The ALJ has recounted the facts thoroughly. Recommended Decision and Order (R. D. and O.) at 2-21. Briefly, Complainant Thomas J. Strawser has been employed by Defendant Exxon Company, U.S.A. (Exxon) since 1981. He began work as a senior engineer in Oklahoma City, Oklahoma. In 1982 he was transferred to the company's Dover-Hennessey Gas Plant located north of Oklahoma City. He then worked as a subsurface engineer assigned to the company's Oklahoma City office and in late 1983 became the temporary field foreman at the Hewitt Field in Wilson, Oklahoma. In 1984 Strawser returned to the Dover-Hennessey Gas Plant as temporary field foreman, and in October of that year he joined the start-up team for a natural gas extraction and processing facility in LaBarge, Wyoming. Although stationed initially in Oklahoma City, the start-up team relocated to LaBarge in March 1985. In his start-up capacity Strawser reviewed processes, instrumentation, pressures, flow rates, metallurgy and blueprints. He also wrote job descriptions for the operations team, trained supervisors and checked vehicle specifications.

   In order for individuals to recover under section 503 of the Rehabilitation Act, they must meet two different criteria which initially may appear contradictory. They must show first, that they are an individual with a disability and second, that despite their disability they can perform the particular duties required for the job. Strawser has shown in a number of different ways that he is an individual with a disability and that despite his disability he is capable of performing the particular duties of the job.

   I. Qualified individual with a disability

Arline, 480 U.S. 273, 283 (1987).

despite the disease, was "qualified."

occur, [3] that he as a result of drinking would take some inappropriate action (or fail to take appropriate action), and [4] that that action or inaction would cause a catastrophe." R. D. and O. at 36, citing Dep't of Labor, OFCCP v. Texas Industries, Inc., 47 Fair Empl. Prac. Cas. (BNA) at 25-26 (this type of formulation represents a "tenuous prediction" which does not begin to approximate the analysis suggested by E.E. Black, Ltd. v. Marshall).

inappropriate action and catastrophe is exponentially lower. . . . The probability of harm is reduced even further if Exxon monitors Strawser's condition through periodic medical examination and random testing.

84-OFC-17, Acting Ass't Sec. Rem. Ord., Jul. 27, 1993, slip op. at 12 (interpreting Mantolete language to "refer to situations that are very clear, evident and obvious, and not subject to serious dispute"). See also Bentivegna v. United States Dep't of Labor, 694 F.2d 619, 621-622 (9th Cir. 1982) (business necessity should not be confused with expediency). Thus, substitution of categorical exclusion for individual evaluation requires that all or substantially all of the individuals with the disability be unable to perform the job safely. Chandler v. City of Dallas, 2 F.3d 1385, 1395 (5th Cir. 1993) (drivers with insulin-dependent diabetes mellitus or extreme vision impairment presented genuine substantial risk of injury as a matter of law under Federal and municipal medical standards for primary transportation jobs). Cf. Rosenfeld v. Southern Pacific Co., 444 F.2d 1219, 1225 (9th Cir. 1971); Weeks v. Southern Bell Telephone Co., 408 F.2d 228, 235 (5th Cir. 1969) (absent proof that entire class of individuals is unable to perform a job, employer may exclude applicant only by showing individual incapacity).

   We disagree with Exxon's premise that a relapse is almost impossible to predict. The ALJ observed:

[T]he evidence establishes that alcoholics, unlike epileptics and diabetics, experience warning signs before they relapse; that the longer an alcoholic remains sober, the less likely he is to relapse; that job problems are the last to appear when an alcoholic relapses, and thus a progression toward alcoholic drinking can be detected long before any job problems appear.

R. D. and O. at 30. Both Drs. Torres and Peoples testified to this effect. T. 330-335, 388-393, 710, 917-918, 929-932; R. D. and O. at 13-15.

program), the success rate for recovering pilots was 85.5% in 1985. The success rate in 1992 was 90-94%. "Success" was defined as having no relapses over a two-year period following return of medical certification. The exhibit explains that the FAA is responsible for issuing pilots medical certificates which document their medical fitness. Each pilot must have a current medical certificate to validate required piloting certificates. Alcoholics may not be certified. If, however, an alcoholic is rehabilitated sufficiently, he may be granted an exemption or "special issuance." The FAA requires that a responsible medical source sponsor any request for a special issuance and imposes a period of surveillance following recertification. Once medical certification is restored, continued certification is contingent on maintaining abstinence and receiving favorable reports from monitoring sources during a 24-month period.

All alcoholics . . . are not the same. There are differences, and there are ways to predict relapse. . . . You go through a person's history. You look them in the eye. You challenge their history of drinking. You do a physical exam on them. You do blood tests if necessary. [Y]ou get a good history of how they're proceeding through treatment, how they're responding to therapy . . . . And you get thorough histories of job performance, family relationships, legal problems . . . . And by looking at all of those different factors, you come up with a judgment as to whether . . . someone is going to relapse or not. As I stated earlier, [Strawser] has a very low risk of relapse.

T. 917-918.

   The ALJ's analogy to Davis v. Meese, 692 F. Supp. 505, 511 (E.D. Pa. 1988) is apposite. There, "the FBI acknowledged that in the category of diabetics, there is a subgroup of individuals [non-insulin dependent diabetics] whose handicap does not pose the same risk as the others." R. D. and O. at 38. Accordingly, the FBI considered those diabetics individually. The ALJ reasoned:

[I]t would appear that the same is true in this situation. Based on the strength of Strawser's recovery and his commitment to continued treatment, it appears that he belongs to a subgroup of recovering alcoholics, namely those with a strong recovery, who can safely perform the designated job in question.

