Mary L. Wilson and Sidney C. Jackson separately seek review of two decisions of the Merit Systems Protection Board (MSPB or Board) sustaining adverse actions by their respective employers, the Social Security Administration (SSA) of the Department of Health and Human Services, 25 M.S.P.R. 681, and the Environmental Protection Agency (EPA).1 Both petitioners were demoted after they failed to achieve the minimally satisfactory level of proficiency in a critical element of their positions as specified by the relevant performance standards. Each contends that these performance standards failed to satisfy the statutory requirement that such standards be based on objective criteria, concluding that the standards are invalid and therefore their demotions cannot be sustained.2

Because these cases raise a common issue concerning the statutory requirements for valid and enforceable performance standards governing federal employees, we have on our own initiative consolidated the appeals for decision. In No. 85-1857 we hold that the performance standard relating to a critical element of Wilson's position was not sufficiently objective as required by statute to allow for a fair and accurate characterization of her performance. On the other hand, in No. 85-1864, we determine that the EPA's performance standard involving a critical element of Jackson's position satisfied the statutory criterion of objectivity. The Board's decision in No. 85-1857 is accordingly reversed; the decision in No. 85-1864 is affirmed.

Section 203 of the Civil Service Reform Act of 1978, Pub.L. No. 95-454, 92 Stat. 1157, amended Chapter 43, Title 5 of the United States Code to provide for a new performance appraisal system for federal employees. Of particular importance here are 5 U.S.C. Secs. 4302 and 4303, which direct executive agencies to develop performance appraisal systems consistent with stated criteria, and authorize Government agencies to penalize or reward employees based on the outcome of a performance evaluation under the proper appraisal system. A central feature of the new appraisal system entails agency-promulgated performance standards, which set forth the degree of proficiency necessary to achieve a given rating (e.g., minimally satisfactory, satisfactory or outstanding). These standards must, "to the maximum extent feasible, permit the accurate evaluation of job performance on the basis of objective criteria ... related to the job in question for each employee or position under the system." 5 U.S.C. Sec. 4302(b)(1).

The statute also provides that an agency may reduce in grade or remove an employee for "unacceptable performance." Id., Sec. 4303(a). "Unacceptable performance" in this context is a term of art; it is "performance ... which fails to meet established performance standards in one or more critical elements of such employee's position." 5 U.S.C. Sec. 4301(3). A "critical element" is "a component of a job consisting of one or more duties and responsibilities ... which is of such importance that 'Unacceptable' performance on the element would result in 'Unacceptable' performance of assigned work." 5 C.F.R. Sec. 430.203 (1984). In sum, if an employee fails to meet the level of work set forth in a performance standard that specifies in objective terms what constitutes satisfactory performance in a critical element of a position, then the employee is subject to demotion or removal.

The question under this statutory scheme which petitioners present is novel to this court: what did Congress mean in 5 U.S.C. Sec. 4302(b)(1) when it required that performance standards contain "objective" evaluation criteria permitting "accurate" job evaluation "to the maximum extent feasible"? As a subissue, we also consider the proper measure by which the MSPB and this court are to judge the sufficiency of contested standards.

The legislative history of the Reform Act generally and of Chapter 43 in particular reveals a conscious desire on the part of Congress to reject the previous system by which the Government evaluated its employees, and to create a new scheme resting on a different foundation. The old system, most recently promulgated in the Performance Rating Act of 1950, 64 Stat. 1098, provided for the summary evaluation of employees' performance under three ratings: satisfactory, unsatisfactory or outstanding. This system ultimately proved ineffective as a management device because (1) it was too subjective to differentiate adequately between competent and incompetent employees, and correspondingly (2) it injected into the appraisal system an element of subjectivity over which any employee could contest the Government's action in lengthy administrative and judicial proceedings. Congress specifically recognized these problems when it drafted the Reform Act:

Although it is difficult to pinpoint the reasons why these figures [i.e., the number of adverse actions taken against federal employees prior to the Reform Act] are low, among the reasons are:

Performance evaluation procedures do not work well enough to distinguish employees whose performance is below an acceptable level to make those charges stick.

