In this case we must ask when an employer's failure to accommodate an employee under the Rehabilitation Act of 1973, 29 U.S.C. § 701 et seq., amounts to a constructive discharge of that employee. Dr. Sharon Johnson left her employment with the National Institutes of Health; she now claims that she was constructively discharged from her job. The district court agreed, holding that NIH had constructively discharged Dr. Johnson because her supervisors knew of her handicapped condition but failed to give her the accommodation that she requested and to which she was entitled under the Rehabilitation Act. We believe that this legal standard would convert every violation of the Rehabilitation Act into a potential claim for constructive discharge. Applying the proper standard for constructive discharge, we find no evidence that NIH intended to force Johnson to resign, and accordingly we reverse the judgment of the district court.

The record in this case demonstrates an employment relationship that was beset on all sides by strains and difficulties. The plaintiff lived some distance from her place of employment, and she suffered from a variety of genuine medical ailments, which did not yield readily to treatment. These problems were exacerbated by a demanding job that by its nature engendered a fair degree of stress. Reviewing the history of this employment relationship, it seems plain (and NIH now concedes) that while NIH accommodated plaintiff's handicaps in some respects, it did not do all that was required under the Rehabilitation Act. When reviewing these facts for a claim of constructive discharge, however, the question becomes whether NIH sought to force the plaintiff to resign.

Dr. Sharon Johnson worked from 1984 to 1986 as the Executive Secretary of the Pathobiochemistry Study Section in the Division of Research Grants for the National Institutes of Health, a branch of the Department of Health and Human Services. The Study Section is a group of scientists who review biomedical research grant applications to determine which applicants should receive NIH grants. Dr. Johnson's duties included organizing and attending meetings for the review of grant proposals, preparing written reports of those meetings, and making site visits to grant applicants. She was also responsible for processing appeals of grant denials and for selecting members of the Study Section. Dr. Johnson's position as chair of the Study Section meetings required that she attend them in their entirety. 5 U.S.C. Appendix 2, § 10(e). The meetings sometimes lasted all day and often went beyond normal business hours.

Unfortunately, Dr. Johnson had a medical condition which made the performance of a demanding job even more difficult. She suffered from idiopathic CNS hypersomnolence, a form of narcolepsy or excessive sleepiness that required her to sleep nine or more hours per night and to take short naps during the day when she became tired. The condition often caused Johnson to fall asleep during the meetings she was conducting. Dr. Johnson also suffered from cardiac arrhythmia, a condition that causes a rapid, erratic heartbeat and that precluded her from taking medication to control her narcolepsy. In addition, Dr. Johnson had had surgery for breast cancer and needed periodic checkups to ensure that the cancer did not recur.

The situation was further complicated by commuting problems. Dr. Johnson and her husband moved to Annapolis, Maryland in 1982. She drove daily from Annapolis to NIH's Bethesda offices, a trip that takes approximately one hour each way. Her narcolepsy, however, frequently required that she pull over for fifteen-minute naps, making it difficult for her to arrive at work at a fixed time.

By memorandum of August 22, 1985, Dr. Johnson requested flexible starting and ending times for her work. At the same time, she requested authorization to change her scheduled work hours more frequently than the permitted July 1 and December 1 dates, so that she could take advantage of "seasonal traffic patterns." In a reply memorandum, Dr. Asher Hyatt, her immediate supervisor, informed Johnson that her request required medical documentation. This memo also stated: "[N]ote that the medical supporting information may be used as documentation in a disability retirement action. Disability retirement may be offered if accommodation cannot be reached." Dr. Johnson sent Hyatt a reply memo with a letter attached from her physician. The letter recommended:

[S]he will need to have flexible working hour schedule with due regard given to possible late arrival and late departure.... She is never expected to obtain full or partial recovery and will always be hampered by this condition. If the patient were given a regular work schedule wherein she could participate in a car pool arrangement where she would not be responsible for driving to and from work the above recommendations would not be necessary. However, I understand that she needs to work late at times to attend meetings associated with her current job situation.

