How Well Is the Government Dealing With Poor Performers? Y General Government Division B-236639 October 2, 1990 The Honorable Ted Stevens Ranking Minority Member Subcommittee on Federal Services, Post Office, and Civil Service Committee on Governmental Affairs United States Senate Dear Senator Stevens: In response to your request, this report discusses how the federal government identifies and deals with employees whose performance has been judged as less than fully successful (i.e., poor performers). The report contains our views on how the management of the process for handling such individuals could be improved. It also presents options for congressional consideration aimed at easing the difficulties supervisors face when dealing with certain employees who have been poor performers for lengthy periods. Copies of this report are being sent to other appropriate congressional committees, the Director of the Office of Personnel Management, agency personnel directors, and other interested organizations. Major contributors to the report are listed in appendix XI. Sincerely yours, Bernard L. Ungar Director, Federal Human Resource Management Issues Executive Summary For years, the common public perception has been that the federal gov- Purpose ernment does little to deal with poor performers, To what extent are federal supervisors identifying poor performers? Are supervisors who do identify poor performers able to deal with them? These are two key questions that GAO sought to answer in this review. This report, a follow-on to an earlier effort in the Social Security Admin- istration, provides a governmentwide perspective.’ The work was requested by the Subcommittee on Federal Services, Post Office, and Civil Service, Senate Committee on Governmental Affairs. To comply with the Civil Service Reform Act of 1978, federal agencies Background implemented performance management systems that require supervi- sors to establish performance plans specifying employee job tasks and performance expectations; monitor, evaluate, and discuss actual per- formance in relation to established standards; and prepare a written performance appraisal categorizing the employee’s overall performance for the rating period. Employees whose work is below fully successful are considered to be poor performers. The agencies’ process for dealing with these employees is to provide assistance so that they can improve their performance to the fully successful level. However, if improvement does not occur, supervisors are expected to propose actions against these employees. These actions can include reassignment, demotion, or removal. To obtain the information for this report, GAO visited 2 locations within each of 10 federal agencies, 3 state governments, 3 local governments, and 12 private corporations. It also sent questionnaires to a govern- mentwide random sample of 650 supervisors to obtain their experiences in identifying and dealing with poor performers during fiscal year 1988. GAO did not evaluate the appeals process. Contrary to general perceptions, supervisors were generally willing to Results in Brief deal with their poor performers and expressed a willingness to deal with them in the future. Although implementing the process was considered unduly difficult and time-consuming, supervisors were generally able to ‘Poor Performers: How They Are Identified and Dealt With in the Social Security Administration (GAO/GGD-89-28, Jan. 27,1989). Page 2 GAO/GGD91-7 Performance Management Executive Summary correct or otherwise deal with the performance problems they identi- fied. Actions were generally taken to deal with employees whose per- formance remained unacceptable. Many supervisors indicated that they found it especially difficult to deal with employees from the estimated 5 percent of their work force who were performing at the level between fully successful and unacceptable (minimally successful). Because supervisors are limited by law in the actions they can take to deal with certain individuals who continue to perform at this level, the government has had to tolerate less than fully successful performance for extended periods of time. About half of the supervisors said they had experienced difficulty in implementing the process for dealing with poor performers. These supervisors cited the significant amount of calendar time that can be involved, a perceived lack of management support, difficulties in using performance standards, a perceived lack of authority to propose per- formance actions, and a reluctance to go through appeal or arbitration. In spite of such problems, most supervisors expressed a willingness to work with poor performers in the future. Because identifying and dealing with poor performers involves subjec- tivity, the current legislative and regulatory framework for dealing with federal employees places heavy emphasis on protecting employees against unfair or arbitrary treatment. Although these safeguards have resulted in a process for dealing with poor performers that may never be perceived as “quick and easy” by supervisors, there does appear to be some potential for making it less difficult. Individual motivation is a key factor in dealing with poor performers. Therefore, agency management must focus on creating an environment within which supervisors are encouraged and motivated to identify poor performers and are properly trained and supported when they attempt to deal with them. GAOalso believes Congress should consider various options, such as those described in this report, to ease the difficulties supervisors encounter when dealing with performance at the minimally successful level for extended periods of time. Page 8 GAO/GGD-91-7 Performance Management Executive Summary Principal Findings Supervisors Have Been According to the questionnaire responses GAOreceived, about 5.7 per- Able to Deal With Many of cent of the estimated 1.6 million employees supervised by the respon- dents performed below the fully successful level in fiscal year 1988. the Poor Performers They About 62 percent of the employees who were identified in the responses Identified as poor performers either improved their performance to fully suc- cessful, voluntarily agreed to vacate their positions, or had performance actions such as demotion or removal proposed against them. The remaining 38 percent were still poor performers at the time GAOcom- pleted its work. GAO'Sanalysis at the 20 locations it visited showed sim- ilar results. The federal government’s process for dealing with poor performers is similar in several respects to the process used by the state and local gov- ernments and private corporations GAOvisited. One major difference is that in many of the nonfederal agencies, employees whose performance remains at the minimally successful level are subject to demotion or removal. At the federal level some employees are subject to demotion or removal for prolonged minimally successful performance, but others, such as General Schedule and Wage Grade employees, are not. Seven of the eight private corporations and two of the four state or local governments GAOvisited that had a minimally successful rating category said their supervisory options for dealing with these employees included demotion or removal. Federal legislation governing General Schedule and Wage Grade employees authorizes demotion or removal only for unacceptable per- formance. Therefore, supervisory options for dealing with minimally successful performers from these employee groups are limited to reas- signment or denial of within-grade pay increases. The options of reassignment and denial of within grade pay increases provide little leverage for federal supervisors and little incentive for employees to improve poor performance, especially since employees near the top of their grade are due such increases only once every 2 or 3 years, Thus, the minimally successful performance can last for extended time frames. For example, about 83 percent of the poor performers iden- tified by supervisors were considered to have been minimally successful, and about 27 percent were minimally successful for 12 months or more. Page 4 GAO/GGDBl-7 Perionnance Management ExecutiveSummary The Processfor GAO found that many of the supervisors who identified poor performers Identifying and Dealing needed to work with them for long periods of time. In the cases GAO analyzed where employees improved to fully successful, supervisors With Poor Performers Can had worked with the employees for periods of time ranging from less Be Difficult to Implement than a month to 44 months. The average period of time was 10 months. About 60 percent of the employees who were still poor performers at the time of GAO'S review had been receiving assistance from 9 to 18 months. About 10 percent of the supervisors who responded to GAO'S question- naire said they would be unlikely to propose a performance action against a poor performer in the