Id. at 38-39. In contrast, a "binge" drinker with no history of abstinence and only recently released from treatment likely would not be appropriate for high risk employment because a relapse would be difficult to predict. T. 389-391 (Dr. Torres). See Altman v. New York City Health and Hospitals Corporation, 903 F. Supp. 503 (S.D.N.Y. 1995)(risk of relapse unacceptably high where physician, who was a binge drinker, had been visibly inebriated while treating patients and only recently had undergone rehabilitation). This determination rests on an individualized assessment of the individual's history, however.

   The ALJ used many of the studies as "a frame of reference" but decided that they "ultimately [were] not conclusive for purposes of this case." He faulted them for relying on self-reporting and noted that those which showed a high rate of relapse relied on data for "rehabilitated individuals during the first year or two after treatment." R. D. and O. at 19, 40; T. 934-935, 1146; DX 48F. "Different studies often had conflicting results." Compare DX 48C with DX 48F at 5 (conflicting rates of inpatient abstinence and relapse). He discussed DX 48I where 29 of 100 alcoholics had been abstinent for three years or more. The study was "significantly biased toward severity of illness," DX 48I at 1148, meaning that of the individuals sampled, 87% had abused alcohol for ten years, 80% had undergone detoxification and 95% required drugs during withdrawal. He questioned whether the study was "representative of the general population of substance abusers . . . ." R. D. and O. at 20.

   DX 48J followed alcoholics for ten years after treatment. A total of 77% of the individuals reporting remission at two years also reported remission at ten years. DX 48J at 50. The criteria for remission permitted degrees of alcohol use, however. Id. at 48. PX 94, the AA

study, also was long-term. Between 20 and 30% of the sample maintained more than five years of sobriety. PX 94 at 10. A total of 91% of individuals who maintained sobriety five years or more remained sober and active in AA for another year. Id. at 22. This study, which began in 1968 and ended in 1989, relied on "completed questionnaires for about 100,000 people." T. 934. According to Dr. Peoples, however,

[i]t's not a particularly well done study, but it does have very, very huge numbers in it. One of the problems with drug addicts and alcoholics is that you lose them to follow up -- you don't have basically good ways of doing long term studies. You can study the first year, the first couple of years, but to get out to many years is a very difficult thing to do.

T. 933-934.

   The ALJ ultimately concluded that this evidence "d[id] not appear reliable, as many of the studies introduced either contradict each other or are not representative of either the general population or of recovering alcoholics with histories similar to Strawser's." R. D. and O. at 40. See T. 919 (Dr. Peoples)("One of the problems of many studies that are done on alcoholism is that they're done on very skewed populations. They're done on people off the street, or . . . on doctors, or . . . on individual groups that don't necessarily apply to the entire disease."). We also note that many of these studies are dated or rely on dated data and thus do not measure the results of more agressive intervention techniques and treatments instituted in recent years.

   In contrast, the ALJ considered the FAA studies to be relevant because they were grounded on an industry-wide monitoring program for recertified pilots. R. D. and O. at 20-21; PX 71 at 15, 20; PX 95. As discussed above, these studies show successful rehabilitation figures of 85.5% and 90-91%. The 1985 study included in its success rate pilots who required a second special issuance, meaning that they may have relapsed after treatment. A total of 79% of those receiving a special issuance had not relapsed since recertification, however. Dr. Torres testified that "[the monitoring program's] success rate in maintaining long-term sobriety is quite impressive." T. 710.

   Although much of the "survey science" is problematic, we rely in part, as did the ALJ, on that which appears most reliable and relevant to Strawser's circumstances, i.e., the FAA studies. In Strawser's case, we have access to his "behavioral science" contained in fitness for duty evaluations, and we rely on it also. Strawser's history is comprehensive. At the time of transfer in September 1989, Strawser's sobriety dated from January 1, 1983. After assignment to the LaBarge gas facility in 1985, Strawser spent five years working successfully and safely in a high risk occupation. At no time during his history with Exxon, beginning in 1981, did alcohol use or abuse affect his employment. A 90% relapse rate simply does not pertain to this individual at this stage of recovery.

   Drs. Torres, Peoples and Gendel agreed. Dr. Peoples testified:

[Strawser's] alcoholism is in remission. He is not drinking. He is working an active program. He is attending 12 step support meetings and has a support [system]. He has done activities that demonstrate his commmitment to AA and to recovery. He also is taking his lithium for his bipolar illness. His recovery is very strong because of the strength of his commitment to recovery. My opinion basically is based on my history, my physical exam, my experience with other alcoholics, my review of . . . his medical records, his personnel file, letters that he has written. . . . From my research and from my examination of him, I feel that he has very little risk for relapse at this point.

T. 908-909. Dr. Torres testified similarly:

[Strawser is an individual] who is highly invested in recovery who has the resources to maintain his recovery, who has resources to deal with adverse situations. [H]is way of life is replete with influences, relationships, choice principles, values, beliefs, traditions and practices where sobriety is essential . . . . I see no evidence that he is disposed towards relapsing. He seems to have a very strong recovery, and that was observed as well by Dr. Gendel, by Dr. Peoples.

T. 775-776.

   The dissent asserts that Strawser disclosed the continuing "allure" of alcohol by confiding to a psychiatrist his "strong desire for a drink." As basis for the assertion, the dissent cites to PX 68 which includes a transcription of Strawser's medical records covering 12 separate meetings. A November 1990 entry concludes: "Feels a drink would be nice.'" In this entry, Strawser also related that he had filed the instant Rehabilitation Act complaint against Exxon, that he expected to look for other employment and that he had experienced "a lot of anxiety, inner turmoil re[garding] this." Dr. Torres testified that he had examined the "drink would be nice" comment in assessing Strawser's fitness for duty. He also testified that the comment signified that "[Strawser] is honest, and he is observant of his relapse potential. He is not trying to look good to the doctor. He is letting it all out." T. 407. In Dr. Torres's opinion, the comment did not indicate "imminence of relapse." Id.