Fear of employee appeals, grievances and the potential problems they create for the employee, the manager and the work unit involved.

S.Rep. No. 969, 95th Cong., 2d Sess. 9, reprinted in 1978 U.S.Code Cong. & Ad.News 2723, 2731. The House report, for its part, said that the new statute would require performance standards and critical elements to "have been adequately defined " by an agency. H.Rep. No. 1403, 95th Cong., 2d Sess. 21 (1978) (emphasis added).

Thus, Congress intended in the Reform Act to initiate and continue a system that would assure that competent employees would be retained and unsatisfactory employees removed from their positions, while keeping to a minimum the need for protracted administrative and judicial intervention resulting from inadequate standards or appraisals. Id. at 4, 1978 U.S.Code Cong. & Ad.News at 2726; see generally Lisiecki v. Merit Systems Protection Board, 769 F.2d 1558 (Fed.Cir.1985) (discussing the legislative history of the Reform Act).

In Wells v. Harris, 1 MSPB 199, 1 M.S.P.R. 208 (1979), the Board reviewed thoroughly the history and structure of Chapter 43. The Board stated that several of the opinions expressed during Congress' consideration of the Reform Act were highly reflective of the legislature's concerns when it called for objective performance standards. For example, the Board considered relevant the testimony of the then-Chairman of the Civil Service Commission before committees of both houses, in which he stated that the old adjectival performance ratings were "useless as a basis for [personnel actions] because they [did] not provide enough information to make any of these decisions.... The increased emphasis on meaningful appraisals ... will also provide [supervisors] with a more effective and equitable means of managing their employees." 1 MSPB at 220-21, 1 M.S.P.R. at 232-34. Congressman Clay of the House Committee on the Post Office and Civil Service noted that "the use of objective criteria in performance appraisals, consistently applied, will benefit Federal employees by providing them with protection against arbitrariness and discrimination." Id. at 233, 1 M.S.P.R. at 246. Senator Stevens of the Governmental Affairs Committee pointed out that "[t]he employee must know the specific criteria which will be used in his evaluation." Id.

In subsequent cases, the Board has used this background to stress that performance standards must "accord[ ] with the statute and promote[ ] the statutory purpose." Callaway v. Department of the Army, 23 M.S.P.R. 592 (1984). Under the Board's formulation, a performance standard must be "reasonable and provide[ ] for accurate measurement of appellant's performance." Huffman v. Department of Health and Human Services, 20 M.S.P.R. 345 (1984). The Board recognizes, as it should, that the "objectivity" requirement in the statute promotes the purposes of the Reform Act by informing employees and their supervisors, as precisely as feasible, of the level of proficiency necessary to achieve a given rating. This objective appraisal system serves to deter the arbitrary personnel actions and protracted appeals which Congress found to be so frequent under the earlier, subjective standard system. As the Board has noted:

When Congress adopted this requirement to remedy deficiencies in the previous system for rating employee performance, it recognized that it would benefit employees by protecting them from arbitrary labeling of their work as unacceptable. In requiring the use of objective criteria in performance standards and communication of the standards to employees, Congress also intended to ensure that employees were made aware in advance of what was expected of their performance.

Siegelman v. Department of Housing and Urban Development, 13 MSPB 27, 29, 14 M.S.P.R. ---- (1983); accord, Baker v. Defense Logistics Agency, 25 M.S.P.R. 614 (1985).

We wholly agree with the Board's formulation, as set forth in Siegelman and other Board decisions, that under the statute's objectivity requirement performance standards must be reasonable, sufficient in the circumstances to permit accurate measurements of the employee's performance, and adequate to inform the employees of what is necessary to achieve a satisfactory or acceptable rating. If the performance standards satisfy this test, then they further the congressional purpose. Employees will be spared arbitrary ratings and adverse personnel actions grounded purely on subjective impressions. Supervisors will feel freer to propose personnel actions against unsatisfactory employees--based on such objective standards--confident that they have made a verifiable decision, and the employee will not be retained because the evaluation process itself was arbitrary or because the agency is reluctant to defend inevitable and protracted appeals.