future. The two most frequently cited reasons for this were the time involved and a perception that they did not have management support. About one-third of the 10 percent cited the difficulty in using performance standards to measure performance and a reluctance to go through the appeal or arbitration processes. Although no one specific problem in dealing with poor performers was cited by more than 22 percent of the supervisors, 51 percent of the supervisors identified some type of problem. This indicates that dealing with poor performers can often be difficult. A Greater Management To make it easier for supervisors to identify and deal with poor per- Commitment Would formers, top management needs to be committed to, and actively involved in, creating an environment in which managers and supervi- Enhance Efforts to Deal sors are encouraged to undertake this important task and are supported With Poor Performers throughout the process. GAO believes that agencies could better demonstrate that commitment and provide a more positive environment for supervisors in several ways. They could begin by improving their oversight of the process for identifying and dealing with poor performers. At 16 of the 20 locations GAO visited, personnel officials said they did not receive or maintain sta- tistics on how poor performance cases were being resolved or how long it took to deal with performance problems. Without such information, it is difficult to identify problem areas, hold supervisors accountable, or become aware of situations where supervisors may need assistance. Agencies could also improve supervisory training and ensure that super- visors receive a greater degree of technical assistance from agency per- sonnel offices. Of the supervisors responding to GAO'S questionnaire, 15 to 29 percent said they had not received training in such areas as the Page 6 GAO/GGD-91-7 Performance Management Executivesummary use of performance standards or assisting poor performers. GAO found that supervisors who either received no training or considered it to be inadequate were more negative than others in their views on dealing with poor performers. recommends that the Director of OPM stress the need for a greater Recommendationsto GAO commitment and active involvement from top management in identi- the Director of OPM fying and dealing with poor performers. GAO also recommends that OPM assist federal agencies in demonstrating this commitment by 9 ensuring that all agencies provide periodic training and the necessary technical assistance to prepare managers and supervisors to deal ade- quately with poor performers; and l requiring agencies to establish methods and procedures for overseeing how well poor performers are being identified and dealt with, including ensuring that the time spent in implementing the process is not more than warranted. Congress should also consider ways to ease the difficulties supervisors Matters for encounter when dealing with General Schedule and Wage Grade Congressional employees whose performance remains at the minimally successful level Consideration after being given a reasonable opportunity to improve. GAO identified four options for congressional consideration ranging from administra- tive to legislative action. (See p. 27.) OPM agreed with GAO'S recommendations and indicated that it would Agency Comments prefer implementing GAO'S suggested options for dealing with minimally successful performers through legislation that would more closely link pay with performance and provide agencies with the authority to act against any employee performing below the fully successful level after being given a reasonable opportunity to improve. (See app. I.) The three federal employee unions commenting on GAO'S report gener- ally agreed with the information it presented. However, two of the three disagreed with some of GAO'S views on minimally successful performers. (See pp. 46,82,84, and 86.) P8ge6 GAO&dGlNU-7PerfomwnceManagement Page 7 GAO/GGD-fM Perfomumce MaMigement Contents Executive Summary 2 Chapter 1 12 Introduction Agencies Use Performance Management Systems to Assess the Performance of Their Employees 12 An Objective of Agencies’ Performance Management 14 Systems Is to Identify and Deal With Poor Performers Objective, Scope, and Methodology 15 Chapter 2 18 Supervisors Have Most Supervisors Are Not Ignoring Performance-Related Problems 19 Been Able to Deal Supervisory Options for Dealing With GS and WG 22 With Many of the Poor Minimally Successful Performers Are Limited State and Local Government and Private Corporation 26 Performers They Supervisors Have More Options for Dealing With Identified Minimally Successful Performers Conclusions 27 Matters for Congressional Consideration 29 Agency Comments 29 Chapter 3 31 The Processfor Not All Supervisors Are Identifying Poor Performers Supervisors Believed That Dealing With Poor Performers 31 32 Identifying and Took Too Much of Their Time Dealing With Poor Supervisors Have Difficulty Using Standards to Measure 33 Performers Can Be Performance Supervisors Perceive a Lack of Authority to Propose 34 Difficult to Implement Performance Actions Supervisors Perceive a Lack of Management Support 35 Supervisors Are Reluctant to Go Through the Appeal and 37 Arbitration Process Conclusions 37 Page 8 GAO/GGD-91-7 Performance Management - Cimtanta Chapter 4 39 Management Supervisors Need Training and Technical Assistance Additional Managerial Oversight Needs to Be Provided 39 41 Commitment Is Conclusions 42 Neededto Enhance Recommendations to the Director of OPM 44 Efforts to Deal With Agency Comments and Our Evaluation 45 Poor Performers Appendixes Appendix I: Comments From the Office of Personnel 48 Management Appendix II: Locations Visited During Our Review 60 Appendix III: Federal Agencies Represented in Our 61 Questionnaire Appendix IV: Questionnaire Used in Our Review 62 Appendix V: Time Supervisors Spent Assisting Poor 74 Performers Appendix VI: Questionnaire Objective, Scope, and 75 Methodology Appendix VII: Comments From the National Federation of 82 Federal Employees Appendix VIII: Comments From the American Federation 84 of Government Employees, AFL-CIO Appendix IX: Comments From the National Treasury 85 Employees Union Appendix X: Consultant Panel 87 Appendix XI: Major Contributors to This Report 88 Tables Table 2.1: Status of Employees Who Were Identified as 20 Poor Performers Table 2.2: Results of Proposed Demotions and Removals 20 Table 2.3: Number of Months Employees Were Minimally 24 Successful Performers Table 2.4: Opportunity Periods of State and Local 25 Governments and Private Corporations Table 3.1: Supervisors’ Opinions Concerning the 34 Adequacy of Performance Standards Table 4.1: Supervisors Who Did Not Receive Training 40 Table 4.2: Supervisors’ Opinions on the Adequacy of 40 Training Table VI. 1: Questionnaire Return 77 Page 9 GAO/GGD-91-7 Performance Management Cadent8 Table VI.2: Universe Estimates and Sampling Errors 79 (g&Percent Confidence Interval) Abbreviations CPDF Central Personnel Data File DC1 data collection instrument GAO General Accounting Office GM General Management GS General Schedule MSPB Merit Systems Protection Board OPM Office of Personnel Management PMRS Performance Management and Recognition System SIB Senior Executive Service WG Wage Grade Pwe 10 GAO/GGD4W7 Perf’ormance Management Page 11 GAO/GGD-91-7 Performance Management Chapter 1 Introduction The Subcommittee on Federal Services, Post Office, and Civil Service, Senate Committee on Governmental Affairs, requested that GAOobtain information on the extent to which federal agencies were identifying and dealing with employees who are poor performers. In January 1989, we issued a report on the results of our review of these activities at the Social Security Administration1 This report discusses how 2 locations in each of 10 federal agencies identified and dealt with poor performers. It also incorporates the results of a questionnaire sent to a random sample of supervisors throughout the government. The Civil Service Reform Act of 1978 requires federal agencies to estab- lish appraisal systems for assessing employees’ performance and prescribes the actions that can be taken against unacceptable per- formers. The act intended to streamline the process for dismissing poor performers. It attempted to balance the public’s need to have its busi- ness performed competently with the rights of employees to be selected and removed solely on the basis of their competence on the job. Under laws existing prior to the Reform Act, an employee could only be dismissed for unacceptable performance if dismissal would improve the efficiency of the federal service. Agencies found it very difficult to prove this to the degree required by the courts. Under current proce- dures, an agency may reduce in grade or remove an employee for unac- ceptable performance without having to demonstrate that the efficiency of the government