   The dissent also attributes Strawser's request to be relieved from shift work to his "fear of alcoholic relapse." In 1987 and early 1988, Strawser was assigned temporarily to the Shute Creek gas processing facility on a "shift work" schedule, i.e., seven days on, seven days off, seven nights on, seven nights off. The assignment also involved a considerable commute. During this period, Strawser experienced severe chest pains and was diagnosed with viral pneumonia. He became dissatisfied with the assignment because it caused problems with his health and interfered with AA attendance. At his request, Strawser was returned to the field foreman position at the LaBarge DeHy in June 1988. Strawser's physician was "concern[ed] that shift work posed a problem for [Strawser's] health." R. D. and O. at 4. Strawser explained to the supervisor at Shute Creek that although he did not currently have a drinking problem, he "didn't want to take any chances." T. 63. With regard to Strawser's request to return to LaBarge, Dr. Torres testified: "[I]t shows good judgment [and] good problem solving skills in that he is aware that this kind of activity stands a significant risk of being detrimental to his recovery, and he took

   Strawser's actions following the use of four ounces of alcohol during the 1991 hunting trip also demonstrate the strength of his recovery. Despite the fact that no one witnessed the lapse, Strawser reported it to his AA sponsor and support group immediately upon his return. He increased his AA activities and with the help of his sponsor "examined the reasons for his lapse and came to some understandings of what changes he needed to carry out in order to bolster his recovery." T. 402 (Dr. Torres). Dr. Torres testified that Strawser's "investment and his way of life is steeped enough in recovery that it . . . promotes continued recovery and . . . serves as a . . . very strong corrective should some drinking occur as it occurred that day." T. 378. Dr. Gendel concurred:

[W]hen one is working with alcoholics in recovery, one learns that there is a point at which people reach a kind of self-sustaining recovery process where they are not working every day not to drink, where the recovery lifestyle is integrated, where the things that support recovery are part of everyday functioning and where, as in this case, a drinking episode is followed rapidly by . . . the use of the tools that support recovery, where it's much more likely that someone will remain free of relapse, and [Strawser is] certainly in that group.

Gendel deposition at 100-101. Dr. Peoples testified that "[i]t does not take a real genius to look at somebody who [in] nine years has had four ounces of alcohol and basically come to the conclusion that he's pretty unlikely to drink." T. 918. What distinguishes Strawser is the capacity to integrate the tools of recovery into his existence. That other recovering alcoholics, who are responsible for elevated relapse rates present in some studies, may not share this capacity should not effectively render Strawser "disabled."

   Exxon asserts that no individual characteristic of a particular alcoholic will serve to reduce the relapse rates shown by the studies. Exxon also asserts that Strawser's work and medical histories are irrelevant and that the chances are between 30% and 90% that on any given day he will report to work impaired. These assertions belie both the expert testimony of Drs. Torres, Peoples and Gendel and common sense. If the relapse rate of any alcoholic really were a totally random 50%, for example, then the chance that Strawser will report to work impaired is the equivalent of a coin toss. Strawser reported to work unimpaired every single day since 1981 -- a period of nine years to date of transfer. This history translates to roughly 2,500 straight coin flips landing heads up. History notwithstanding, Exxon asserts that the chances are still 30-90% that Strawser will show up for work tomorrow impaired. We simply cannot agree. The most reliable predictor of how Strawser will perform his job tomorrow is how he performed his job over the past nine years.

   The relevant FAA studies hardly substantiate categorical exclusion of all rehabilitated alcoholics from positions designated safety-critical. R. D. and O. at 40-41. Rather, categorical exclusion is an expedient means of avoiding any risk where individualized assessment would distinguish between those persons who have rehabilitated themselves successfully and those who have not. At bottom, Exxon's "never-ever" policy is based on a judgment that rehabilitated alcoholics are forever disposed to relapse, certainly a "myth, fear or stereotype" associated with alcoholism. In the instant case, the reality is contrary -- for an individual like Strawser who has

maintained sobriety for years, any fear of relapse is not well-grounded.

   We decline to abandon the section 503 requirement of individualized inquiry in these circumstances. R. D. and O. at 38-44. The Exxon Valdez notwithstanding, section 503 of the Rehabilitation Act mandates that Exxon not behave like the cat that sat on a hot stove. That cat will never sit on a hot stove again. Of course, the cat will never sit on a cold stove either. Section 503 requires an evaluation of the temperature of the particular stove, not just a blanket policy to avoid all stoves.

   Finally, Exxon asserts that Strawser's transfer to a non-safety critical position constituted reasonable accommodation. To the contrary, the ALJ found that Strawser was qualified to perform the job of field foreman at the LaBarge, Wyoming, gas facility without accommodation. R. D. and O. at 28-31 (Strawser had performed the job successfully and safely for five years prior to transfer and continued to function as a field foreman at Gillette thereafter.).

   Transfer to the Gillette oil field was discrimination, not "accommodation." Exxon regarded Strawser as being disabled, whereas in reality he was able to perform the job at LaBarge as well as any unimpaired individual with the requisite training and experience. These circumstances "are analogous to capable workers discriminated against because of their skin color or some other vocationally irrelevant characteristic." Vande Zande v. State of Wis. Dep't of Admin., 44 F.3d 538, 541 (7th Cir. 1995). The issue of whether the discrimination includes an employer's failure to make reasonable accommodation arises only where the disability "interferes with the individual's ability to perform up to the standards of the workplace . . . ." Id. As the court in Mantolete v. Bolger recognized, "two questions must be asked," the first being whether the individual presently is "qualified to perform the essential requirements of the job without a reasonable probability of substantial injury to [himself] or others." 767 F. 2d at 1423. If the response to this question is affirmative, "then employment cannot be denied based upon the handicap." Id. Only if the response to the first question is negative does a second question concerning reasonable accommodation need to be asked. In Strawser's case, and based on the instant record, the response to the first question clearly is affirmative and, therefore, responding to the second question is unnecessary.

peer identification." R. D. and O. at 16 (footnote omitted).

