewed Section 211(a) as barring review of all decisions of the Administrator implementing a veterans benefits program -- regardless of whether those decisions arguably required the administrator to consider additional statutes, such as the Administrative Procedure Act, 5 U.S.C. 706(2)(A) and (E), or the Rehabilitation Act. Indeed, it is difficult to conceive of a case in which a veteran claiming a "disability" under a veterans' benefits statute could not also claim a "handicap" under the Rehabilitation Act. The predictable consequence of a rule allowing judicial review for veterans claiming Rehabilitation Act protection is that such claims would routinely be recited in lawsuits challenging VA benefits decisions. The result would be to involve the courts in "day-to-day determination and interpretation of Veterans' Administration policy" (Johnson v. Robison, 415 U.S. at 372) -- the very result Section 211(a) was intended to preclude. 3. Petitioners' third theory for avoiding the jurisdictional bar of Section 211(a) is that the statute does not apply here because the VA did not expressly "decide" the question whether the agency's regulation on alcoholism violated the Rehabilitation Act. Petitioners' theory is incompatible with established principles of administrative law. Even in cases where judicial review of agency action is available, courts employ the exhaustion doctrine in order to insure that the agency has had an opportunity to address the legal questions presented. Among the salutary effects of the exhaustion doctrine are that the agency decision may dispose of the matter in a fashion that makes further review unnecessary and that, should judicial proceedings ensue, the court will have the benefit of the agency's views. Weinberger v. Salfi, 422 U.S. 749, 765 (1975); McKart v. United States, 395 U.S. 185, 200 (1969); cf. Bowen v. City of New York, No. 84-1923 (June 2, 1986), slip op. 16-17. Petitioners' tortured reading of Section 211(a) proceeds from just the opposite premise. Under their theory, judicial review would exist only where a court does not have the benefit of the agency's view on the subject, viz., only on those questions of law raised in the administrative proceedings which the Administrator did not expressly "decide." /13/ Conversely, it is only by "deciding" an issue that the Administrator could trigger the statutory proscription of judicial review. There is no evidence that Congress intended to make the jurisdiction of the federal courts hinge on the absence of an agency's statement of its views on a particular question of law. Nor is there any reason to believe that Congress would have crafted the peculiar system petitioners posit: a system that discourages parties from raising legal issues in agency proceedings (in order to preserve judicial "review") and simultaneously encourages the agency to express views on issues that otherwise would not be reached (in order to preserve the finality of agency adjudication and to defeat judicial review). See 86-737 Pet. App. 30a (Scalia, J., dissenting). II. THE VA'S CRITERIA FOR APPLYING THE "WILLFUL MISCONDUCT" STANDARD OF THE VETERANS' BENEFITS LAWS IS A REASONABLE MEANS OF IMPLEMENTING THOSE LAWS AND IS NOT INCONSISTENT WITH THE REHABILITATION ACT Because the courts lack jurisdiction in these cases, it is unnecessary for this Court to consider petitioners' contention that the challenged VA regulation and the VA's denials of benefits on these cases are violative of the Rehabilitation Act. However, should this Court determine that the jurisdictional issue is not dispositive, the decisions below should be affirmed on the alternative ground that petitioners' Rehabilitation Act claims are without merit. In challenging the VA's policy on alcoholism, petitioners and amici curiae raise a subject that has long bedeviled scholars in a variety of disciplines: the crafting of a normative cultural view on alcohol and alcoholism. The striking disagreements the subject provokes are reflected in the various descriptions of alcohol as "a valuable food and commodity, a 'gift of God'" (Blume, Public Policy Issues, in Alcoholism and Related Problems: Issues For The American Public 179 (1984)) and as the "verray sepulture (o)f mannes wit and his discrecioun" (G. Chaucer, The Canterbury Tales (The Pardoner's Tale) in Chaucer's Poetry 314 (E. Donaldson ed. 1958). The complexities of the subject are magnified in these cases by the effort to frame the legal issues within the nomenclature of medicine. The briefs of petitioners and their supporting amici appear to proceed from the assumption that the dispositive issue in the case is whether alcoholism is a disease and, in urging reversal, rely on recent medical literature to support the proposition that it is. But the issue in these cases is not a medical issue, it is the legal issue of determining Congress's intent in enacting the relevant statutes. /14/ It is thus of little consequence to the legal analysis whether alcoholism is or is not a disease because that term does not foreclose the possibility of voluntary conduct; nor is the search for the congressional inteent underlying statutes enacted in the late 1970's advanced by citations to the medical literature circa 1987. In any event, as we will discuss, despite substantial recent medical progress in the undestanding of alcoholism, the observations in this Court's plurality opinion in Powell v. Texas, 392 U.S. at 522 (footnote omitted; emphasis in original) retain their validity: the inescapable fact is that there is no agreement among members of the medical profession about what it means to say that "alcoholism" is a "disease." One of the principal works in the field * * * concludes that "a disease is what the medical profession recognizes as such." In other words, there is widespread agreement today that "alcoholism" is a "disease" for the simple reason that the medical profession has concluded that it should attempt to treat those who have drinking problems. There the agreement stops. /15/ We therefore focus our analysis of the statutory issues in these cases on the language Congress adopted and its programmatic context, the legislative history, and the longstanding agency interpretation. A. VA Regulations Have Historically Considered Some Forms of Alcoholism to be "Willful Misconduct" Barring Disability Pensions For many years, disability compensation for veterans has been subject to a statutory bar where "the disability is the result of the veteran's own willful misconduct." 38 U.S.C. 310 (disability compensation for injuries suffered or diseases contracted in line of duty). See also 38 U.S.C. 410 (survivors' benefits); 38 U.S.C. 521 (compensation for non-service connected disabilities). /16/ Over the decades in which disability benefits programs for veterans have been administered, the definition of "willful misconduct" has achieved general acceptance in many specific applications, including its application to alcoholism. The VA first had occasion to consider the "willful misconduct" standard in relation to the consumption of alcoholic beverages in an administrative decision in 1931 (Administrator's Decision No. 2 (Mar. 21, 1931) (J.A. 133-137)). In that case, the Administrator granted compensation to veterans who had been paralyzed from drinking "jamaica ginger," on the ground that the substance they drank was not known to be poisonous. In dictum, however, the Administrator stated that "if in the drinking of any beverage for the purpose of enjoying its intoxicating effects, excessive indulgence leads to disability, willful misconduct would undoubtedly inhere in the act" (id. at 136). This dictum was later incorporated in a regulation and manual, which stated the test to be: "Was there excessive indulgence and