Petitioners do not reject the Board's general interpretation of the objectivity requirement. However, they urge further that, to reach these goals, the Government must provide its employees with precise quantitative or numerical standards. They contend that only such mathematical or percentage standards can completely remove subjective judgment from the employee evaluation process, as Congress intended. We cannot agree that this refinement is mandatory.

The statutory terms "accurate evaluation," "objective criteria," and "to the maximum extent feasible"--taken together with Congress' stated objective (in the legislative history) to obtain performance standards that were "adequately defined"--all suggest that Congress desired standards sufficiently specific that compliance vel non with those standards could be verified in the particular instance by others than the official appraisers of the employee's performance. A parallel way of putting the same Congressional objectives is that a standard should be sufficiently precise and specific as to invoke a general consensus as to its meaning and content. But the legislative language does not suggest any necessary requirement for numerical measurement, and it is not at all difficult to think of many positions in which such strictly quantitative criteria would be unrevealing, bizarre, or counter-productive.

Furthermore, Congress clearly expressed its intention to allow the agencies "great flexibility to choose or develop their own systems. Agencies should determine what type of performance appraisal methods best suited their needs." S.Rep. No. 969, supra, at 42, 1978 U.S.Code Cong. & Ad.News at 2764. This strong emphasis on "great flexibility" for the agencies runs directly counter to the rigidity of an obligation for numerical or quantitative standards. Of course, the agency must always live up to the statutory demand that the standard allow for reasonably accurate measurement of performance and protect employees against arbitrary treatment. However, we reject outright, as has the Board, any absolute rule regarding the formulation of performance standards (including that they pose a stringent numerical requirement).

We now apply these tests for valid standards to the Wilson and Jackson cases before us.

Mary Wilson was a social insurance representative assigned to the SSA's Arlington, Va., office. She had supervisory authority over fourteen other agency employees in that office. A critical element of her position required that she "[d]irect[ ] work activity--issue[ ] orders and instructions on what, how and by whom action should be taken." A performance standard for this critical element defines minimally acceptable performance for this activity as:

Coordinates controls, and directs activities of subordinate staff to insure adequate service to the public by appropriate management principles. Assignments and instructions to staff are hastily made and sometimes misunderstood. Direction of work activities is occasionally effective in achieving objectives.

In 1983, Wilson's supervisors concluded that she had not met this minimally acceptable level of performance and effected her demotion and reassignment to another office.

Wilson appealed to the MSPB challenging both the agency's specific allegations and the validity of the minimally acceptable performance standard. The Board's presiding official found that substantial evidence supported some (though not all) of the agency's charges, including failure to cure work backlogs and failure to review certain social security claims. With regard to the validity of the standard, the presiding official said that much of the standard (e.g., the part which required that Wilson's instruction be "occasionally effective") amounts to "almost no standard at all." She did point out, however, that the standard required Wilson "to insure adequate public service" by the employees in her charge. The presiding official concluded that this aspect of the standard was valid, and sustained the agency's action. The full Board declined review.

We rule, however, that the standard the Board sustained cannot be upheld as within the requirements of the Civil Service Reform Act. Rather, it is almost a parody of a proper standard. First, the standard permitted Wilson to issue instructions which are "sometimes misunderstood." Nothing indicates how Wilson or her supervisors were to interpret "sometimes" in this context. In the normal course of human interchange, instructions are sometimes misunderstood. This is to be expected for even the clearest and most efficient manager. When the phrase appears in a standard defining minimally acceptable performance, we assume it must mean even less than this normal, unavoidable, inevitable, occasional miscommunication. The standard as stated is, however, so vague and inexact that it is impossible to apply in a verifiable fashion or to discover the level of proficiency which the SSA actually intended by the phrase. It is exactly the type of standard that left Wilson open to arbitrary appraisal in contravention of the Reform Act.