   An individual with a disability is responsible for informing the employer that accommodation is necessary. The regulations provide:

The contractor shall invite all applicants and employees who believe themselves covered by the Act and who wish to benefit under the affirmative action program to identify themselves to the contractor. The invitation shall state that the information is voluntarily provided, that it will be kept confidential, that refusal to provide it will not subject the applicant or employee to any adverse treatment, and that it will be used only in accordance with the Act and the regulations in this part. If an applicant or employee so identifies himself or herself the contractor should also seek the advice of the applicant or employee regarding proper placement and appropriate accommodation.

R. D. and O. Appendix (emphasis added). Strawser's performance was evaluated by his supervisor regularly. Both Drs. Torres and Peoples testified that Strawser needed little, if any, monitoring. R. D. and O. at 31. Dr. Torres testified: "I don't believe that Mr. Strawser in particular needs to be monitored. His sobriety is very stable. His commitment to maintaining his lithium treatment is very stable. His responsiveness to directions from treatment personnel is stable." T. 681. Dr. Torres also testified that Strawser was fit for duty in a high risk job, including a job that involved no direct supervision. T. 378-380. Strawser attended AA on his own time and without Exxon's assistance. In short, Exxon was not required to make any modifications or adjustments "in its ordinary work rules, facilities, terms, and conditions" of employment to enable Strawser to work. Vande Zande v. State of Wis. Dep't of Admin., 44 F.3d at 542. Accordingly, Exxon was not entitled to choose the means of accommodation, i.e., involuntary transfer, because Strawser did not require any accommodation.

current assignment would pose an undue hardship. Transfer to Gillette limited Strawser's promotional opportunities and reduced his status from field foreman at a state-of-the-art gas facility to foreman at an oil field -- additional considerations militating against transfer.

   III. Remedy

   The ALJ recommended that Strawser be offered reinstatement to the position of field foreman at the LaBarge, Wyoming, facility with the seniority and pay he would have received had he not been transferred and that Exxon reimburse him for moving costs and the loss realized on the sale of his house. The ALJ rejected Plaintiff's argument that Mrs. Strawser should be reimbursed for lost wages. We agree with and adopt these findings. R. D. and O. at 11-12, 44-45.

   The ALJ declined to recommend any further action to abate the violation. Plaintiff excepts to the omission, arguing that Exxon should be directed to discontinue its policy of categorical exclusion. As Plaintiff points out, the Rehabilitation Act authorizes the Department of Labor to "take such action" on any complaint of noncompliance "as the facts and circumstances warrant, consistent with the terms of [the contractor's] contract and the laws and regulations applicable thereto." 29 U.S.C. § 793(b). Certainly an order directing Exxon to discontinue a policy that violates the affirmative action/nondiscrimination mandate of section 503 is an "action" which is "consistent with" the Rehabilitation Act.

   Plaintiff also argues that the authority to order a party to cease discriminatory practices exists under analogous remedial legislation such as Title VII of the Civil Rights Act of 1964. Dothard v. Rawlinson, 433 U.S. 321, 331-332 (1977). Cf. U.S. Dep't of Labor v. Honeywell, Inc., and Teamsters Local 1145, Case No. 77-OFCCP-3, Sec. Dec., June 2, 1993, slip op. at 30-34 (broad remedial authority under Executive Order No. 11,246 and 41 C.F.R. § 60-2.1).

   The ALJ recognized that while Exxon "was attempting to deal with the problem of alcoholism and drug abuse in the workplace and the serious human and environmental damage that that problem can cause," the nondiscretionary policy implemented to address the problem was "impermissibly inflexible." R. D. and O. at 43. In particular, the "across-the-board policy prohibiting rehabilitated individuals from holding designated positions does not differentiate between those who have been successful in rehabilitating themselves and those who have not." Id. at 43-44. It thus violated the section 503 "mandate of affirmative action and non-discrimination in employment," and absent application of the policy "on a case-by-case basis," Exxon risked future violation. Id. at 44.

   The ALJ noted, however, that implementation of the policy in conformance with certain of Exxon's directives could permit individualized inquiry. The directives state that any employee working in a designated position must report "1) past, present or future alcohol or substance abuse, including participation in rehabilitation programs; 2) arrests or traffic tickets received while under the influence of alcohol or drugs; 3) arrests for public intoxication, possession, sale or distribution of controlled substances; and 4) any other incident involving alcohol or drugs." Id. at 42. The ALJ observed:

It appears from these directives that, based on factors such as how long ago the incident occurred, how many incidents there were, and what job the employee held at the time of the incident, an employee with a substance abuse history could nevertheless be retained in a designated position. . . . Exxon nevertheless has contended throughout this proceeding that Strawser's transfer out of his designated job was required under the 1989 policy. Regardless, it is clear that no thought was given to maintaining Strawser in his designated position at [the] DeHy when the policy was implemented.

Id. at 42-43.

   Individualized inquiry would not be burdensome. Of the 3,000 employees working in designated positions, "if they reflect the population as a whole, approximately 10% will be alcoholic." R. D. and O. at 42. Because of Exxon's pre-employment screening, the percentage may be less. (According to Exxon, only about 40 employees were transferred when the policy took effect.) Exxon would be required to evaluate only the individuals whom it knew to have histories of substance abuse, rather than all 3,000 employees filling designated positions. Id.

   In these circumstances, we deem it appropriate to direct Exxon to discontinue the policy of categorically excluding individuals with a history of alcoholism from consideration for designated positions.