After consulting with his supervisor and a physician from Occupational Medical Services at NIH, Hyatt sent Johnson a note saying "The only further accommodation I can make is to the hours of a car-pool. I strongly suggest you try to find an appropriate car-pool. There is a locator in Bldg. 31." No adjustment was made to Johnson's work hours to make them more regular, but around this time all employees were sent a memo informing them of the agency's "Flextime Policy," which permitted up to fifteen minutes variance in arrival and departure times. In addition, Dr. Johnson was permitted to change her work hours four times a year, rather than the standard two times permitted for other employees.

Dr. Johnson joined a carpool soon after receiving the memo from Hyatt. Meeting the carpool required a twenty-five to thirty minute drive, which, on those days when it was her turn to drive the carpool, made her drive longer than before. In addition, Johnson's duties as chair of the Study Section did not permit her to drive every day with the carpool. She did not, however, communicate her dissatisfaction to Dr. Hyatt.

In February 1986, Dr. Johnson requested that she be allowed to stay in a hotel in Bethesda at government expense for two nights during a three-day Study Section meeting. Dr. Hyatt refused this request, which was contrary to NIH policy.

On April 7, 1986, Dr. Johnson requested five weeks of Leave Without Pay ("LWOP") to begin April 9. Accompanying this request was a letter from Johnson's psychologist advising that Johnson needed at least a month off because she was suffering from "extreme physical and psychological exhaustion," which the psychologist attributed to Johnson's long commute and stress from her NIH workload. On April 8, Dr. Friedman, Johnson's second-level supervisor, approved that request, which would become effective when Johnson completed the summaries of the last meetings and reported on the status of grant applications scheduled to be considered at the next round of meetings.

On April 11, Dr. Johnson requested fifty-two hours of advance sick leave for routine medical examinations. In a memo, Dr. Hyatt approved twenty-nine hours of leave that Johnson had already taken, but informed her that in the future she would have to comply with NIH rules and first request in writing any advance sick leave. On April 16, Johnson spent two hours at a local hospital for a medical exam. When she submitted a slip for sick leave later that day, Hyatt denied it on the grounds that Johnson had not obtained prior approval. She was, however, allowed to take annual leave for this time. The next day, Hyatt sent Johnson a memo advising her that she would receive no more sick leave until she had again accrued positive sick-leave time, and that no leave of any kind would be granted unless a request was made in writing three days in advance. Johnson regarded this requirement as a unique imposition upon her and as a refusal to accommodate her handicap; her supervisors viewed it as a reasonable response to what had become a substantial sick-leave deficit and as an effort to "control the situation, so that we knew where we were and where Dr. Johnson was with her leave requests."

On April 18, Dr. Johnson submitted a handwritten memo on the status of her work to Dr. Friedman. Friedman requested a typed copy, which Johnson produced. Hyatt reviewed the memo and informed Friedman that Johnson's work was "essentially complete." On April 21, Johnson gave James Pike, the executive director of her section, a letter from her psychiatrist, Dr. Thomas Goldman, stating that she was suffering from stress and needed a leave of absence. Johnson also had her attorney call Pike and request the LWOP. On that same day, Johnson filed an EEO complaint, alleging that NIH had not accommodated her handicap and had discriminated against her on the basis of her sex and handicap. Later that day, Pike granted Johnson her LWOP starting April 22.

On May 6, Dr. Goldman sent Pike a letter discussing Johnson's medical and mental condition. Goldman thought that Johnson did "not appear to have the resources to adapt effectively to the stress level in her work environment." In the letter, Goldman recommended that:

From her description of her job situation and the information I have about it, it is doubtful that a truly reasonable accomodation [sic] can be provided her which would substantially reduce the level of psychic stress to the point where she could comfortably cope without constant aggravation of her symptoms. Therefore, barring some major change in her work environment, the best alternative from the medical standpoint would be retirement.