The same criticism applies to that portion of the standard that requires that Wilson's direction of her subordinates' work be "occasionally effective." The word "occasionally" indicates less than a majority of times, but we cannot be certain or even reasonably sure how much less in this or any instance. Moreover, the SSA placed itself in the awkward position of having to prove a negative; in order to show that petitioner was not occasionally effective, it must, because of the imprecision of "occasionally," prove that she was never effective. The presiding official was quite correct when she noted that this portion of the standard was meaningless.3

The Board sustained this performance standard because it also required Wilson to "insure adequate service to the public." We cannot agree that this phrase standing alone satisfies the statutory requirement of objectivity. Every government employee must, of course, serve the public as the duties of his or her position dictate. We assume that failure to serve the public in a meaningful way itself constitutes unsatisfactory performance. This type of performance standard, however, is fatally similar to the pre-Reform Act summary adjectival rating of "satisfactory." It does not specify what duties the employee must perform in order to receive a satisfactory rating. As we have pointed out in Part II, supra, Congress intended in the Reform Act to do away with just such broad, general, overly-subjective criteria. In addition, if that portion of the standard has any acceptable meaning, it is in direct conflict with the linked requirements that the employee's instructions and assignments can be "hastily made," or "sometimes misunderstood," and only "occasionally effective." The latter kind of performance can hardly "insure adequate service to the public." The result is that neither the separate parts of the standard, nor the standard as a whole, provide adequate guidance to Wilson and her supervisors--nor could the existence of the standard's components be verified or found to exist apart from the appraiser's own subjective evaluation.4

We conclude in Wilson's appeal that the SSA abused its discretion, by violating the Reform Act, when it promulgated the performance standard setting forth the minimally acceptable level of proficiency expected of Wilson in the critical element concerning the management of her subordinates. The converse is that the performance standard on the basis of which the SSA demoted Wilson did not properly inform her of what was required to achieve an acceptable performance rating. That standard was not valid because it failed to comply with the statutory mandate. Accordingly, the agency's decision to demote Wilson, based on that invalid standard, cannot be sustained.

Wilson's agency has suggested that if we find Wilson's performance standard to be invalid, as we do, that we remand the case to the MSPB for consideration by the Board under Chapter 75 of Title 5 (relating to adverse personnel actions taken "for the efficiency of the service"). We are directed to this court's recent decision that an agency has the choice of proceeding under either Chapter 43 or Chapter 75 when an employee fails to perform satisfactorily. Lovshin v. Department of the Navy, 767 F.2d 826 (Fed.Cir.1985) (en banc ). The Lovshin situation, in which the agency made an a priori choice between two statutorily authorized procedures, is much different from the SSA's present position, in which it seeks to change the nature of the proceeding completely after the fact. As the Board has noted in response to a similar request:

[T]he establishment of valid performance standards is an important substantive right of an employee under Chapter 43. There is no comparable provision requiring the establishment of valid performance standards for long-term measurement of performance under Chapter 75. [Citation omitted.] Moreover, if the agency had established a valid performance standard under Chapter 43, appellant might have met the standard and no action would have been brought. To allow the agency to proceed with a Chapter 75 action is to subject appellant to an adverse action where, but for the agency's failings, there might have been no action.

Callaway, supra. In Callaway, the Board properly refused a request, as here, to change the nature of the action at the Board level. Because of the Lovshin decision, however, we leave open the question whether the agency may now proceed with a Chapter 75 action against Wilson, beginning at the agency level and following the procedures required by that Chapter, based on the same or a similar set of operative facts--if the facts are thought to warrant such a proceeding. But we cannot accede to the Government's request to remand Wilson's case to the Board for reconsideration now under Chapter 75 on the very same record.