was it the proximate cause of the injury or disease in question" (id. at 139-140). In 1964 the VA clarified its policy on alcoholism in two significant respects. Taken together these clarifications showed that the willful misconduct standard did not disqualify all alcoholics from receiving benefits and that large categories of persons disabled by alcoholism were expressly permitted to obtain benefits. The Administrator announced these policy modifications in an administrative decision (Administrator's Decision No. 988 (Aug. 13, 1964) (J.A. 138-146). First, the Administrator drew a dictinction between primary alcoholism and alcoholism "secondary to and a manifestation of an acquired psychiatric disorder" and held that the latter condition, secondary alcoholism, is not to be considered as willful misconduct (id. at 143). /17/ The Administrator's 1964 decision also removed the "willful misconduct" label from those alcoholics, both primary and secondary, whose conditions had produced derivative disabling effects, including "cirrhosis of the liver to gastric ulcer, peripheral neuropathy, vitamin deficiency, chronic brain syndrome or simply acceleration of debility of age" (id. at 144). In 1972, the VA issued its present regulation. Drawing upon the language of the 1931 "jamaica ginger" decision, it states (38 C.F.R. 3.301(c0(2)): If, in the drinking of a beverage to enjoy its intoxicating effects, intoxication results approximately and immediately in disability or death, the disability or death will be considered the result of the person's willful misconduct. In issuing the regulation, the VA announced that it was intended to incorporate the principles of the 1964 administrative decision. 37 Fed. Reg. 20335, 20336 (1972) (proposed regulation); 37 Fed. Reg. 24662 (1972) (final regulation). Consequently, the VA Manual incorporates the 1964 decision, stating that alcoholism is not a compensable disability unless it is "secondary to and a manifestation of a psychotic, psychoneurotic or psychophysiologic disorder," in which case the rating official is to "consider the alcoholism part and parcel of the disability and rate as one disease entity, e.g., schizophrenia with alcoholism." VA Manual M21-1, ch. 50, Subchapter XII (see S. Zimberg, supra, at 41). Thus, by 1977, when Congress enacted the provision extending for disabled veterans the ten-year limitation on educational benefits, the "willful misconduct" exclusion had a long history and an established administrative interpretation. B. The 1977 Amendment of the G.I. Bill Incorporated the VA's Existing "Willful Misconduct" Test Since the 1940's the VA has granted educational assistance to eligible veterans. Congress added the ten-year delimiting period in 1974 (Pub. L. No. 93-337, Section 1(1), 88 Stat. 292). In 1977, Congress amended the statute to provide for the first time for extensions of the delimiting period for veterans who were prevented from initiating or completing their education "because of a physical or mental disability which was not the result of such veteran's own willful misconduct." Pub. L. No. 95-202, Title II, Section 203(a)(1), 91 Stat. 1439, 38 U.S.C. 1662)(a)(1). The legislative history of the 1977 amendment shows that in choosing the term "willful misconduct" Congress was aware that it was requiring the same test that was already in the statute as a bar to disability pensions, and that Congress intended to adopt the interpretation the VA had already given to that term in the disability pension context. The report of the Senate Veterans' Affairs Committee makes explicit the legislative awareness of and acquiescence in the VA interpretation of "willful misconduct." The Senate Report states that "(i)n determining whether the disability sustained was a result of the veteran's own 'willful misconduct', the Committee intends that the same standards be applied as are utilized in determining eligibility for other VA programs under title 38. In this connection, see 38 CFR part III, paragraphs 3.1(n) and 3.301 (the "willful misconduct" regulation), and VA Manual M21-1, section 1404." /18/ S. Rep. 95-468, 95th Cong., 1st Sess. 69-70 (1977). This Court has often stated that committee reports "contain the authoritative source for finding the Legislative intent" and "represen(t) the considered and collective understanding of those Congressmen involved in drafting and studying proposed legislation" (Garcia v. United States, 469 U.S. 76, 76 (1984), quoting Zuber v. Allen, 396 U.S. 168, 186 (1969)). The legislative history of the 1977 amendment thus confirms the conclusion otherwise indicated by settled principles of statutory construction: that Congress's use of the term should be given the same meaning in all applications. See Morrison-Knudsen Construction Co. v. Director, OWCP, 461 U.S. 624, 633 (1983); Bob Jones University v. United States, 461 U.S. 574, 586-587 & n.10 (1983); Sedima, S.P.R.L. v. Imrex Co., 473 U.S. 479, 489 (1985). The conclusion that Congress chose to have the VA apply the same test in the same manner to both disability pensions and education benefits is well-supported by the factors this Court ordinarily employs in assessing whether an administrative interpretation is faithful to legislative intent. As this Court observed in an analogous context in Alcoa v. Central Lincoln Peoples' Utility District, 467 U.S. 380, 390 (1984): "(t)he subject under regulation is technical and complex. (The agency) has longstanding expertise in the area, and was intimately involved in the drafting and consideration of the statute by Congress. Following enactment of the statute the agency immediately interpreted the statute in the manner now under challenge." Moreover, the Administrator, whose interpretation is entitled to considerable deference, has consistently and prominently interpreted the term "willful misconduct" with regard to alcoholism both before and after passage of the 1977 amendment; Congress was aware of the Administrator's interpretation and therefore should, at the very least, be viewed as having implicitly approved that interpretation when it employed the same term in the 1977 amendment (City of Pleasant Grove v. United States, No. 85-1244 (Jan. 21, 1987), slip op. 5-6; United States v. Sheffield Bd. of Comm'rs, 435 U.S. 110, 131-135 (1978)). In fact, Congress's approval is explicitly stated in the Senate report. It is, accordingly, clear that in 1977, when Congress enacted the extension of the delimiting period for disabled veterans whose disability was not caused by willful misconduct, Congress understood precisely how that language would be interpreted in cases where the claimed disability is alcoholism. C. The 1978 Rehabilitation Act Amendment did not Alter the Effect of the 1977 G.I. Bill Amendment Despite the clarity of the situation in 1977, petitioners and amici supporting them contend that subsequent events have overtaken Congress's explicit approval of the relevant VA regulation. Although cast in different forms, petitioners' principal argument is that amendments to the Rehabilitation Act in 1978 effectively nullified the VA regulation. Thus, petitioners contend that alcoholism should be regarded as a disease, that the conduct of drinking is inseparable from the disease itself, and therefore that the Rehabilitation Act prohibits the VA from concluding that an alcoholic's drinking can ever have a volitional component. /19/ Under this argument, an alcoholic veteran who had been disabled by drinking during the statutory period is ipso facto entitled to an extension of the delimiting period for educational benefits. /20/ Since the same "willful misconduct" standard applies also to the grant of veterans' disability pensions, acceptance of petitioners' argument would lend credence to