would be improved. The Reform Act requires federal agencies to establish systems that pro- Agencies Use vide employees with feedback on how well they carry out their job Performance responsibilities in relation to management’s expectations. The Office of Management Systems Personnel Management (OPM) is responsible for playing a major role in this process by developing the policies under which each agency is to to Assessthe operate and reviewing their performance appraisal systems to deter- Performance of Their mine whether they meet the act’s requirements. Employees As required by the Reform Act, the federal agencies we reviewed had performance management systems that required supervisors to (1) establish performance plans that specify each employee’s job tasks and ‘Poor Performers: How They Are Identified and Dealt With in the Social Security Administration @ii’-89-28, Jan. 27,1989). Page 12 GAO/GGD-91-7 Performance Management Chapter 1 Introduction the performance standard that is expected, (2) monitor employee per- formance against performance expectations, and (3) prepare periodic performance appraisals to determine the employee’s level of performance. Performance Planning Supervisors are to begin the appraisal process by preparing a perform- ance plan. This plan should be prepared in consultation with the employee and contain the employee’s job tasks and the standards by which the employee’s performance will be assessed. In the plan, an employee’s job tasks can be divided into critical and non- critical tasks. Critical tasks are required by OPM regulation and are defined as major job tasks that are of sufficient importance that per- formance at the unacceptable level requires corrective action and may be the basis for removing the employee or reducing his or her grade level. Noncritical tasks, which are optional, are defined as job tasks that are important enough to require measurement but are not a key element of the employee’s job. Agencies’ procedures require that, to the extent possible, each perform- ance standard be specific and permit appraisal based on objective cri- teria clearly stated in terms of quality, quantity, and/or timeliness. Performance Monitoring Periodically throughout the appraisal period, the supervisor is expected to discuss the performance plan with the employee. The employee is to be informed of the level of performance and how it compares with the standards contained in the performance plan. At any time during the appraisal period, a supervisor can compare an employee’s performance against performance standards. The supervisor should call to the employee’s attention areas in which performance indi- cates a need to improve and take positive steps to help the employee improve his or her performance to at least the fully successful level. Performance Appraisal At the end of the appraisal period, the supervisor must determine the employee’s level of achievement toward reaching each individual job task by comparing actual performance against established standards. OPMrequires agencies to have at least three levels for assessing indi- vidual job tasks. All of the agencies we visited used either three or five, with the middle level constituting fully successful performance. Page 13 GAO/GGD-91-7 Performance Management Chapter 1 Introduction The levels of performance assigned to an employee’s job tasks are to be used as the basis for making a summary assessment of the employee’s performance, which, according to OPMregulations, must be made at one of five levels. As is the case with individual job tasks, performance at the middle level is considered to be fully successful. The two levels above fully successful constitute excellent and outstanding perform- ance, while the two levels below are for performance that is minimally successful and unacceptable. When employees are performing at either the minimally successful or An Objective of unacceptable levels, agencies’ policies and procedures require supervi- Agencies’ Performance sors to assist these individuals to improve their performance to the fully Management Systems successful level and to take appropriate action when efforts are not suc- cessful. The process for dealing with poor performers can be undertaken Is to Identify and Deal at any time during the appraisal period. It usually begins with informal With Poor Performers counseling and closer supervision. OPMguidance suggests that when a supervisor discusses poor perform- ance with an employee, the supervisor make a’note to the file, with a copy to the employee, which documents the matters discussed and any assistance offered. If unacceptable performance continues, the super- visor must inform the employee that he or she is being given an opportu- nity period to improve performance. Neither the Reform Act nor OPM regulations require that this notification be in writing; however, OPM strongly recommends a written notice to maintain a complete agency record. Supervisors who give an employee an opportunity period to improve performance identify the employee’s deficiencies, the action to be taken by the employee to improve his or her performance, the assistance to be provided by the supervisor, and the length of the opportunity period. The agencies we visited did not require a specific length of time for the opportunity period, stating instead that supervisors were expected to give the poor performer a reasonable period to improve his or her per- formance, depending on the circumstances in each case. Five of the agencies provided general guidance related to the timing of opportunity periods. Three of the five suggested a minimum time frame, and two provided information on how long an opportunity period might nor- mally be expected to last. Of these latter agencies, one said an opportu- nity period would normally last 60 days; the other considered a reasonable period to be from 30 to 60 days. Page 14 GAO/GGD-91-7 Pedormance Management Chapter 1 Introduction If an employee’s performance is unacceptable at the conclusion of the opportunity period, action can be proposed to remove or demote the employee. The Reform Act requires that the employee is entitled to advance written notice of the proposed action that identifies specific instances of unacceptable performance by the employee and an opportu- nity to respond to the proposed action. The employee is also entitled to a written decision that specifies the instances of unacceptable perform- ance on which the action is based. Most federal employees who are removed or demoted have the right to appeal to the Merit Systems Pro- tection Board (MSPB), which is responsible for adjudicating employee appeals of removals and demotions for unacceptable performance.2 In instances where an appeal is made, the burden of proof is on the employing agency. If employees are in an organization represented by a union, they can either use the grievance and arbitration process con- tained in the labor-management contract or appeal to MSPB, but not both. The other category of poor performance is minimally successful, and actions that can be taken to deal with these individuals vary. Continued minimally successful performance can lead to removal if the employee is a member of the Senior Executive Service (SES).Also, under the Per- formance Management and Recognition System (PMRS),~ minimally suc- cessful performers do not receive a full general pay increase and such performance can lead to reassignment, demotion, or removal. However, minimally successful performance by General Schedule (GS) and Wage Grade (WG) employees can only result in reassignment or the denial of any within grade salary increase they may be due. The law only autho- rizes the removal or demotion of a GS or WG employee if performance is unacceptable. Our objective was to determine how federal agencies were identifying Objective, Scope,and and dealing with poor performers. For the purpose of this review, we Methodology considered poor performers to be those employees performing below the fully successful level-either at minimally successful or unacceptable. Our review did not include an assessment of how poor performers in SES were identified and dealt with. We visited 2 locations within each of 10 federal agencies. Using information obtained from OPM, the agencies and ‘Whether an employee has appeal rights is governed by such factors as the nature of the employee’s appointment and his or her tenure. For example, probationary employees do not have appeal rights. 