Later in the month, Hyatt extended Johnson's LWOP at her request.

On June 4, Dr. Johnson applied for disability retirement, a possibility that had been discussed among her supervisors at NIH. Dr. Johnson testified at trial that she wanted to continue work, but that the difficulty she was having in getting accommodation for her handicap and the stress she was experiencing in her work made retirement her only option. On June 5, Hyatt drafted an evaluation, backdated to April 1, 1986, stating that Johnson's performance had been unsatisfactory. This evaluation did not reflect Hyatt's views, but he wrote it believing that only her attorney would use the evaluation to enable Johnson to receive disability retirement. The government concedes, as it must, that Hyatt's action was "wrong and a serious mistake." Hyatt cut out Johnson's signature from a prior evaluation and attached it to the false evaluation. Hyatt claims that he prepared the false evaluation at the request of Johnson's attorney, but the attorney denied it at trial.

On June 6, Pike wrote to Johnson advising her that she would have to be psychiatrically examined before returning to work. Johnson testified that she believed that an examination was impermissible for this purpose, but, fearing discharge, she submitted to the request. During this exam, Johnson was shown the evaluation fabricated by Hyatt. The government states that the psychiatric examination was for the purpose of assisting Johnson's application for disability retirement. Johnson testified at trial that the psychiatrist told her this during the examination.

On August 5, Johnson met with an NIH official to discuss possible reassignment. At this point, Johnson was offered a position at GS-13, a grade lower than her GS-14 position, which she declined. The parties dispute whether she was offered the same salary for the GS-13 position. On August 20, 1986, Dr. Johnson resigned. Subsequently, she was awarded disability retirement benefits.

Johnson eventually filed this lawsuit, alleging violations of Title VII and the Rehabilitation Act. After a bench trial, the district court rejected the Title VII claims but found that NIH had not reasonably accommodated Johnson's handicap and had constructively discharged her by failing to act in the face of known intolerable conditions. The district court awarded Johnson $57,638.39 in backpay and ordered reinstatement. Separately, the district court awarded Johnson $106,526.63 in attorneys' fees. The government now appeals from the district court's determination that it constructively discharged Johnson.*II.

The government claims that the district court applied an erroneous legal standard in finding that NIH constructively discharged Dr. Johnson. In this circuit, the standard for constructive discharge requires a plaintiff to show both intolerable working conditions and a deliberate effort by the employer to force the employee to quit. Bristow v. Daily Press, Inc., 770 F.2d 1251, 1255 (4th Cir.1985). The government challenges only the district court's finding of deliberateness. Deliberateness can be demonstrated by actual evidence of intent by the employer to drive the employee from the job, or circumstantial evidence of such intent, including a series of actions that single out a plaintiff for differential treatment. EEOC v. Clay Printing Co., 955 F.2d 936, 944-46 (4th Cir.1992). Accord Johnson v. Bunny Bread Co., 646 F.2d 1250, 1256 (8th Cir.1981) ("fact that employees were treated identically rebuts any inference" of constructive discharge).

This standard for deliberateness, however, has evolved in cases of age, sex and race discrimination. In these cases, there is ordinarily a relevant pool of workers whose treatment can be compared to the treatment accorded plaintiff. In contrast, the Rehabilitation Act demands more of the federal government than simple equality of treatment--the government must affirmatively take steps to accommodate employees with handicaps, unless accommodation would impose undue hardship on the government. 29 U.S.C. § 791; 29 C.F.R. § 1613.704. Treating disabled workers the same as workers without a disability falls short of satisfying the requirements of the Rehabilitation Act. See Southeastern Community College v. Davis, 442 U.S. 397, 410-11, 99 S. Ct. 2361, 2369, 60 L. Ed. 2d 980 (1979). This duty of affirmative accommodation complicates the application of a standard requiring evidence of differential treatment to establish deliberate intent to discharge. The number of individuals with handicaps in any agency is likely to be small, and those individuals will have different handicaps, each requiring different accommodations on the part of the government. The traditional standard of constructive discharge therefore does not neatly translate to the context of the Rehabilitation Act.