the quite surprising notion that the government is obligated to pay disability pensions to all persons claiming to suffer from alcoholism. Petitioners' argument is seriously flawed. As we now discuss, the 1978 amendments to the Rehabilitation Act did not repeal the recent congressional judgment that extensions of the delimiting period would be unavailable to persons who bear some responsibility for their disabilities. Nor does the Rehabilitation Act prevent the government from making reasonable distinctions among types of handicaps. Before we turn to that discussion, some further context for the issues is useful. We have said that the issue in this case is not whether alcoholism is a disease, nor is there an issue whether alcoholism is a handicap cognizable under the Rehabilitation Act (see 43 Op. Att'y Gen. 12 (1977)). For, elven if it is a disease, it would not necessarily follow that all alcoholics are disabled. The statute requires that a veteran suffer from a disability in order to qualify for an extension of his delimiting period. Moreover, even if one accepts the disease conception of alcoholism and considers only those veterans whose alcoholism is disabling, that would not necessarily preclude scrutiny of the veteran's conduct or his volition as a contributing cause of his disability. Alcoholism, like drug addiction, may well be a disease as petitioners argue; but, as even the medical authorities cited by amici recognize, alcoholism "cannot be reified but reflects a collection of various symptoms and episodic behaviors that collectively make up perhaps as many alcoholisms as there are alcohol abusers" (G. Vaillant, The Natural History of Alcoholism 3 (1983)). In short, alcoholism is not a unitary condition. Substance abuse, particularly alcoholism, has multiple forms and ranges of severity, /21/ and is acknowledged to be unlike other diseases in the sense that it frequently involves a significant volitional element, in both its genesis and its treatment. /22/ As we show, it is reasonable in the context of this particular disease, unlike other diseases, for the VA to draw a distinction based on a showing of an underlying psychological disorder or disabling derivative effects; it is a distinction that is paralleled in the medical literature and that faithfully implements Congress's decision not to permit extensions of the delimiting period (or disability pensions) for veterans whose disabilities were caused by willful misconduct. Although the VA policy may not be expressed in precisely the terms medical science would use, and although the policy may not produce in an individual case the same conclusion another arbiter might reach, the VA policy provides a reasonable and workable accommodation of modern medico-psychological evidence and Congress's instructions in the veterans' benefits statute. Perhaps, as medical science makes further advances, some modification of the statutory standard will be indicated; but that is a judgment to be made by Congress, not by the courts. 1. The Rehabilitation Act does not prohibit reasonable distinctions among different types of handicaps When first enacted in 1973, the Rehabilitation Act's bar on discrimination against handicapped persons did not apply to federal government programs; it was limited to federally-funded programs and activities. Pub. L. No. 93-112, Tit. V, Section 504, 87 Stat. 394. The original statutory definition of a handicapped individual was a person who has a physical or mental disability that results in a substantial handicap to employment (Pub. L. No. 93-112, Section 7(6), 87 Stat. 361). Congress expanded that definition in 1974 to include, for purposes of Section 504, a person who has a physical or mental impairment which substantially limits one or more major life activities, has a record of such impairment, or is regarded as having such impairment. Pub. L. No. 93-516, Tit. I, Section 11(a), 88 Stat. 1619. In the Rehabilitation, Comprehensive Services, and Developmental Disabilities Amendments of 1978 (Pub. L. No. 95-602, Tit. I, 92 Stat. 2955), Congress further amended the Rehabilitation Act in two ways that are significant to this case. Section 504 was amended to prohibit discrimination "under any program or activity conducted by any Executive agency or by the United States Postal Service," and required the heads of those agencies to promulgate regulations prohibiting discrimination against handicapped persons (Pub. L. No. 95-602, Tit. I, Sections 119, 122(d)(2), 92 Stat. 2982, 2987, 29 U.S.C. 794). Congress also amended the definition of "handicapped person" to state that for purposes of Section 504, in the employment context, the term "does not include any individual who is an alcoholic or drug abuser whose current use of alcohol or drugs prevents such individual from performing the duties of the job in question or whose employment, by reason of such current alcohol or drug abuse, would constitute a direct threat to property or the safety of others" (Pub. L. No. 95-602, Tit. 1, Section 122(a)(6), 92 Stat. 2985, 29 U.S.C. 706(7)(B)). Petitioners' principal contention is that the 1978 amendments to the Rehabilitation Act invalidated the VA regulation concerning alcoholism. Since Congress had expressly approved that VA regulation in 1977, petitioners' argument is reduced to the proposition that the Rehabilitation Act amendments implicitly repealed the 1977 VA Benefits Law amendments. /23/ The short answer to petitioners' argument is, as this Court has stated repeatedly, that a subsequent statute will not be held to repeal an earlier statute by implication, especially where the later enactment is general and the earlier statute is specific. Radzanower v. Touche Ross & Co., 426 U.S. 148, 153 (1976); Morton v. Mancari, 417 U.S. 535, 550-551 (1974); Silver v. NYSE, 373 U.S. 341, 357 (1963); Posadas v. National City Bank, 296 U.S. 497, 503 (1936); Wood v. United States, 41 U.S. (16 Pet.) 342, 363 (1842). Moreover, there are strong indications that Congress regarded the 1977 G.I. Bill amendment and the VA regulation as retaining their full force notwithstanding the 1978 Rehabilitation Act amendment. In 1979, the Senate Committee on Veterans' Affairs revisited the VA's interpretation of the "willful misconduct" test for extensions of the delimiting period as that test applied to alcohol or drug dependence disabilities. Although, upon reexamination, the Committee preferred to have the VA grant a "delimiting period extension when the veteran was prevented, during part of all of the ordinary 10-year delimiting period, from using GI Bill benefits by a drug or alcohol disability and the veteran has recovered from the disability," the Committee recognized that new legislation would be needed to effect that result since the VA's contrary interpretation had been expressly endorsed in the Senate Report accompanying the 1977 amendment. S. Rep. 96-314, 96th Cong., 1st Sess. 25 (1979). Accordingly, the Committee concluded that "in light of the legislative history (of the 1977 amendments), the VA has had little choice but to deny such extensions involving alcohol and drug abuse or addiction disabilities" (ibid.). On four occasions between 1979 and 1984 the Senate passed bills containing amendments that would have eliminated the "willful misconduct" test for VA educational benefits extensions (while retaining the test for VA disability benefits), but none of the bills passed the House. See S. Rep. 98-604, 98th Cong., 2d Sess. 38-39 (1984) (summarizing legislative history). There is not the slightest suggestion in the legislative history that, as petitioners' theory assumes, modification of the G.I. Bill was regarded as unnecessary in light of the Rehabilitation Act's 1978 amendments. Rather, "despite the (Senate) Committee's strongest urgings, the House would accept neither the GI Bill nor the rehabilitation program provision for delimiting-period extensions based on drug or alcohol disabilities" (id. at 39); see 126 Cong. Rec. 27578 (1980) (remarks of Senator Cranston, Chairman, Committee on Veterans' Affairs) ("the House was adamant in its refusal to accept this provision"). If, as petitioners urge, the Rehabilitation Act amendments invalidated the VA regulation on alcoholism or superseded that regulation with respect to the educational benefits delimiting-period extension, that fact seems to have eluded Congress. The various refinements of petitioners' theory are similarly flawed. Petitioners contend, for example, that the Rehabilitation Act proscribes distinctions among different types of handicaps and that the VA runs afoul of that standard by imposing only on veterans disabled by alcoholism (and drug dependence) the burden of proving that their disability was not the result of willful misconduct. The range of handicaps covered by the Rehabilitation Act is extensive, including a wide variety of conditions from mental illness, alcoholism and drug addiction, to all types of serious physical afflictions. /24/ Obviously, different handicaps present different problems; no one would contend, for example, that a blind person and a paralyzed person must be treated in identical fashion. The Rehabilitation Act's general prohibition of discrimination against handicapped persons has never been construed to preclude government agencies from recognizing the differences among handicaps in determining how best to deal with them. /25/ In each case, the nature of the particular handicap must be considered in determining what is required by the Act. See Southeastern Community College v. Davis, 442 U.S. 397 (1979) (deaf applicant for nursing school); Doe v. New York University, 666 F.2d 761 (2d Cir. 1981) (medical school applicant with history of mental illness). These considerations are particularly relevant where the differing treatment of a particular handicap is required by statute. HEW's implementing regulations under the Rehabilitation Act of 1973 (applying to federally-funded programs) provide that "exclusion of a specific class of handicapped persons from a program limited by Federal statute or executive order to a different class of handicapped persons" is not prohibited by the Act. 42 Fed. Reg. 22676, 22679 (1977), promulgating 45 C.F.R. 84.4(c); see 28 C.F.R. 41.51(c) (HEW coordinating regulation for federally-assisted programs), 28 C.F.R. Pt. 39 (Department of Justice regulations), 38 C.F.R. 18.404(c) (VA regulation for federally-assisted programs), 52 Fed. Reg. 25124 (1987) (VA notice of proposed regulation for federally-conducted programs). This Court has "recognized (the HEW) regulations as an important source of guidance on the meaning of section 504." Alexander v. Choate, 469 U.S. 287, 304 n.24 (1985); School Bd. of Nassau County v. Arline, No. 85-1277 (Mar. 3, 1987), slip op. 5; Consolidated Rail Corp. v. Darrone, 465 U.S. 624 (1984); Southeastern Community College v. Davis, 442 U.S. 397, 413 (1979). The HEW regulation is persuasive authority here. In the 1977 G.I. Bill amendment, Congress allowed handicapped persons to utilize educational benefits beyond ten years following their discharge, but restricted this special allowance to cases involving "willful misconduct." /26/ In doing so, Congress obviously focused on alcoholism and drug addiction, as evidenced not only by the Senate report's citation and approval of the VA's regulation (discussed at pages 30-31 supra), but its use of the same language that had already been interpreted in the context of disability pensions to include alcoholism and drug addiction. Thus, in 1977 Congress mandated separate treatment of drug and alcohol addiction under a "willful misconduct" standard; the HEW regulation and subsequent implementing regulations by other agencies properly recognize that it was never the intent of the Rehabilitation Act to alter existing statutes requiring separate treatment of different handicaps. 2. There is a reasonable basis for treating alcoholism differently from other handicaps for purposes of veterans' benefits programs As we have explained, Congress directed the VA to apply the same "willful misconduct" standard to applicants for educational benefits extensions as to applicants for disability pensions. In assessing the reasonableness of the VA's interpretation of "willful misconduct," it is therefore necessary to consider the consequences of that interpretation in both contexts, yet petitioners do not explain how the Rehabilitation Act would require a change in the VA regulation interpreting the willful misconduct standard when that standard is applied to the educational benefits program but not when it is applied to disability pensions. /27/ Congress has clearly stated the reasons for its reluctance to grant disability pensions to alcoholics or drug addicts (S. Rep. 96-314, supra, at 25-26): In the context of (the disability compensation) program, the rate and duration of benefits depend directly upon the severity and duration of the disability. Thus, an individual receiving benefits under that program for alcoholism or drug addiction would have a strong financial incentive -- in the form of a higher rate of compensation or the continuation of receipt of compensation -- in the worsening or prolongation of the disability, both of which are to some extent within his or her control because they depend upon the amount, frequency, and duration of his or her consumption of alcoholic beverages or drugs. One may differ, as a matter of policy, on the question whether extensions of the ten-year limitation on educational benefits for alcoholic veterans should be measured by the same standard as disability pensions. In its several recent efforts to enact legislation repealing the "willful misconduct" bar to educational benefits extensions while retaining it for disability pensions, the Senate Committee on Veterans' Affairs has urged that a different standard be used for each program on the ground that educational benefits cases involve recovered alcoholics who should be rewarded for their recovery and encouraged to go to school (S. Rep. 96-314, supra, at 25-26): T he Committee would anticipate that the GI Bill educational assistance would have considerable value to the social and economic rehabilitation of veterans who have recovered from disabilities related to alcohol or drugs. See also S. Rep. 98-604, supra, at 38-39. Congress's repeated rejection of that argument, however, has not been unreasonable. Under the statute, a qualified veteran -- whether handicapped or non-handicapped, alcoholic or non-alcoholic -- has ten years from the date of his discharge to utilize the G.I. Bill educational benefits. 