3PMRSis the pay for performance system established for the government’s General Management (GM) 13 through 16 managers and supervisors. Page 16 GAO/GGD-91-7 Performance Management Chapter 1 Intxoductlon locations were judgmentally selected to obtain a mix of those with rela- tively large numbers of identified poor performers and those with few or no poor performers. To learn how organizations outside the federal government dealt with poor performers, we visited 3 state governments, 3 local governments, 12 private corporations within the geographical areas of the GAO regional offices involved in this review, and sent a letter of inquiry to the states we did not visit. We did not verify the information obtained from these organizations because their policies and procedures were not the subject of this review. We were primarily interested in identifying any aspects of their operations that might be used as ideas for improving how the federal government deals with poor performers.* Appendix II lists the locations we visited. We visited OPM and MSPB to determine their responsibilities and to obtain information on how well the agencies’ processes were working, We also interviewed personnel officials at each agency’s headquarters and at each of the 20 locations to determine policies and procedures for identi- fying and dealing with poor performers and to obtain their opinions on how well the process was working in their location. Additionally, we interviewed 68 judgmentally selected managers and supervisors at the 20 locations. In selecting these, we included both managers and supervi- sors who had poor performers and those who did not. The 20 locations we visited employed about 85,000 persons in fiscal year 1988. According to agency records at the 20 locations, 373 employees received less than fully successful appraisals in fiscal year 1988. We designed a data collection instrument (DCI) to obtain informa- tion on how these poor performers were dealt with and the results of supervisors’ efforts to improve their performance. Supervisors com- pleted DCISon 340 of these employees. The DCIS for the remaining 33 employees were not completed because the supervisors were no longer at the agencies. We verified the information in about 18 percent of the DCISto personnel records to test the accuracy of the responses we received. To obtain additional information on how the government identifies and deals with poor performers, we sent a questionnaire to a random sample of 550 supervisors selected from OPM’S database of 252,366 civilian 4Where appropriate, we have incorporated a discussion of state and local governments’ and private corporations’ procedures into the text of this report. Page 16 GAO/GGBSl-7 Performance Management chapter 1 Introduction supervisors. The purpose of the questionnaire was to obtain supervi- sors’ opinions and perceptions on the quality of the standards used to measure performance, the process for dealing with poor performers, management support of the process, and taking performance actions against poor performers. We also obtained information on their exper- iences in dealing with poor performers in fiscal year 1988. We mailed the questionnaire in April 1989 with a follow-up mailing in May 1989. Of the 550 supervisors in our sample, 396 (72 percent) returned usable questionnaires. The supervisors who responded were employed in 28 federal agencies (see app. III). The questionnaires we received repre- sented approximately 181,704 of the 252,366 civilian supervisors identi- fied by OPM'S database. Our sample was designed to yield estimates that are precise within * 5 percent of the true population at a 95-percent confidence level. However, in some instances where only a subset of the population responded to a question, our results could be less precise because of the decrease in sample size. Estimates with sampling errors greater than & 5 percent are noted in the report. A copy of the question- naire is in appendix IV. As part of our effort to analyze the data and develop the report issues, we convened an eight-member panel of consultants representing private corporations, academia, and the federal government. We discussed the issues and findings in this report with them to obtain their views on how the process for dealing with poor performers was working and how it might be improved. Panel members are listed in appendix X. Our review, which was done in accordance with generally accepted gov- ernment auditing standards, did not include an assessment of the gov- ernment’s appeal or arbitration processes. It took place between June 1988 and August 1989. Our scope and methodology are discussed in greater detail in appendix VI. This appendix also contains the universe estimates and estimate ranges that appear in this report. Written comments on a draft of this report were provided by the Director of OPM, the Executive Director of the National Federation of Federal Employees, the Field Services Department Director of the Amer- ican Federation of Government Employees, and the National President of the National Treasury Employees Union. OPM'S comments are summa- rized and evaluated in chapters 2 and 4 and are included in appendix I along with our additional analysis. The comments received from the employee unions are discussed in chapter 4 and included in appendixes VII, VIII, and IX. Page 17 GAO/GGD-91-7 Perf’ormance Management Chapter 2 Supervisor Have BeenAble to Deal With Many of the Poor Performers They Identified Contrary to general perception, most supervisors have been dealing with their poor performers. Supervisors often assisted poor performers by recommending formal training or by providing on-the-job training, additional supervision, and counseling on job performance. When employees’ performance remained unacceptable, supervisors proposed performance actions against them. We found that 62 to 70 percent of the poor performers identified by supervisors either improved their performance to fully successful, vol- untarily agreed to vacate their positions, or had performance actions such as demotion or removal proposed against them. Also, about 76 per- cent of the supervisors responding to our questionnaire indicated a will- ingness to deal with poor performers in the future. We found, however, that supervisors were experiencing problems dealing with GS and WG employees who continued to perform at the mini- mally successful level. Unlike PMRS and SES employees, whose continued performance at the minimally successful level can lead to demotion and/ or removal, supervisors’ options for dealing with poor-performing GS and WG employees are limited to reassignment or the denial of their peri- odic within-grade increases. For employees near the top of their grade, such increases are due only once every 2 or 3 years. Thus, these options provide employees with little incentive to improve. The limited supervi- sory options for dealing with such situations have resulted in employees remaining in the minimally successful rating category for extended periods of time. The federal process used to deal with poor performers has many simi- larities to the process used by the state and local governments and pri- vate corporations that we visited. One notable exception, however, relates to how many of these organizations dealt with the group of employees causing problems for federal supervisors-minimally suc- cessful performers. Of the eight corporations that had this rating cate- gory, seven had a policy that such employees could be removed if their performance did not improve. Page 18 GAO/GGD-91-7 Periomwnce Management chapter 2 Supervlsom Have Been Able to Deal With Many of the Poor Performers They Identified Federal Personnel Manual states that one purpose :of the perform- Most Supervisors Are OPM’S ante appraisal system is to help managers and supervisors recognize Not Ignoring and deal more effectively with performance deficiency problems. The Performance-Related policies and procedures at the 10 agencies we visited state that supervi- sors are expected to assist poor performers and propose performance Problems actions against employees who are performing at the unacceptable level after being provided with an opportunity to improve. All of the poor performers whom supervisors have identified and may be dealing with are not necessarily documented through the appraisal process. For example, supervisors are authorized to work with poor per- formers without preparing performance appraisals. Also, poor per- formers may either improve their performance or vacate their position at any time during the appraisal period. Consequently, there are more poor performers throughout the govern- ment than the number that are formally appraised as such. For example, OPM data show that about 0.6 percent of federal employees were rated below fully successful in fiscal year 1988. However, according to the questionnaire responses we received from supervisors, an estimated 89,500, or 5.7 percent, of the estimated 1.57 million employees they supervised performed below the fully successful level at some time during fiscal year 1988.1 We estimated that these employees received annual salaries of approximately $2.7 billion. The information in table 2.1 was compiled from the questionnaire and the 340 DCIS to show the results of supervisors’ efforts to deal with their poor performers. Overall, our questionnaire