Sidney Jackson was an environmental scientist with the EPA responsible for reporting on seminars and meetings of permittees and licensees of both the EPA and similar state and local agencies. A critical element of this position required Jackson to:

Concisely define information needed for developing comprehensive reports. Establish format for obtaining and presenting information. Identify all available information resources and coordinate information development with other program offices. Verify accuracy of information presented. As example, write, edit and publish technical newsletter on relevant Agency and state permit assistance activities.

The corresponding performance standard (applied in this case) required that:

Draft reports are completed in a timely manner, address all pertinent issues, and require minimum revisions. Information developed and reported is coordinated with other relevant program offices. Recognize, report and recommend solutions to potential problems.

According to EPA's charges, Jackson's reports on licensee and permittee seminars and workshops were often late and required extensive revisions. Initially, the agency proposed Jackson's removal, but decided later to demote him to a lower grade.

On appeal, the Board's presiding official found that substantial evidence supported the agency's allegations.5 Jackson contended that, notwithstanding his problems, the performance standard was invalid because it was not sufficiently objective. The presiding official concluded that the EPA had not abused its discretion in devising this standard. We agree.

The performance standard relating to Jackson's preparation of reports requires that they be completed in a timely manner, that they address all relevant issues, and that they require minimum revisions. These are sufficiently objective and precise in the sense that most people will understand what they mean and what they require--and Jackson could adequately determine (in the words of the presiding official) "whether the standard [was] attained and exceeded." Jackson correctly notes that this standard allows for some subjective judgment on the part of his evaluators. This is particularly true of the portion prohibiting more than minimum revisions. But we cannot conclude, as does Jackson, that this is per se a fatal flaw in the standard. Jackson's position required that he write reports as a professional; to argue that he should be judged in a purely mechanical way, as by the number of pages he wrote (or the bare number of errors committed, as if all errors were equal), is absurd. As the Board has recognized, some tasks may be rated only with a certain modicum of subjective judgment. See Siegelman, supra, 13 MSPB at 29, 14 M.S.P.R. at ---- ("[W]ork at appellant's level is not susceptible to the kind of mechanical judgment-free rating which she argues is required"). Congress recognized that these situations would exist when it provided in 5 U.S.C. Sec. 4302(b)(1) that the objectivity requirement must be met only "to the maximum extent feasible." The challenged standard as a whole strikes us as within the statute, adequately gauged and informative to Jackson, and readily attainable.6

Moreover, EPA did not base its action solely on the performance standard. The standard provided the initial impetus for placing Jackson on an individual development plan. At that point, EPA instructed him on precisely how he could achieve a satisfactory rating. The standard was thus fleshed out and implemented in detail. In that respect, the current case is similar to Shuman v. Department of the Treasury, 23 M.S.P.R. 620 (1984). Shuman was a Revenue Officer with the Internal Revenue Service (IRS). As such, her position required that she exercise judgment in the performance of her duties. As a result, the performance standard by which she was evaluated required an exercise of some subjective judgment on the part of her supervisor. The evidence in that case disclosed that the IRS specifically informed Shuman before it took action (following an unsatisfactory appraisal) of those areas in which it considered her weak, and appropriate methods by which she could improve. When Shuman failed to improve in these areas, after receiving a specific and detailed notice of the improvement the IRS expected, the Board held that the IRS could properly rate her unsatisfactory and take appropriate action.

The EPA's actions with regard to Jackson were comparable. The performance standard reasonably informed Jackson of what, how and when he was to write. When he failed to meet EPA's expectations, he was informed of the specific areas in which he needed improvement and of the specific tasks by which he could satisfy EPA's concerns. Jackson then knew very precisely the level of proficiency he would have to achieve to retain his position. When he failed to complete those tasks in a satisfactory manner, EPA could properly take action.

For these reasons, the Board's decision in No. 85-1857 is reversed; the decision in No. 85-1864 is affirmed.

No. 85-1857 REVERSED,

No. 85-1864 AFFIRMED.