38 U.S.C. 1662(a)(1). Persons in petitioners circumstances therefore had some meaningful access to the educational benefit program. Cf. Alexander v. Choate, 469 U.S. at 301-306 (Rehabilitation Act is not violated by state Medicaid rule that had disparate impact on the handicapped). Moreover, eligible persons affected by alcoholism have available to them, during the delimiting period and thereafter, medical and rehabilitative services provided by the VA. 38 U.S.C. 1610. Even persons such as petitioners, if they are pursuing a full-time educational program when their delimiting period ends and if their request for an extension is denied, remain eligible for VA educational loans after the delimiting date (38 U.S.C. 1662(a)(2)(A)). In these circumstances, Congress might well conclude that ten years is time enough for a veteran to resolve a drug or alcohol problem -- that it would be inappropriate, and would indeed communicate the wrong incentives, to tell veterans that there is no time limit on how long the VA will wait for them to resolve their problems and proceed with their publicly-funded education. Just as this Court has concluded that conduct induced by alcoholism may be subject to deterrence through criminal sanctions, despite medical opinion labelling alcoholism a "disease" (Powell v. Texas, supra), so too is it permissible for Congress to decide that the "disease" label does not require that alcoholism be categorized for all purposes with other disabilities less likely to involve elements of volition. Medical concepts developed in the context of diagnosis and treatment do not control the administrative decision of how best to interpret and further the purposes of the veterans' benefit statute (see note 14, supra). /28/ Moreover, even those medical writers who label alcoholism a "disease" for purposes of diagnosis and treatment, perceive that the label embraces a range of conditions including some in which volition plays a significant role that may be affected by appropriate incentives. /29/ Even proponents of the "disease" concept recognize that an alcoholic may be able to exercise control over the amount he drinks, depending on the individual and on the professional or other support he or she receives. /30/ Indeed, even the view expressed by some medical commentators, that alcoholism represents a physical condition, either inherited or acquired, does not negate an element of volition: "'physical' hypotheses do not claim that there is no volition in the alcoholic's excess drinking but that, partly because of physical abnormality, the alcoholic is one who faces a choice which is (increasingly) more difficult than for most people." Fingarette, The Perils of Powell: In Search of a Factual Foundation for the "Disease Concept of Alcoholism," 83 Harv. L. Rev. 793, 805 (1970). /31/ Petitioners and the amici criticize the court of appeals' decision in McKelvey for separating the conduct of an alcoholic from his underlying condition, allowing the VA to label the conduct of excessive drinking as "willful misconduct." In fact, however, this approach coincides with a widely used and frequently successful approach to the treatment of alcoholics (Fingarette, supra, 83 Harv. L. Rev. at 807 n.64): Alcoholics Anonymous maintains that alcoholism is a "disease," but not that drinking is involuntary. On the contrary, the entire approach in Alcoholics Anonymous is to enlist the voluntary cooperation of the alcoholic, to appeal to him on moral-religious-pragmatic grounds to voluntarily abstain from drinking, and to engage in reciprocal self-help along these lines with his brother AA members. /32/ This approach is also compatible with recent studies, which stress that the "disease" concept of alcoholism is perfectly consistent with an insistence that the alcoholic assume responsibility for his or her own conduct (G. Vaillant, supra, at 299): /33/ In conveying the concept that alcoholism is a disease to the patient, it is important also to underscore that alcoholism is a disease that is highly treatable, but that like the treatment of diabetes, treatment of alcoholism will require great responsibility from the patient. Even ardent advocates of the "disease" concept recognize that a person with the genetic predisposition to alcoholism may exercise personal responsibility and choose not to drink (Talbott, Alcoholism and Other Drug Addictions: A Primary Disease Entity, 75 J. Med. Ass'n Ga. 490, 493 (1986), cited by amici American Medical Association, et al. (Br. 13 n.7)): There are millions of people in this country who have the genetic predisposition, but who for cultural, health, personal, or religious reasons will not abuse, so they never manifest the disease. In sum, Congress decided to apply a "willfulness" test to the participation of alcoholics in veterans' benefits programs. That decision is consistent with medical authority, which recognizes that many alcoholics are not completely helpless but retain a significant degree of volition. In its regulations and policies implementing the statute, the VA has acted reasonably in focusing the "willfulness" determination on the presence or absence of an underlying psychiatric disorder or a derivative organic disorder. The VA's approach -- which Congress has endorsed -- makes a reasonable distinction among types of handicaps, based on the "willfulness" test and substantial medical authority. It is therefore perfectly consistent with the Rehabilitation Act. /34/ CONCLUSION The judgments of the courts of appeals should be affirmed. Respectfully submitted. CHARLES FRIED Solicitor General RICHARD K. WILLARD Assistant Attorney General LAWRENCE G. WALLACE Deputy Solicitor General JERROLD J. GANZFRIED Assistant to the Solicitor General ANTHONY J. STEINMEYER ROBERT V. ZENER Attorneys AUGUST 1987 /1/ A companion provision (38 C.F.R. 3.301(c)(3)) establishes similar standards for determining when drug usage constitutes "willful misconduct" rendering the recipient ineligible for a disability pension or an educational benefits extension. /2/ She expressed no view on the merits (Pet. App. 38a). /3/ In response to the court's request at oral argument, the General Counsel of the VA took the position that the Rehabilitation Act does not invalidate the VA's interpretation of the "willful misconduct" language in Section 1662. The court held, however, that the General Counsel's letter did not constitute a "decision of the Administrator" for purposes of Section 211(a), since it was written after the case started. Pet. App. 7a-8a. /4/ This Court stated in Walters that (473 U.S. at 311) "(t)he process is designed to function with a high degree of informality and solicitude for the claimant. There is no statute of limitations, and a denial of benefits has no formal res judicata effect; a claimant may resubmit as long as he presents new facts not previously forwarded. See 38 C.F.R. Sections 3.104, 3.105 (1984)." /5/ In Section 211(a) Congress has provided "clear and convincing" evidence (Abbott Laboratories v. Gardner, 387 U.S. 136, 141 (1967)) of its intent that VA benefits decisions not be subject to examination by the courts. Congress has evidenced this intent in terms that are "'fairly discernible' in the details of the legislative scheme" and that are sufficiently clear to overcome "the general presumption favoring judicial review of administrative action" (Block v. Community Nutrition Inst., 467 U.S. 340, 351 (1984), quoting Data Processing Service v. Camp, 397 U.S. 150, 157 (1970); see Bowen v. Michigan Academy of Family Physicians, No. 85-225 (June 9, 1986), slip op. 5 & n.4). That conclusion is supported by reference to the sources to which this Court ordinarily turns in assessing legislative intent: the express language of the statute, "'the structure of the statutory scheme, its objectives, its legislative history, and the nature of the administrative action involved'" (Lindahl v. OPM, 470 U.S. at 779 (quoting Block v. Community Nutrition Inst., 467 U.S. at 345)). Where, as in these cases, there is "persuasive reason to believe" that Congress sought to preclude review, then that intent must be respected. Morris v. Gressette, 432 U.S. 491, 501 (1977); Abbott Laboratories v. Gardner, 387 U.S. at 140. /6/ The 1957 amendment changed the statute to preclude judicial review of the Administrator's decision "on any question of law or fact concerning a claim for benefits or payments under any law administered by the Veterans' Administration." Act of June 17, 1957, Pub. L. No. 85-56, Section 211(a), 71 Stat. 92. The text of the 1933, 1940 and 1957 versions of the statute appear in Daylo v. Administrator of Veterans' Affairs, 501 F.2d 811, 813 (D.C. Cir. 