analysis showed that about 98 percent of an estimated 50,014 supervisors who had poor performers provided counseling, additional supervision, and/or training to assist these individuals. ‘This report contains information from (1) a questionnaire sent to a random sample of supervisors from OPM’s database and (2) DCIs. Unless otherwise stated, the results from the questionnaire are extrapolated to represent the universe. The DC1information represents the actual number of DCIs that were completed. Page 19 GAO/GGD-91-7 Perlormance Management Chapter 2 Supervbwr~ Have Been Able to Deal With Many of the Poor Performers They Identified Table 2.1: Statue of Employeea Who Were Identified a8 Poor Performers Employees in questionnaire8 Emploveerr in DClr Status of employee Number Percent Number Percent lmoroved to fullv successful 71 38 116 34 Voluntarily agreed to: Demotion 0 0 4 1 Reassianment 15 8 28 a Fiesionation 13 7 16 5 Retirement 9 5 11 3 Other 3 2 17 5 Subtotal 40 22 76 22 Performance action orooosed I I 4 2 46 14 Still a poor performer 70 38 102b 30 Total 165 100 340 100 Note: The employees in our DCls were rated as either minimally successful or unacceptable. The employees identified as poor performers in the questionnaires were not necessarily rated as such aThe questionnaire data in this table are not extrapolated to represent the universe. ‘At the time of our review, these employees had been poor performers for an average of 12 months and the range was 2 to 32 months. Table 2.2 shows the results of the 46 proposed actions-15 for demotion and 31 for removal-identified from the DCIS.As the table shows, pro- posals for demotion and removal were often resolved in other ways. Table 2.2: Results of Proposed Demotlonr and Removal8 Demotion Removal Actions completed: Demotion 8 1 Removal 0 11 Reassignment 5 4 Resignation 0 7 Retirement 0 7 No action taken 0 1 Subtotal 13 31 Actions in process 2 0 Total 16 31 The following examples obtained from the DCISillustrate how some supervisors have dealt with their poor performers. . Bill, a GS-1 1 computer programmer, was informed of his poor perform- ance on October 3,1988. His supervisor counseled and worked with him Page 20 GAO/GGBI)l-7 Performance Management Chapt43r 2 Super&on Have Bean Able to Deal With Many of the Poor Performers They Identifkd informally, and on November 1, 1988 (about 1 month later), his per- formance improved to fully successful. According to Bill’s supervisor, Bill just needed an additional push to get some tasks accomplished in a more timely manner. l On July 1, 1988, Bob, a GS-12engineer, was informed of his poor per- formance. According to his supervisor, Bob was not self-motivated and historically had not been a strong performer. The supervisor increased the amount of supervision and put Bob on a performance improvement plan. On October 4, 1988 (about 3 months later), Bob improved his per- formance to fully successful. . John, a GS-12,was informed at a counseling session on February 8, 1988, that his performance was less than fully successful. To assist John, his supervisor (1) provided him with an opportunity period to improve his performance, (2) increased the amount of his supervision, (3) provided on-the-job training, (4) counseled and worked with him informally, and (6) referred him to counselling services. The efforts were unsuccessful, and the consensus among supervisory and training officials was that John could not perform at the Gs-12 level. On April 16, 1988, the super- visor proposed that John be removed; on May 16,1988 (about 3 months after he was informed of his poor performance), John was removed. As shown in the following examples obtained from our IICI analysis, supervisors were sometimes willing to spend significant amounts of time assisting their poor performers.2 . Joe, a GS-13engineer, was informed during a counseling session on June 18, 1986, that although there were several things he did well, one area of his work was minimally successful. The supervisor increased the level of supervision, and on November 1,1988 (over 28 months later), Joe’s performance improved to fully successful. . Sarah, a minimally successful GM-E, was informed during a counseling session on March 16, 1988, that her performance was less than fully successful. The supervisor indicated that Sarah was not able to delegate tasks and had trouble following up on assignments and bringing them to completion. Sarah’s supervisor counseled her, provided on-the-job training and increased supervision, and arranged for Sarah to receive formal training. Sarah’s performance did not improve. On February 12, 2Supervisors usually begin to deal with a poor performer by providing informal counseling and super- vision. In our LKX, we did not ask supervisors to distinguish between the time spent working with poor performers informa.lly and the time spent during formal opportunity improvement periods. We also did not ask supervisors to comment on the reasonableness of time frames for specific cases of poor performance. Page 21 GAO/GGDgl-7 Performance Management Chapter 2 Supervisors Have Been Able to Deal With Many of the Poor Performers They Identiaed 1989 (about 11 months after she was informed of her poor perform- ance), she voluntarily agreed to accept a demotion to a GS-14 position. The supervisors of the 116 employees in our DCI analysis who improved to fully successful worked with their employees an average of about 10 months before their performance improved to that level. The times ranged from less than a month to 44 months. The median time was 8 months.3 Supervisors said they spent over 1 year each assisting 31 of these employees. The time supervisors spent assisting the employees in our DCISis included in appendix V. Even though the process can take a significant amount of time, about 76 percent of the supervisors responding to our questionnaire indicated a willingness to continue to deal with poor performers in the future. About 83 percent of an estimated 89,600 poor performers identified by Supervisory Options supervisors were performing at the minimally successful level. We do for Dealing With GS not have data on how many of these poor performers were GS, WG, or and WG Minimally PMRSemployees, However, it is probable that most of the poor per- formers were in the GSand WG groups since these groups constitute Successful Performers about 93 percent of the total number of employees in the three groups Are Limited combined. The minimally successful performance level poses a problem for super- visors of GSand WG employees because they are limited in the actions they can take to deal with employees who remain at this level for extended periods of time. These employees cannot be rated as unaccept- able because their performance in relation to the standards by which they are assessed has not been poor enough to warrant such a rating. Similarly, their performance has not been determined to be good enough to meet the standards for a fully successful rating. According to OPM reg- ulations and discussions with OPMofficials, supervisors can only reas- sign such employees or deny their within-grade pay increases. These options provide little leverage for the supervisor and little incentive for the employee to improve performance, especially since employees near the top of their grade are due such increases only once every 2 or 3 years. Thus, the minimally successful performance can last for extended periods of time. 3Data show that the agencies that provided criteria for an opportunity period did not deal with their poor performers any faster than the agencies that did not. Page 22 GAO/GGD91-7 Performance Management chapter 2 Supervlmm Have Been Able to Deal With Many of the Poor Performers They Identifled SomeSupervisors View An estimated 33,966, or 19 percent, of the supervisors in our question- Limited Options for naire indicated that not being able to demote or remove a minimally suc- cessful employee has decreased the likelihood that they will deal with Dealing With Certain these employees in the future. Further, about 69 percent of the 88 per- Minimally Successful sonnel officials, managers, and supervisors we interviewed said that Employees as a Problem minimally successful performers are a problem. The following are exam- ples of comments we received from interviews and questionnaires regarding the minimally successful performance level. . One manager we interviewed said that minimally successful employees do not present any problem in his division because they are all rated “fully successful.” . Another manager indicated that existing regulations were not very helpful in dealing with minimally successful employees because little can be done other than to reassign them and they can remain in that level of performance “forever.” . A supervisor commented, “I have seen people that are minimal that are dead ended in jobs they never leave . . . and they are poor performers for years and