   We find that Complainant Thomas Strawser is a qualified individual with a disability under section 503 of the Rehabilitation Act and that Defendant Exxon Corporation, d/b/a Exxon Company, U.S.A., failed to demonstrate that its 1989 Drug and Alcohol Policy, as applied to Strawser, is consistent with business necessity and safe job performance. Exxon thus violated section 503 of the Rehabilitation Act when it precluded Strawser from working in the designated position at issue.

of section 503 and the rules, regulations and orders issued thereunder which have been found to have been violated in this case.

   SO ORDERED.

       DAVID A. O'BRIEN
       Chair

       JOYCE D. MILLER
       Alternate Member

Member Karl J. Sandstrom, dissenting:

   This case presents the Board with the difficult task of identifying the unobvious point where the law strikes the balance between public safety and the rights of disabled individuals in the workplace. Where one yields to the other is a not a question committed to the Board's discretion. Rather, our task is to examine the statute and its regulations in the light of the case law and determine what the law requires of employers in choosing between competing public demands. If the law permits an employer to take one course of action, even if our judgment is that the public would be better served by a different employment practice, our obligation is to affirm the employer's decision. For the reasons more fully given below, I depart from the Board's decision because I find that the law affords employers more leeway in making these decisions than the Board recognizes.

   I begin my dissent in agreement with the Board's finding that Complainant is for the purposes of Section 503 of the Rehabilitation Act an "individual with a disability." Even without taking into consideration Complainant's bi-polar disorder and its connection to or influence upon Complainant's alcoholism, Complainant's history of alcoholism alone is sufficient to establish that he is an individual with a disability. See Tinch v. Walters, 765 F.2d 599, 603 (6th Cir. 1985) (a recovered alcoholic is an individual with handicaps within the meaning of the Rehabilitation Act). That alcoholism is a serious disease that can substantially impair an individual's mental and physical

functioning is beyond dispute. Exxon offered substantial and unrefuted scientific evidence on this very point. See DX 48G, 48H and 48MM. It is then somewhat disingenuous for Exxon to contend that it does not consider Complainant's alcoholism to be a covered disability. The fact that Exxon does not consider alcoholism to be a disqualifying condition for non-safety related jobs is better evidence of its general compliance with its obligation under the Rehabilitation Act than it is proof that it does not consider alcoholism to be a disability. Regardless of whether Exxon considers alcoholism to be a disability, it is well settled for the purposes of Rehabilitation Act that it is a covered disability. Teahan v. Metro-North Commuter Railroad Company, 951 F.2d 511, 517 (2nd Cir. 1991).

   Because I find that alcoholism can manifest itself in the substantial impairment of physical and mental functioning, I find no reason to conclude that Complainant was a covered individual with a disability because Exxon regarded him as having a disability. Complainant is a covered individual because he suffers from the disease of alcoholism. Complainant's admirable and sustained efforts to control his alcoholism only underscore the terrible and persistent nature of this disease. To suggest, as the majority does, that Complainant is a covered individual largely because Exxon regarded him as such is to diminish Complainant's own efforts to control his disease and Exxon's legitimate public safety concerns. Once it is established that an individual suffers from a disability, there is no reason to proceed with an analysis of whether the defendant regarded the individual as disabled. At this point of the analysis, defendant's state of mind is irrelevant. The focus instead should turn to second step of the required legal analysis and whether the individual is "otherwise qualified" with or without reasonable accommodation. School Board of Nassau County, Florida, et al v. Arline, 480 U.S. 273, 287 (1987).

   It is at this stage of the analysis that I part company with my colleagues. The majority has concluded that Complainant must be permitted to return to his previous position without the imposition of any accommodation. I question the confidence with which the majority issues this finding, because immediately upon rendering it, the majority notes Exxon's failure to demonstrate that an accommodation by means of testing, supervisor evaluation and continued AA attendance would constitute an undue hardship. Why even suggest an accommodation where none is necessary unless one entertains doubts about the initial determination. Inasmuch as the majority later in its decision concludes that Complainant could refuse the imposition of any accommodation, Exxon is left to wonder why the Board would speculate about an accommodation which the

company could not legally impose. Because I believe the majority's decision is constructed on a very shaky legal and factual foundation, I suspect that the majority speculates about possible accommodations because it fears that its finding that no accommodation may be imposed on Complainant may be upset on appeal. A review of facts of this case demonstrates why such fears would be well grounded.

   Complainant was employed by Exxon as a Field Foreman at a major natural gas field in Wyoming. In this largely unsupervised position Complainant was responsible for the development and -- in the event of an accident -- the implementation of contingency plans to protect the safety of the personnel working at the plant and the people of the surrounding community. It is undisputed that an accidental leak at this field could result in a major disaster, endangering the health and lives of the people working there and those living in the surrounding area. Nor is it disputed that Complainant's position was critical to the prevention and containment of such accidents. Concededly, the possibility of such an accident was remote if for no other reason than such fields would not be legally allowed to operate if they posed an imminent threat to the public safety. For that matter, the accident at Three Mile Island prior to its occurrence would have been deemed a remote possibility. It is the catastrophic consequences not the likelihood of an accident that justify great and costly precautions in these situations.