1974). /7/ This Court's reading of the 1970 amendment, as intended to reaffirm the original meaning of the statute, is confirmed by the fact that in the 1970 amendment Congress retained the effective date of October 17, 1940. Pub. L. No. 91-376, Section 8(a), 84 Stat. 790. Had the 1970 amendment been intended as a change, presumably a 1970 effective date would have been used. /8/ The courts have articulated only two limited exceptions -- not applicable here -- to the statutory bar on judicial review: where a challenge is made to the constitutionality of a statute (Johnson v. Robison, supra) and where a challenge to a VA policy is made, not by a party seeking benefits, but by a third party affected by the policy who has no opportunity to be heard in administrative proceedings. See, e.g., Evergreen State College v. Cleland, 621 F.2d 1002 (9th Cir. 1980); University of Maryland v. Cleland, 621 F.2d 98 (4th Cir. 1980); Merged Area X (Education) v. Cleland, 604 F.2d 1075 (8th Cir. 1979); Wayne State University v. Cleland, 590 F.2d 627 (6th Cir. 1978). The cases in the latter group were not brought by veterans seeking review of denials of benefit claims. Rather, they were brought by "educational institutions interested in the overall administration of the VA educational benefits program" (86-622 Pet. App. 18a). Petitioners argue (Br. 46 n.31) that in Wayne State and Merged Area X the educational institutions were joined as plaintiffs by individual veterans; but there is no indication in these opinions that the veterans were seeking to challenge their individual benefits determinations, or even that the VA had denied benefits to those veterans. In fact, the courts' rationale makes it clear that had the plaintiffs challenged the outcomes of individual benefits determinations, the courts would have held that Section 211(a) barred jurisdiction (604 F.2d at 1078; 590 F.2d at 632). In situations where neither of these limited exceptions obtains, the general rule has prevailed, precluding judicial review of benefits decisions at the behest of the veteran. E.g., Barefield v. Byrd, 320 F.2d 455 (5th Cir. 1963); Milliken v. Gleson, 332 F.2d 122 (1st Cir. 1964), cert. denied, 379 U.S. 1002 (1965); Redfield v. Driver, 364 F.2d 812 (9th Cir. 1966); Fritz v. Director of Veterans Administration, 427 F.2d 154 (9th Cir. 1970); Wickline v. Brooks, 446 F.2d 1391 (4th Cir. 1971), cert. denied, 404 U.S. 1061 (1972); Ross v. United States, 462 F.2d 618 (9th Cir. 1972); De Rodulfa v. United States, 461 F.2d 1240 (D.C. Cir.), cert. denied, 409 U.S. 949 (1972); Holley v. United States, 352 F. Supp. 175 (S.D. Ohio 1972), aff'd without opinion, 477 F.2d 600 (6th Cir.), cert. denied, 414 U.S. 1023 (1973); Anderson v. VA, 559 F.2d 935 (5th Cir. 1977); Rosen v. Walters, 719 F.2d 1422, 1424-1425 (9th Cir. 1983); Pappanikoloaou v. Administrator of the Veterans Administration, 762 F.2d 8, 9 (2d Cir. 1985); Roberts v. Walters, 792 F.2d 1109 (Fed. Cir. 1986). It is true that, as petitioners state (Br. 42-45), there came a time long after Section 211(a) was enacted, and more than a decade after it was last amended, when the Administrator of the VA wrote that Section 211(a) "does not protect every decision, policy, or action of the VA from judicial scrutiny" (130 Cong. Rec. S6160 (daily ed. May 22, 1984)). That statement was not made in connection with the enactment, amendment, or reenactment of Section 211(a). The same is true of the floor statements, cited by petitioners (Br. 44-45), made by several individual members of the Senate in 1984, which are in any event inconclusive (see 130 Cong. Rec. S6158 (daily ed. May 22, 1984) (remarks of Senator Cranston) ("the VA and the Justice Department, despite my urging, refuse to agree to stop raising in court the bar to judicial review -- section 211(a) of title 38 -- as a defense in all cases challenging VA regulations")). These snippets of subsequent legislative materials thus provide no guidance as to Congress's intent in enacting Section 211(a) in 1933 and amending it in 1940 and 1970. /9/ VA policy is typically embodied in two forms: (1) regulations appearing in Title 38 of the Code of Federal Regulations, and (2) internal manuals that amplify and explain the regulations. The present cases involve both types of policy enactment: the VA regulation (38 C.F.R. 3.301(c)(2)) and the VA's Adjudication Manual, which expands upon the terse language of the regulation. /10/ There is presently pending a suit challenging the VA's regulations and policies with respect to veterans' claims for disabilities allegedly resulting from exposure to herbicides in Vietnam. Nehmer v. Veterans Administration, No. 86-6160TEH (N.D. Cal.). /11/ Nor is it accurate to suggest that judicial review of challenges to policies and regulations would not greatly burden the VA with litigation because, unlike lawsuits challenging individual benefits decisions, the validity of a policy or regulation need be litigated only once. Nothing requires all possible challenges to a regulation by various persons to be presented in a single lawsuit, and a regulation could be challenged unsuccessfully by different persons many times. Revisions in the regulations would be subject to similar challenges. And even if it were true that each policy or regulation would be the subject of litigation only once, the sheer volume of VA policies and regulations would produce the substantial, costly litigation Congress precluded. /12/ The First Circuit, in the context of a statute barring judicial review of actions taken under the Federal Employees Compensation Act (5 U.S.C. 8128(b)), recently characterized an argument identical to petitioners' as a distortion of the statute. Paluca v. Secretary of Labor, 813 F.2d 524, 527 (1987). The First Circuit stated (ibid. (footnote omitted; emphasis in original)): It would create the absurd result of permitting a court to strike down a policy statement of the Secretary, notwithstanding the court's inability to review any subsequent individual adjudications for conformance with its policy decision. As recently stated by the Court of Appeals for the District of Columbia, it is individual determinations that "have traditionally been accorded more rather than less judicial protection against agency error than generally applicable rules. Compare Bi-Metalic Investment Co. v. State Board of Equalization, 239 U.S. 441 (1915) with Londoner v. Denver, 210 U.S. 373 (1908)." Gott v. Walters, 756 F.2d 902, 915 (1985) (Scalia, J.). * * * (Plaintiffs') suggestion makes no sense, because the Secretary would be free, in individual adjudications, to ignore anything we say, and because it conflicts with one of the major purposes of Section 8128(b): "that the Secretary should be free to make the policy choices associated with disability decisions." Rodriguez v. Donovan, 769 F.2d 1344, 1348 (9th Cir. 1985). /13/ The extreme reach of the theory is evident on the record in No. 86-737, where petitioner McKelvey did not even raise any question under the Rehabilitation Act in the administrative proceedings. Had he raised the Rehabilitation Act issue, and had that issue been expressly rejected by the VA, then presumably Section 211(a) would bar judicial review of that "decision." Petitioner's theory thus would discourage parties from raising before the agency questions on which judicial review may later be sought. /14/ See Powell v. Texas, 392 U.S. at 526 (plurality opinion) (footnote omitted) (pointing out "the conceptual difficulties inevitably attendant upon the importation of scientific and medical models into a legal system generally predicated upon a different set of assumptions"); id. at 541 (Black, J., concurring) ("Medical decisions concerning the use of a term such as 'disease' or 'volition,' based