years.” . Another supervisor said “The inability to remove a marginal employee is a serious problem. I recommend that, after two go-day improvement periods, those remaining marginal become subject to removal, etc.” . A third supervisor said that “the minimally successful performer is the most difficult to deal with since there are so many gray areas.” GS and WG Employees Can Some employees remain at the minimally successful level for an Remain Minimally extended period of time. For example, the 194 employees in our DCIanal- ysis who were rated minimally successful performed at this level from Successful for Extended 11 days to about 44 months, and averaged about 10 months. Because 68, Periods of Time or about 30 percent, of these minimally successful performers were still receiving assistance at the time we completed our analysis, the final average for the group will be even longer. The responses to our questionnaire also provided evidence of the extended periods of time during which employees can remain minimally successful. An analysis of these responses from supervisors indicated that about 4.7 percent of their approximately 1.67 million subordinates had performed at the minimally successful level at some time during fiscal year 1988. An estimated 20,189, or 27 percent, of these individ- uals were said to have been performing at this level for 12 months or more. Page 22 GAO/GGDI)l-7 Performance Management Chapter 2 Supervisors Have Been Able to Deal With Many of the Poor Performera They Identifled Table 2.3: Number of Months Employees Were Minimally Successful Performers Range of time performance was minimally successful Percenr Less than 6 months 46 6 to less than 12 months 27 12 months or more 27 aThe sampling error for these estimates is greater than 5 percent. See appendix VI One of the cases cited in our earlier report on the Social Security Admin- istration illustrates the amount of effort exerted and the frustration experienced by supervisors when attempting to help minimally suc- cessful employees improve their performance. In a progress report, a supervisor told an employee that her performance for the previous 3 months was minimally successful and that her within-grade increase would be denied. The employee continued to perform at the minimally successful level for over 3 more years. During that time, we identified at least 10 progress reviews held with the employee as well as an offer of additional training, which the employee declined. Finally, after 3 l/2 years, the supervisor stopped formal efforts to encourage the employee to improve, even though her performance was still only minimally successful. OPMregulations require agencies’ performance appraisal systems to have five summary rating levels, including a level for minimally successful performance. Also, several agencies are using a minimally successful category for rating individual performance elements. Because the Civil Service Reform Act only provides supervisory options for dealing with unacceptable performance by GSand WGemployees, supervisors are lim- ited in the options they can take to deal with individuals whose per- formance falls into the minimally successful category. Legislation Contains There is precedent under current federal law for taking additional Additional Options for actions to deal with the minimally successful performer. For example, 5 U.S.C. 4314(b)(4) states that any senior executive who twice in any Supervisors to Consider period of 3 consecutive years receives less than fully successful ratings When Dealing With PMRS shall be removed from the Senior Executive Service. and SESEmployees Also, Public Law 101-l 03, enacted in September 1989 to extend PMw for ~~-13 through -15 employees, provides for placing any employee whose performance has been rated less than fully successful on a performance Page 24 GAO/GGD91-7 Performance Management Chapter2 Supetira Have Been Able to Deal With Many of the Poor Performera They Identified Chapter 2 Supervi~om Have Been Able to Deal With Many of the Poor Performers They Identified improvement plan. If the employee fails to attain at least the fully suc- cessful level after a reasonable time, the employee can be reassigned, reduced in grade, or removed. In May 1990, the administration submitted a proposal to reform the pay of federal white collar (PMRSand GS) employees. Under the proposal, employees performing at the fully successful level or above would be eligible for performance-based additions to their pay. However, employees judged to be doing less than fully successful work would not receive any pay increases. The performance appraisal systems and the processes for dealing with State and Local poor performers used by the state and local governments and private Government and corporations we visited are generally similar to those of federal agen- Private Corporation cies. The state and local governments’ and private corporations’ systems generally require supervisors to (1) establish performance plans con- Supervisors Have taining tasks and standards, (2) monitor employee performance against More Options for expectations, and (3) prepare performance appraisals. When a poor per- Dealing With former is identified, these systems generally allow for a period within which to improve performance. However, contrary to federal supervisor Minimally Successful options for minimally successful GS and WGemployees, some nonfederal Performers supervisors have the additional options of demotion and removal. We visited a total of 18 state and local governments and private corpo- rations. Of the 17 from which we received information regarding an opportunity period for improvement, 16 said that they provided one. Their responses as to how long their opportunity periods lasted are sum- ” marized in table 2.4. Table 2.4: Opportunity Periods of State and Local Governments and Private Time frame Number -__- -___---__ Corporation8 1.5 months 1 3 months 4 4 months 1 1 to 3 months 3 1 to 4 months 1 2-----_ to 4 months . A’ Chapter 2 Supervkrs Have Been Able to Deal With Mnny of the Poor Performers They Identified Like federal agencies, the state and local governments and private cor- porations generally use the opportunity period to discuss the (1) employee’s current level of performance, citing those areas requiring improvement; (2) specific level of performance expected in each of the identified deficiencies; (3) time frames for improvement; and (4) steps management will take to support improvement. If job performance remains unacceptable at the end of an opportunity period, the state and local governments’ and private corporations’ procedures generally call for action to remove, demote, or reassign the employee. We found that in the private sector and at the state and local levels, removal and demotion actions can also be taken for minimally suc- cessful performers. Of the 12 private corporations we visited, 8 had a rating category equivalent to minimally successful, and 7 had a policy that minimally successful employees could be removed. For example, one corporation’s procedures stated that the minimally successful employee is someone who is not consistently meeting the requirements of the job. The procedures further stated that the corporation could not afford to have someone in that situation very long and that there either must be improvement, early consideration for reassignment, or a deci- sion to remove the employee. Of the six state or local governments we visited, four had a rating category equivalent to minimally successful; two had a policy allowing minimally successful employees to be demoted or removed. Thirteen of the other states we contacted also had such a rating category, and 12 allowed minimally successful employees to be demoted or removed. Seventeen of the organizations we visited also provided an avenue of appeal. At the state and local governments we visited, the process was similar to that of the federal government in that employees could appeal to state or local personnel boards. However, the process for the private corporations we visited differed in that the highest level of appeal- other than filing a lawsuit-was the corporation president or chief exec- utive officer. The relationship between employers and employees in the private sector-even with regard to performance problems-is generally per- ceived as being governed almost exclusively by the doctrine of “employ- ment-at-will,” which can be characterized as the right of an employer to fire an employee for any reason and at any time. However, the employ- ment-at-will doctrine has been significantly eroded as a result of several Page 26 GAO/GGD-91-7 Performance Management chapter2 511pervioor11Have Been Able to Deal With Many of the Poor Performers They Identif’ied A fourth option would be to amend the Civil Service Reform Act to pro- vide supervisors with the additional options of demoting and/or removing GS and WGemployees who perform at the