   Complainant is a recovering alcoholic. He was first diagnosed to be an alcoholic in 1980. Two of his siblings are also alcoholics. He began receiving treatment for his alcoholism in 1981. Complainant entered a treatment center in 1981 for a twenty-eight day program and became active in Alcoholics Anonymous. Complainant continued to experience difficulty in achieving sobriety. His AA sponsor noticed that he exhibited certain manic tendencies and speculated that these tendencies might have been contributing to his inability to achieve sobriety. At his sponsor's suggestion, he reentered the center for a five day treatment program. During treatment for his alcoholism, Complainant was diagnosed as suffering from bi-polar disorder, which is also known as manic depression. His physician prescribed lithium to control his manic depression and to aid in his quest for sobriety.

to sustain sobriety while at LaBarge. Complainant continued his attendance at AA meetings and followed AA precepts while at LaBarge. His fear of alcoholic relapse did lead him to request along with his doctor that Exxon accommodate him by taking him off shift work. Exxon honored this request. When his manic symptoms reappeared in 1990, Complainant sought medical advice and was put back on lithium. He remains on lithium to this day. During his treatment, he disclosed to his doctor the continuing allure of alcohol. Complainant also admits to drinking during the pendency of this case while on a hunting trip in October of 1991.

   Under these facts, I am unable to find that Complainant without accommodation is to be considered as a matter of law to be otherwise qualified under section 504. "An otherwise qualified person is one who is able to meet all of a program's requirements in spite of his handicap." Southeastern Community College v. Davis, 442 U.S. 397, 406 (1979). This includes any requirement reasonably designed to protect the public safety. As the Fifth Circuit held in Chiari v. City of League City, 920 F.2d 311, 316 (5th Cir. 1991): "a significant risk of personal injury can disqualify a handicapped individual from a job if the employer cannot eliminate the risk."

   In Bradley v. University of Texas M.D. Anderson Cancer Center, 3 F.3d 922 (5th Cir. 1993), the circuit court had occasion to expand upon its analysis of the relationship between safety demands and a finding that a person is "otherwise qualified." The case involved a surgical assistant who, upon revealing that he was HIV positive, was transferred to another position in the hospital. In upholding the transfer from a challenge under the Rehabilitation Act, the court found, at 924:

While the risk is small, it is not so low as to nullify the catastrophic consequences of an accident. A cognizable risk of permanent duration with lethal consequences suffices to make a surgical technician with Bradley's responsibilities not "otherwise qualified."

   These cases are in accord with the regulations implementing Section 501 of the Rehabilitation Act which define a "qualified handicapped person" as one who "can perform the essential functions of the position in question without endangering the health and safety of the individual or others . . . ." C.f. 29 C.F.R. § 1613.702(f).

   Measured against this standard, I believe Exxon is fully

justified in conditioning Complainant's continuing employment on his acceptance of a reasonable accommodation designed to satisfy Exxon's legitimate safety concerns. Although it is undisputed that Complainant's education and experience provide him with the requisite skills to function as a field foreman, and the record demonstrates that he has done so for five years in an exemplary manner, the risk of relapse remains ever present, as do the catastrophic consequences that could accompany an accident attended to by an individual in an impaired condition. Exxon has an obligation, as did the defendant in Altman v New York City Health and Hospitals Corporation, 903 F. Supp. 503 (1995) (Hospital refused to reinstate alcoholic physician as Chief of its Department of Internal Medicine), "to weigh carefully the public safety concerns that an undetected relapse might entail." at 509.

   In finding that Complainant requires no accommodation, the majority largely ignores the testimony of Complainant's own consulting experts. Dr. Torres testified that Complainant should be monitored by his supervisors, should be subject to random testing, should be required to attend AA meetings and continue under psychiatric care. (T. 686-689). Dr. Peoples similarly recommends alcohol and drug screening and monitoring. (T. 997-1002). Lastly, Dr. Gendel, who was consulted regarding the issue of bi-polar disorder in relationship to Complainant's alcoholism notes the need for monitoring and psychiatric treatment. (PX 65). These recommendations are rooted in a recognition that even though an individual with bi-polar disorder may have his alcoholism under control, there is no cure for either alcoholism or bi-polar disorder. Continuing attention is therefore required and in those instances where public safety is implicated, employer action is necessary.

    In Hogarth v. Thornburgh, 833 F. Supp. 1077 (S.D.N.Y. 1993), the court was confronted with the question whether the F.B.I. could terminate a clerk suffering from bi-polar disorder if his condition were controllable with medication. Judge Francis wrote at 1083:

The second consideration -- the probability of a relapse -- also favors the Government. Mr. Hogarth's psychiatrist testified that once his condition had been controlled with medication the chances of a recurrence were minimal. The Government successfully rebutted this argument in several ways. First Dr. Siegle testified that lithium therapy is no longer viewed as a guarantee against recurrence of bipolar disorder as it once was. Second, the effectiveness of any regimen of medication depends on the patient's compliance.

On this basis, along with evidence of Mr. Hogarth's post-termination relapse, the court found the plaintiff was not "otherwise qualified". In the instant case Exxon did not terminate Complainant but seeks only to impose certain accommodations to protect the public safety on his continuing employment.

   To find that Complainant requires no accommodation, the majority must downplay the significance of Complainant's volitional use of alcohol during the pendency of this matter. According to his own testimony, Complainant's resort to alcohol was the result of stress or in his own words "everything just got to me . . . ." (T. 255). I cannot agree with the majority that Complainant's "lapse" ultimately demonstrated his control. For me it demonstrates the continuing allure, particularly in times of stress, that alcohol holds for him. This conclusion is supported by other actions of Complainant. To aid in his recovery, Complainant upon doctor's advice, sought relief from shift work, a request which Exxon honored. In November 1990 Complainant confided to his doctor his strong desire for a drink. (PX 68). At that time his physician ordered additional medication to control his tension and anxiety. It is clear then from the evidence that Complainant's recovery is not complete. Under the circumstances Exxon's effective options are limited. Exxon is unable to determine the amount of stress in Complainant's life or guarantee Complainant's compliance with his recovery plan; therefore, to protect the public and its employees, Exxon is restricted to taking those actions that are within its control.