as they are on the clinical problems of diagnosis and treatment, bear no necessary correspondence to the legal decision whether the overall objectives of the criminal law can be furthered by imposing punishment."); M. Guttmacher, The Role of Psychiatry in Law: The Irresistible Impulse 53-54 (1968) ("It is my earnest hope that the psychiatrist will not be assigned the task of expressing gradations of responsibility for the alcoholic offender but will be required only to furnish clinical diagnoses, when that is possible."). /15/ A recent article cited (at 5) in the amicus brief of the American Medical Association, states (Categories, Careers, and Outcomes of Alcoholism, The Lancet 719 (Mar. 29, 1986) (emphasis added)): Although the medical and neuropsychiatric consequences of excessive drinking have long been established as matters of clinical concern, the notion that alcoholism -- the thing itself -- is a "disease" is more recent and more controversial. /16/ The "willful misconduct" test is derived from the benefits statute for World War I veterans. Act of Oct. 6, 1917, ch. 105, Section 300, 40 Stat. 405. /17/ By one estimate, 20% to 30% of alcoholics fall within this "secondary" category. See American Medical Association Amicus Br. 7. One of the medical texts cited by amicus National Council on Alcoholism, Inc. (Br. 14, 19, 26), and by amici American Medical Association, et al. (Br. 8), describes as a "major group * * * those alcoholics who develop drinking problems after the onset of a major psychiatric disorder such as manic-depressive illness, anxiety-panic disorder, or schizophrenia. * * * These individuals can be considered secondary alcoholics" (S. Zimberg, The Clinical Management of Alcoholism 153 (1982)). The same text stresses that "(a) distinction must be made between the primary and the secondary alcoholic" (id. at 20; see id. at 40-41). The author of that text, Dr. Zimberg, submitted an affidavit in the district court in No. 86-622 (J.A. 56-80). An article cited as authoritative by the amici American Medical Association, et al. endorses distinctions between primary and secondary alcoholism, describing a survey indicating that secondary alcoholism is more difficult to control. Categories, Careers, and Outcomes of Alcoholism, The Lancet 719 (Mar. 29, 1986). In this survey, the primary alcoholics had a "lower intensity of drinking, and fewer social problems than most in the secondary categories," while the secondary alcoholics had "the poorest clinical outcomes with, as might be expected, higher rates of social and forensic problems" (ibid.). /18/ The cited Manual provision states that "(b)asic principles for application in deciding cases involving alcoholism are stated in Administrator's Decision No. 988 * * *." Administrator's Decision No. 988 is the 1964 decision upon which the VA regulation is based (see J.A. 138-146). /19/ The very existence, let alone extensive membership, of the Alcoholics Anonymous organization (see pages 45-47, infra) is not easily reconciled with this aspect of petitioners' contentions. /20/ Amici American Medical Association, et al. makes a somewhat different argument. Conceding that the Rehabilitation Act still allows the G.I. Bill's "willful misconduct" standard to be applied to alcoholics, the amici argue that the Act requires the determination to be made on a case-by-case basis, without the aid of any regulation requiring the factfinder to focus on the presence or absence of an underlying psychological disorder. /21/ See, e.g., E. Jellinek, The Disease Concept of Alcoholism 35-41 (1960) (identifying five different types of alcoholics, only two of which are regarded as suffering from a "disease"); S. Zimberg, supra, at 20 (distinguishing between primary and secondary alcoholism); Bohman, et al., Maternal Inheritance of Alcohol Abuse, 38 Arch. Gen. Psychiatry 965, 968 (1981) (describing genetically different types of susceptibility to alcoholism, each producing a different form of alcoholism); Cloninger, et al., Inheritance of Alcohol Abuse: Cross-Fostering Analysis of Adopted Men 38 Arch. Gen. Psychiatry 861, 867 (1981) (identifying two types of alcohol abuse with different genetic and environmental causes); Schuckit, Genetic Aspects of Alcoholism, 15 Annals of Emergency Medicine 991-992 (1986) (distinguishing primary alcoholism from secondary alcoholism, which occurs "in the context of another major preexisting psychiatric illness"); American Psychiatric Ass'n, Diagnostic and Statistical Manual of Mental Disorders 129-138, 163-170 (3d ed. 1980) (hereinafter DSM-III) (identifying seven organic mental disorders attributed to the ingestion of alcohol, all of which are distinguished from alcohol abuse and alcohol dependence). "(A)lcohol abuse reflects a multidetermined continuum of drinking behaviors whose determinants are differently weighted for different people and include culture, habits, social mores, and genes" (G. Vaillant, supra, at 17). See Zimberg, Office Psychotherapy of Alcoholism in Alcoholism and Clinical Psychiatry 218 (J. Solomon ed. 1982) ("alcoholism does not occur as an all-or-none phenomenon but has varying degrees in the same individual"). /22/ One medical commentator states that "(l)ike an automobile driver who chooses to drive rapidly down a busy highway in a car with defective brakes and ends up spending two years in an orthopedic rehabilitation clinic, the alcoholic may consciously have made some early decisions related to his eventual disorder. But such conscious choice becomes less and less important with the passage of time" (G. Vaillant, supra, at 17). /23/ Of course, Congress did not expressly repeal the 1977 VA amendments. Moreover, neither the language nor the legislative history of the 1978 Rehabilitation Act amendments evidences any congressional awareness of an effect on the veterans' educational benefit extension provision or any congressional intent to repeal a standard it had endorsed just one year earlier. /24/ In proposing the original regulations under the Rehabilitation Act, HEW referred to "the diversity of existing handicaps and the differing degree to which particular persons may be affected." 41 Fed. Reg. 20296 (1976). The regulations define "handicapped persons" in terms of "physical or mental impairments" which substantially limit one or more major life activities (or are regarded as doing so); "physical or mental impairment" includes "orthopedic, visual, speech, and hearing impairments, cerebral palsy, epilepsy, muscular dystrophy, multiple sclerosis, cancer, heart disease, diabetes, mental retardation, emotional illness, and drug addiction and alcoholism." 