minimally successful level for extended periods. On a case-by-case basis, supervisors could choose from among the available options after giving due consideration to the individual facts and circumstances. Employees would, of course, retain their rights to appeal or arbitration. This option would enable supervisors to deal more effectively with poor-performing GS and WG employees and would make supervisory options for doing so more con- sistent with those now available for PMRSand SIB personnel. We favor a combination of the third and fourth options. Such actions would result in more consistency across employee groups regarding the opportunity to earn extra pay for good performance and the denial of pay increases for poor performance. They would also result in more con- sistency across employee groups regarding the case-by-case options from which supervisors can choose for dealing with poor performance. Finally, such actions could be taken without affecting employee rights for appeal or arbitration. Congress should consider ways to ease the difficulties supervisors Matters for encounter when dealing with GS and WG employees who have performed Congressional at the minimally successful level for lengthy periods. The range of Consideration options considered should include the following: . have agencies change performance standards so that what is now con- sidered to be minimally successful performance becomes unacceptable; . eliminate the minimally successful rating category by law or OPM regulation; . enact legislation to link pay to performance more closely for GS and WG employees, while giving supervisors authority to deny pay increases to long-term minimally successful performers; and . enact legislation to give supervisors the additional options of demoting and/or removing rs and WGemployees who remain at the minimally suc- cessful level after being given a reasonable opportunity to improve. OPMconcurred in our reservations about the first two options we Agency Comments presented. With regard to option 2, OPMpointed out that because agen- Y ties have some discretion in determining the number of rating levels to Page 29 GAO/GGDBl-7 Performance Management chapter 2 Supervlsor~ Have Been Able to Deal With Many of the Poor Performers They Identified and have resulted in performance at the minimally successful level for extended periods of time. One option for enabling supervisors to deal more effectively with mini- mally successful performers would be to redefine performance stan- dards so that performance currently considered minimally successful would become unacceptable. Making the standards more stringent could result in some minimally successful employees moving to the unaccept- able rating category. In all probability, however, there would still be employees who would meet the revised standards for minimally suc- cessful performance and remain at that level for extended periods. Also, the inconsistency in supervisory options across employee groups would remain. A second option would be to eliminate the minimally successful rating category completely so that employees who would otherwise have been rated as minimally successful would have to be rated as either unaccept- able or fully successful. This could be accomplished if agencies opted to or were required to eliminate the minimally successful rating category from assessment of individual job tasks and if OPMchanged its regula- tions regarding summary ratings for GSand WGemployees. The problem with this option, however, is that-in all likelihood-the performance of many employees who are minimally successful would not be consid- ered poor enough to be rated as unacceptable and would thus be rated as fully successful. Therefore, although the rating category would be elimi- nated, the type of performance it identifies would not. Also, the dis- parity in supervisory options across employee groups would still exist. A third option would be to enact legislation that more closely links pay to performance for Gs and WGemployees. Such legislation could give supervisors the discretion to deny pay increases to employees who per- form below the fully successful level for extended periods without a legitimate reason. Alternatively, legislation could be enacted to prohibit pay increases to employees rated below the fully successful level as pro- vided for by the administration’s May 1990 proposal for white collar (PMRSand GS)employees. Denying pay increases to poor performers could serve as an incentive for them to either improve their perform- ance or vacate their positions. It would also put them on a more consis- tent basis with PMRSand SESpersonnel relative to pay for performance. However, this action would not by itself preclude the employee from choosing to continue performing at the minimally successful level for an extended period. Neither would it completely eliminate the inconsis- tency in supervisory options across employee groups. Page 29 GAO/GGD-91-7 Performance Management The Processfor Identifying and DealingWith Poor PerforormemCaylBe Difficult to Implement Although most supervisors have been dealing with their poor per- formers, a significant number have not. Personnel officials, managers, and supervisors believed that poor performers were not always being identified through the appraisal process because supervisors did not want to follow the process for dealing with them. Many of the supervisors we contacted indicated that identifying and dealing with poor performers can be a difficult and time consuming task. An estimated 91,770, or 61 percent, of the supervisors experienced one or more problems, including the amount of time involved in dealing with such individuals, difficulty in using existing performance stan- dards to measure performance, a perception that they did not have suf- ficient authority or lacked management support in proposing performance actions, and a reluctance to go through the appeal or arbi- tration process. Although no one problem was cited by more than 22 percent of the responding supervisors, taken collectively, these problems indicate that the task of identifying and dealing with poor per- formers may be more difficult than it needs to be. Personnel officials, managers, and supervisors said that poor per- Not All Supervisors formers are sometimes rated fully successful, and thus not formally Are Identifying Poor identified, because supervisors did not want to use the process to deal Performers with them. Our questionnaire analysis provided further evidence that not all supervisors are identifying poor performers. An estimated 19,730, or 11 percent, of the supervisors would be unlikely to use their agency’s process to deal with poor performers in the future, primarily because the process takes too long and uses too much of their time. Other reasons cited by respondents included the view that the process disrupted working relationships and that they were aware of problems encountered by other supervisors who had used the process to deal with their poor performers. According to the questionnaire responses of those supervisors who had poor performers, an estimated 4,130, or 8 percent, had poor performers that they did not assist. Rather than deal with these individuals, the supervisors indicated that they reduced the employee’s workload, gave the employee easier work, hoped that the situation would work itself out, or reassigned the employee to another unit. Some examples of supervisors’ questionnaire comments concerning this matter follow. Page 31 GAO/GGD-91-7 PerPormance Management Chapter 2 , Supervisora Have Been Able to Deal With Many of the Poor Performers They Identified be used in assessing individual job tasks, the minimally successful cate- gory could be eliminated for such tasks without legislative change. How- ever, as also pointed out by OPM, the use of this approach would deny agencies the flexibility of having a minimally successful rating category. OPM agreed with our third option- adopting pay for performance for GS and WG employees. Concerning our fourth option, OPM agreed that super- visors need additional authority to deal with employees who continue to perform at the minimally successful level after being given a reasonable opportunity to improve. OPM suggested that such authority could be pro- vided by changing the wording of the law to authorize performance- based actions against employees who continue to perform “below the fully successful level.” OPM prefers this approach for dealing with minimally successful per- formers because it would retain the flexibility agencies now have to use or not use the minimally successful rating level while at the same time allowing supervisors to act against employees rated as minimally suc- cessful in one or more individual performance elements. OPM also pointed out that this approach would conform with the approach recently adopted under the PMRS program. We believe that OPM’s pro- posal