   Exxon's fear of relapse is not baseless. Although the risk of relapse cannot be precisely quantified, the best available science strongly suggests that Exxon's fear is well grounded. The studies introduced in the hearing below document the risk. Exhibit 48D, the October 1989 edition of Alcohol Alert, published by the National Institute on Alcohol Abuse and Alcoholism reports: "There is evidence that approximately 90 percent of alcoholics are likely to experience at least one relapse over the 4-year period following treatment. Despite some promising leads, no controlled studies have shown any single or combined intervention that prevents relapse in a fairly predictable manner." It goes on to report on a study that "revealed that most relapses were associated with three high risk situations: (1) frustration and anger, (2) social pressure and (3) intrapersonal temptation." I would note that each of these risk factors is largely beyond an employer's ability to control. A second study, Exhibit 48F, recites the results of a Rand Corporation study that found among the alcoholics who received treatment and were followed "that only seven percent of men

abstained throughout the course of the study's four year follow-up."

   The risk of relapse is also well documented in Exhibit 48I, a study by Dr. George Vallant of Dartmouth Medical School. His study, one of the few prospective studies on relapse, found that 95 percent of the alcoholics who received treatment and then were followed resumed drinking at some point during the eight year follow-up. He also concluded that "to a remarkable degree, relapse occurs as independent of conscious free will and motivation." Even the studies relied upon by Complainant support a finding that the risk of relapse is substantial. For alcoholics who remain active in AA, only twenty to thirty percent were able to maintain sobriety for five years. (PX 94). Of those who reported five years of sobriety, nine percent of those who remained active in AA relapsed in the sixth year. As the record demonstrates, these studies tend to err on the low side in calculating relapse because of the reliance of the studies on self-reporting. Only about fifty percent of those interviewed are truthful about their consumption of alcohol. (T. 1148). This fact alone is a good reason why Exxon in this case should not be compelled to rely on a complainant's good faith compliance with his recovery program.

   As serious as the risk of relapse is among the general population of alcoholics, the evidence suggests that it is heightened in persons suffering from bipolar disorder. (DX 48L, M and N). Based on these studies and the hearing record, I find Exxon's fear of relapse to be well substantiated. I do not dispute that decision makers and employers among them could benefit from studies that more precisely identified individual characteristics that were implicated in relapse. Regrettably, the science has not progressed to that stage. The question then becomes: absent such evidence, is an employer when confronted by a substantial public safety concern compelled to rely on the unvalidated by scientific study recommendations of consulting psychologists? In light of the potentially catastrophic consequences of a wrong decision, uncertainty calls for prudence. An employer is given latitude in these circumstances. Huber v. Howard County, 849 F.Supp. 407, 414 (Md. 1994). Therefore, I conclude Exxon could reject the opinions of the consulting witnesses in this case in favor of the available science and condition Complainant's continued employment on his acceptance of a reasonable accommodation.

   My colleagues suggest the law is otherwise and Complainant was denied the individualized assessment that the law requires. I respectfully disagree. As the court in Carr v. Reno, 23 F.3d 525, 530 held: " individualized inquiry' need be no more extensive than

regulations. Chandler, at 1394.

   Similarly, in Ward v. Skinner, 943 F.2d 157 (1st Cir. 1991) the petitioner, a truck driver with a history of epilepsy but who took anticonvulsant drugs and had been seizure-free for seven years challenged a Department of Transportation regulation that prohibited persons with a history of epilepsy from driving trucks in interstate commerce. Although conceding that the risk of someone like the petitioner having a seizure while driving was very low, the circuit court upheld the regulations pending further general studies and investigations, "designed to turn up general features tending to show greater or lesser, degrees of risk." Ward, at 164.

   I find Judge VanArtsdalen's discussion of this issue in Davis v. Meese, 692 F. Supp. 505 (E.D. Pa. 1988) particularly persuasive. At issue in that case were the regulations of the Federal Bureau of Investigation which excluded insulin-dependent diabetics from applying for positions as special agents or investigative specialists. At 517, Judge VanArtsdalen responds to an argument very similar to Complainant's:

Plaintiff argues that determination of who, among the insulin-dependent diabetics, is or is not qualified must be made on an individualized case-by-case basis. The difficulty with this seemingly logical solution is that there exists no reliable method of determining in advance those insulin-dependent diabetics who do not present a substantial risk of having a sudden and unexpected severe hypoglycemic episode while on a duty assignment. The experts agreed that under certain circumstances which might arise while on a duty assignment, any insulin-dependent diabetic would be at risk of having a severe hypoglycemic occurrence. There are tests , such as the insulin infusion test, which are helpful in predicting those more likely to have severe hypoglycemic occurrences and also helpful in predicting those more likely to have frequent and recurring severe hypoglycemic occurrences. However, the evidence fails to establish that any known technology may, on any individualized case-by-case basis, reliably predict the probability or frequency that any individual insulin-dependent diabetic would have a sudden and unexpected severe hypoglycemic occurrence in the future. No known method establishes, within reasonable medical certainty, that the risk as to any individual is nonexistent, minimal or even very small, or any basis to quantify such risk in any and all situations reasonably likely to occur on the job.

Faced with the medical uncertainty, Judge VanArtsdalen concluded that deeper individual inquiry was not required.

for advancement. In fact the opportunities for promotion were probably greater in Gillette than they were at LaBarge. (T. 231-236). Recent case law supports a finding that the proposed accommodation was reasonable.

   In Kerno the plaintiff, a salesman, suffered from obsessive compulsive disorder, post traumatic stress disorder with a secondary reactive depression and chronic depression. He found that these conditions made it difficult for him to work with a new supervisor. He asked his employer to redraw the sales regions so that he could work under a different regional sales manager. His employer refused to consider this alternative and instead offered to transfer him to his choice of open sales positions. Contending that a transfer would complicate his medical treatment and disrupt his family life, he declined the transfer and sued under the Americans with Disabilities Act (ADA), 42 U.S.C. § 12101 et seq., alleging that his employer had failed to reasonably accommodate him. In granting summary judgment for the defendant, the court found that no jury could find a transfer under which the plaintiff retained all aspects of the job he held including job description, title, compensation and benefits to be an unreasonable accommodation. Kerno, at 23.