In this case, the district court held that NIH had constructively discharged Dr. Johnson because her supervisors were on notice of her need for changes in her working conditions, but failed to affirmatively accommodate her. We think that this standard sweeps too broadly; rather than cabining the doctrine of constructive discharge, the standard adopted by the district court threatens to convert every failure to accommodate under the Rehabilitation Act into a potential claim for constructive discharge.

The consequences of regarding every failure to accommodate an employee as a constructive discharge would be significant. Under this standard, if a plaintiff can prove that the government has violated the Act, she can automatically quit her position without first resorting to the administrative and judicial remedies provided by Congress to mediate these disputes while the employment relationship can still be salvaged. See 29 U.S.C. § 794a(a)(1) (incorporating the provisions for conciliation agreements and equitable remedies available to federal employees under 42 U.S.C. § 2000e-16 and § 2000e-5(f) through (k)). In the interval between a constructive discharge and reinstatement of the employee, both sides lose. The plaintiff may lose the advantages of continuity and understanding gained by being on the job site, and the government irretrievably loses productivity. The plaintiff may recover the backpay that she would have received for working, but the government cannot recover the lost work that it would have received for that compensation. Moreover, once the employment relationship has terminated, the parties may harden their positions and become unable to resolve their dispute. It is far better for all concerned to resolve the dispute while the employment relationship is ongoing.

For these reasons, we believe the district court erred as a matter of law in grounding its finding of constructive discharge upon the premise that NIH failed to afford Dr. Johnson reasonable accommodation. The evidence here is insufficient to show a deliberate intent to discharge an employee when there has been an attempt to accommodate that same employee, even though the accommodation falls short of satisfying the requirements of the Rehabilitation Act. The plaintiff must present some evidence that the employer intentionally sought to drive her from her position.

In this case, Dr. Johnson failed to meet this burden. NIH did not ignore Dr. Johnson's requests--NIH responded, but frequently in ways that she found unsatisfactory. Rather than the thirty-minutes flexibility in her starting time that Johnson requested, NIH gave her fifteen. She was authorized to join a carpool, but without the regularized hours that she wanted. NIH granted her initial request for Leave Without Pay, albeit under the condition that she complete certain necessary work before departure. NIH also offered Johnson a different position, but Johnson refused it because it was at a lower grade. In other instances, NIH complied fully with Dr. Johnson's requests, permitting her to change duty hours four times a year instead of twice like other employees, and later granting her extended Leave Without Pay. NIH could have done more, and should have done more to conform to the requirements of the Rehabilitation Act, but it did not simply turn its back on Johnson's needs. Finally, once Johnson requested disability retirement, a solution urged by her own psychiatrist, personnel at NIH tried to help her obtain approval for her application. The evaluation falsified by Hyatt, while plainly inexcusable, fails to demonstrate an intent by NIH to force Johnson from her position because the evaluation was prepared only after Johnson herself requested disability retirement. The evidence put forth by Johnson may well demonstrate a lack of flexibility or magnanimity on the part of her supervisors, but what it does not demonstrate is a deliberate intent to force Johnson from her job. NIH tried, but failed, to accommodate Johnson's handicaps.

We hasten to add that ours is not a sweeping holding. We recognize that a complete failure to accommodate, in the face of repeated requests, might suffice as evidence to show the deliberateness necessary for constructive discharge. And, of course, deliberateness can always be shown through direct evidence that an employer intended to force an employee from his or her job. However, in this case of partial or imperfect accommodation in the context of an employment relationship that was difficult for everyone involved, we do not think that the elements of a constructive discharge have been met.

For the above reasons, the judgment of the district court is

REVERSED.