28 C.F.R. 41.31(a) and (b)(1). /25/ Indeed, Congress has itself enacted legislation providing benefits only to persons suffering from certain handicaps (e.g., 29 U.S.C. 721(a)(5), 706(13) ("severe handicaps")), legislation that grants priorities only to persons with a specific disability (e.g., Pub. L. No. 93-516, Tit. II, Sections 201-210, 88 Stat. 1622-1630 (blindness)), and legislation that expressly excludes persons with certain handicaps from eligibility (e.g., Rehabilitation Act Section 7(7)(B), 29 U.S.C. 706(7)(B), which, in specified circumstances, excludes persons who abuse alcohol or drugs from the definition of "handicapped person"). /26/ Given the limited resources available for veterans' benefits, Congress may reasonably place limitations on eligibility for such benefits, and it is Congress's prerogative to determine the manner in which priorities will be set and resources allocated. Cf. McDonald vs. Board of Elections, 394 U.S. 802, 809 (1969). /27/ There is presently pending in the Third Circuit a case in which an alcoholic veteran is challenging the VA "willful misconduct" regulation as applied to claims for disability pensions on the basis of alcoholism. Buck v. Veterans Administration, No. 86-1656. The Third Circuit has stayed proceedings in that case pending this Court's decision in the present cases. Amicus Vietnam Veterans of America states (Br. 2-3) that one of its reasons for participating here is its participation in Buck. /28/ The medical community has found it useful to apply the "disease" label to alcoholism for reasons that have nothing to do with the administration of a benefits program. See Fingarette, The Perils of Powell: In Search of a Factual Foundation for the "Disease Concept of Alcoholism," 83 Harv. L. Rev. 793, 809-812 (1970). One researcher has summarized these reasons (Blume, Public Policy Issues, supra, at 186-187): Acceptance of alcoholism as a disease is the basis for public policies promoting early intervention; treatment; third-party coverage, such as public and private health insurance; services for families of alcoholic persons, with or without the participation of the alcoholic member; prevention services for children of alcoholic parents; and wide-ranging research into the biology, epidemiology, psychology, and sociology of alcoholism. /29/ See note 21, supra. "(M)any authorities flatly oppose (the "loss of control") orientation, and, while recognizing that there may be physical factors at work, consider it essential that 'there is a whole series of voluntary actions in the act of drinking; and there has to be a choice involved.'" Fingarette, supra, 83 Harv. L. Rev. at 808 n.66; see DSM-III at 169 (diagnostic criteria for alcohol abuse include "continuation of drinking despite a serious physical disorder that the individual knows is exacerbated by alcohol use"). /30/ A survey of the literature in 1970 concluded that even authorities "who use 'loss of control' language ultimately introduce serious qualifications. * * * (W)e are told not that the alcoholic has no control of his drinking, but that he has greater or lesser control, widely varying in degree according to the circumstances and the individual." Fingarette, supra, 83 Harv. L. Rev. at 802 (emphasis in original). More recent authority, cited in this case as supporting the "disease" concept, bears out this description. Thus, one writer states (Wolf, Alcohol and Health: The Wages of Excessive Drinking, in Alcoholism and Related Problems: Issues for the American Public 29 (1984), cited by amicus National Association of Addiction Treatment Providers (Br. 11 n.6)): There has been much dispute, but no consensus, about the various parts played by genetic inheritance, the social pressures, special features of individual growth and maturation, and the insidious effect of ethanol itself in the cause of alcoholism. Apparently, all are pertinent, and the behavioral characteristics of alcoholics are partly innate and partly shaped by learning or practice. See also S. Zimberg, supra, at 8: Alcoholism can best be considered a disease with a multiple causality. The factors believed to contribute to the development of alcoholism are both psychological and sociocultural, along with physiological factors which probably operate on a genetic basis. The "disease" approach to alcoholism has itself been substantially revised (Kissin & Hanson, The Bio-psycho-social Perspective in Alcoholism, in Alcoholism and Clinical Psychiatry 12 (J. Solomon ed. 1982) (footnote omitted)): In the old disease concept of alcoholism as promulgated by Jellinek and expanded upon by both psychoanalysis and Alcoholics Anonymous, the alcoholic was seen as more or less preordained to become alcoholic and to remain quintessentially alcoholic even after he became abstinent. * * * (T)he old disease concept of alcoholism has largely been discarded. In the new approach to the disease concept a variety of influences -- physiological, psychological, and social -- are seen as contributing to the development of problem drinking * * *. /31/ The studies cited by amici to support the genetic component of alcoholism concede that the genetically-predisposed alcoholic retains volition in the sense of being susceptible to outside authorities telling him he should not drink (Bohman, et al., Maternal Inheritance of Alcohol Abuse: Cross-Fostering Analysis of Adopted Women, 38 Arch. Gen. Psychiatry 965 (1981), cited by Amicus National Council on Alcoholism, Inc. (Br. 25 n.13); see Cloninger, supra, at 867): (T)he critical importance of sociocultural influences in most alcoholics suggest that major changes in social attitudes about drinking styles can change dramatically in the prevalence of alcohol abuse regardless of genetic predisposition. The most recent study cited by the amici also states that even among genetically predisposed individuals, "personal values and intentions" play a role in excessive drinking. Peele, The Implications and Limitations of Genetic Models of Alcoholism and Other Addictions, 47 J. Studies on Alcohol 63 (1986). /32/ The AA treatment has also been described as a way of "transform(ing) externalization of responsibility into self-responsibility." G. Vaillant, supra, at 204; see S. Zimberg, supra, at 118 ("The only requirement for AA membership is a desire to stop drinking."). "'There is no longer any question that Alcoholics Anonymous has been responsible for the sobriety of more alcoholics than any other method, social, religious, or medical. Psychiatrists are now "believers."'" M. Hayman, Alcoholism -- Mechanism & Management 174-175 (1966), quoted in Fingarette, supra, 83 Harv. L. Rev. at 807 n.64. One of the studies cited by amici acknowledges that while the data are inconclusive, AA treatment as well as "personal resolve" has helped cure many alcoholics (West, Alcoholism and Related Problems: An Overview, in Alcoholism and Related Problems: Issues for the American Public 17 (1984)): AA is known to have helped hundreds of thousands of alcoholics to maintain sobriety. Other alcoholics, even without the help of AA, have also been able to remain sober through personal resolve reinforced by family support, religious conversion experiences, psychotherapeutic intervention, and effective contact with medical facilities and health personnel. /33/ The role of self-responsibility in the cure of alcoholism is also stressed in S. Zimberg, supra, at 67-69: The patient should be told that the therapist, the patient's spouse, or anyone else cannot stop a problem drinker from drinking. Only the patient's own efforts, utilizing the treatment tools, will result in a recovery from the problem drinking. * * * * * The patient's recovery ultimately depends on his weighing the benefits of sobriety against the "benefits" of drinking. Only the patient can make this determination. /34/ This Court's decision in School Bd. of Nassau County v. Arline, supra, does not require the conclusion that the Rehabilitation Act prevents the VA from carrying out the "willful misconduct" test for G.I. Bill extensions and disability pensions imposed by the veterans' benefits statute. The Court's opinion in Arline concluded (slip op. 9) that the purpose of the Rehabilitation Act is to assure that handicapped individuals are not denied benefits because of "prejudiced attitudes or * * * ignorance" or "society's accumulated myths and fears" (ibid.). In this case, unlike Arline, the VA is implementing a specific congressional decision to utilize a "willful misconduct" standard; and it has done so in a manner that accepts psychiatric testimony in individual cases to determine whether a psychiatric disorder is involved in the particular case. Thus, the reconciliation here of the Rehabilitation Act with a more specific congressional directive -- a matter not addressed in Arline -- is fully consistent with the Court's rationale in that case. APPENDIX