   In Guice-Mills v. Derwinski, 967 F.2d 794 (2nd Cir. 1992) the plaintiff was a head nurse who suffered from a number of maladies, among them depression, severe anxiety, insomnia and migraine headaches. Upon her physician's recommendation, she requested a change in her duty hours. In response to her request, her employer offered her a position as a staff nurse with the requested hours and at the same pay as a head nurse. Unwilling to accept the demotion, she sued under the Rehabilitation Act claiming that her employer had failed to reasonably accommodate her disability. In rejecting her claim, the circuit court held that "(w)hen an employer offers an employee an alterative position that does not require a significant reduction in pay and benefits, that offer is a "reasonable accommodation" virtually as a matter of law. Guice-Mills, at 797.

   Lastly, in Kuehl the plaintiff, a door greeter at a wholesale store, was diagnosed as suffering from chronic tibula tendinitis which prevented her from standing throughout her shift at work. She requested that she be given a stool to sit on as needed during the course of her shift. Her employer denied the request, but offered her a position as a cashier or in the alternative, agreed to split her shift to enable her to reduce the stress on her legs. She sued under the ADA. The court ruled that "(t)he plaintiff, by rejecting two reasonable alternative accommodations has lost any

status she may have had under the ADA, as a qualified individual with a disability." Kuehl, at 803.

   With these cases serving as guideposts, I find no basis upon which it can be concluded that Exxon's proffered accommodation was anything less than reasonable. Indeed I find it to be generous and fairly considered. In the present case, the majority's opinion presupposes that testing, supervisor evaluation and AA attendance would be a more effective accommodation. Regardless of the merits of that judgment, for the reasons given above, I think legally it is of no consequence. Nonetheless I think the majority's view is suspect and merits some comment.

   Some of the shortcomings of the monitoring and supervision program that the majority touts are set out in Altman v New York City Health and Hospitals Corporation, 903 F. Supp. at 511. In Altman, the plaintiff, a recovering alcoholic, resisted his demotion from Chief of Internal Medicine to staff physician and contended that the defendant hospital should in lieu of demoting him have subjected him to the "professional monitoring" program proposed by the Committee on Physicians' Health. The court identified two problems common to such monitoring programs:

First, it must be noted that plaintiff's history suggests that he was able to consume substantial quantities of alcohol without his impaired condition being noticed by his professional colleagues. Thus, even reasonable monitoring could not insure that plaintiff would be performing his duties without his judgment being impaired by alcohol.

Then later:

In addition, the people perhaps best equipped to monitor plaintiff's performance and behavior in a less formal way -- the departmental staff members with whom he interacted on a day-to-day basis while performing his clinical and administrative duties -- would have been placed in the awkward position of having to judge his judgment, and if necessary, having to disobey the instructions of their supervisor. They also would have had to face the unsettling truth that, if they reported another relapse to plaintiff's superiors, they would effectively have brought about the end of plaintiff's professional career.

    These are among the reasons that Exxon rejected a monitoring and supervision program in favor of its policy. (See testimony of

Dr. Hayden). The majority is nevertheless concerned that Exxon's policy may actually discourage self-reporting and drive alcoholics underground. This is a legitimate fear but not necessarily the inevitable outcome of Exxon's policy. Under Exxon's policy, an individual suffering from alcoholism has three choices. First, he could disclose his alcoholism and be transferred to an equivalent non-designated position while maintaining his pay and benefits and his entitlement to medical treatment. Second, he could fail to disclose and maintain sobriety. Third, he could fail to disclose and relapse and run the risk of the relapse being detected through testing, medical records or reports of his colleagues. The third choice poses the greatest risk to the individual because if his relapse is detected, he would lose his job and benefits. If he chooses courses one or two, Exxon and the public safety would benefit. I find no reason to believe that outcome three is more likely than the others. Even if I did, I believe employers retain the discretion to craft safety policies provided that those policies are not a mere pretext for discriminating against individuals with disabilities.

   Exxon continues to examine its policies in light of the available evidence. Its policies are the matter of much debate in the corporation. (See testimony of James Rouse). As it examines the experience of other companies and organizations, it intends to regularly revisit its policy to determine whether or not it best serves the interests of the company and public. Rather than compelling businesses to adopt the one model endorsed by the majority (i.e FAA's), employers should be encouraged to constantly rethink and when appropriate modify their policies. There is no right answer to combating the dangers posed by substance abuse in the workplace. There are only partial, evolving solutions. Exxon's response may not be the best, but as it was applied to Complainant, it did not violate the Rehabilitation Act.

   This brings me to my last point. My dissent is addressed to the facts of this case. I do not hazard a guess as to whether Exxon's policy as it has been applied in other cases would violate the Rehabilitation Act. Exxon may or may not have over designated the number of positions as safety sensitive. The accommodations offered other employees may or may not be reasonable. For that matter, by the time this decision is issued Exxon may have changed its policy. Claims under the Rehabilitation Act are fact specific. With only the illumination of the facts of a particular complaint, I would not arrogate the responsibility to supervise general compliance with the law. Employers should be given the opportunity to adjust their safety policies in light of specific decisions, rather than be compelled to discard their policies in response to a general indict. Wrongs against specific individuals can be corrected. Public safety should not be compromised pending review and implementation of an alternative policy. Therefore I believe the Board's order is unnecessarily and overly broad insofar as it addresses the treatment of employees whose cases are not before this Board.

       KARL J. SANDSTROM
       Member