provides a viable way to implement our fourth option. Page 30 GAO/GGD-91-7 Performance Management w-3 TheProeoeeforI4len~endD@alingmtb Poor Pelfomlela can Be Dmcalt toImplement working with individuals whose performance had improved. As pointed out in chapter 2, this took as long as 44 months. Some of the comments we received from supervisors regarding time included the following. It takes months of a supervisor’s time to develop sufficient documenta- tion to remove or demote an employee. It takes so long to get rid of someone that supervisors just “put up with” the poor performer. The actions leading to an employee’s removal are too long and take an inordinate amount of the supervisor’s time. Even with the Civil Service Reform Act, it is still very difficult and extremely time-consuming to take removal action. Based on an analysis of our questionnaire responses, an estimated 18,354, or 10 percent, of the supervisors would be unlikely to propose a performance action in the future. Of these, an estimated 11,930, or 65 percent, would be unlikely because the process took too much of their time.2 To evaluate the performance of their subordinates effectively, federal Supervisors Have agency supervisors must have valid performance standards and reliable Difficulty Using methods of monitoring and measuring performance in relation to pre- Standards to Measure scribed standards. Federal law (5 U.S.C. 4302(b)(l)) requires that each performance appraisal system shall provide for performance standards Performance that will “to the maximum extent feasible, permit the accurate evalua- tion of job performance on the basis of objective criteria related to the job in question. . .” The Federal Personnel Manual states that, to the extent feasible, performance standards should be realistic, reasonable, objective, and measurable. The agencies’ performance standards describe measures of performance in both non-numeric terms (i.e., descriptive words) and numeric terms (i.e., quantifiable). Table 3.1 summarizes supervisors’ opinions on the adequacy of performance standards. “The sampling error for this cstimatc is greater than 5 percent. See appendix VI. Page 33 GAO/GGDBl-7 Performance Management Chapter 3 The Process for Identifying and Dealing With Poor Performers Can Be Difficult to Implement l A supervisor said that the easiest approach (and one proven successful by other supervisors) was to ignore the problem and assign any essential work to other employees. . A supervisor responded that he would not propose a performance action. He wrote that, historically, his organization has passed on problem employees to a new supervisor instead of dealing with the problem or confronting the employees. He said that if he inherited a problem employee, he would not put his neck on the line to attempt to correct a long-standing deficiency unless he was positive that manage- ment would support the action. l A supervisor commented that most supervisors try to ignore the problem if possible and hope it will go away. . A supervisor said that the system puts a burden on the supervisor to prove and re-prove the employee’s failings to a point that it is not worth the time and effort for supervisors to pursue that course of action. . A supervisor commented that she had a poor performer who had been an ongoing problem. The employee had been placed in jobs she was unable to handle for the last 5 years. Rather than dealing with the real problem of her inability to supervise and handle multiple tasks with constant interruptions, she was passed around, consequently causing problems for many people. Over 60 percent of the 88 personnel officials, managers, and supervisors interviewed at the 20 locations we visited also expressed a concern that not all supervisors were identifying their poor performers. Some of the reasons they gave for this nonidentification were that . the process is too time-consuming, . supervisors want to avoid confrontation, and l supervisors believe they lack management support. Our questionnaire analysis showed that supervisors spent an average of Supervisors Believed about 5 hours a week providing assistance to each poor performer they That Dealing With had; our DCI analysis showed that this time commitment could extend Poor Performers Took for several months.1 For example, of the 102 employees who were iden- tified in our DCIS as poor performers and who were still in that category Too Much of Their at the time we completed our review, 38 had been rated as unacceptable. Time Fourteen of these individuals had been assisted for periods of time ranging from 13 to 32 months and actions to deal with them still had not been proposed. Supervisors also had to spend long periods of time ‘The sampling error for this estimate is greater that 6 percent. See appendix VI. Page 32 GAO/GGD91-7 Performance Management The Proccae for Identifyhg and Dealing With Poor Performem Can Be Diffhlt to Implement An estimated 28,907, or 16 percent, of the supervisors perceived a lack Supervisors Perceive a of adequate management support for taking performance actions Lack of Management against unacceptable performers. Examples of their comments follow. Support l Most upper level managers are reluctant to support actions dealing with employee performance. They often use a common “cop out” such as “the employee’s performance is your own/our failure to deal with perform- ance.” The end result often leads to frustration and a passive attitude with first and second level supervisors and managers, . Increasingly, management is apathetic to supervisors trying to work through or with a performance problem. Often, evidence and documen- tation clearly indicate the problems, yet management refuses to be responsive to the supervisor. . Management discourages giving anyone a less than fully successful rating because of the problems that can arise if the employee decides to fight the rating. . Higher levels of management and some involved in personnel work may be supportive to little or no extent, often creating confusion and frustra- tion for the employee and supervisor. The following LZI case highlights the lack of management support that can be experienced by a supervisor. Jack, a GS-13 management analyst, was first recognized as a poor per- former on March 27, 1988. He was provided with informal feedback, counseling, and written comments on his completed work. During his progress review on June 29, 1988, his supervisor provided him with written comments on the specific tasks that needed to be improved so that he could be rated at the fully successful level. Jack’s supervisor also offered to meet weekly with him to discuss assignments and per- formance, recommended specific training that could improve perform- ance, and referred him to a counseling service. Jack was given until September 30, 1988, to improve. Before the end of the rating period, the supervisor recognized that Jack’s performance had further deteriorated and sought assistance from both her immediate supervisor and Employee Relations. It was decided that Jack’s rating would be delayed and he would be given a 60-day written notice to improve. Jack’s supervisor started writing the notice in August 1988; it was not completed until October 7, 1988. According to the supervisor, the notice was delayed because of the complexity of Jack’s position, coupled with Employee Relations’ concern that all docu- mentation be absolutely defensible (e.g., the deficiencies be accurately Page 35 GAO/GGD-91-7 Performance Management Chapter 3 The Process for Identifying and DeaUng With Poor Performers Can Be Difficult to Implement Table 3.1: Supervisors’ Opiniona Concerning the Adequacy of Non-numeric Performance Standards standards Numeric standards Number Percent Number Percent Adequate 107,829 75 69,286 648 Neither adequate nor inadequate 20,189 14 18,354 17 Inadequate 16,519 11 20,648 19 BThesampling error for this estimate is greater than 5 percent. See appendix VI. Among supervisors citing problems with standards, the most frequently stated problem was that it was difficult to measure the employee’s per- formance against the standards. About 33 percent of an estimated 18,364 supervisors who indicated that they would be unlikely to pro- pose a performance action in the future said the standards were not suf- ficient to support the action.3 Lack of authority was viewed as a problem by an estimated 39,002, or Supervisors Perceive a 22 percent, of the supervisors. This sense of insufficient authority was Lack of Authority to higher among supervisors who had experience dealing with poor per- Propose Performance formers than among those who had not. About 32 percent of the super- visors with poor performers perceived a lack of authority.4 Only about Actions 18 percent of the supervisors without poor performers had this problem. Our questionnaire analysis showed that supervisors who believed they lacked authority to propose actions were less likely to do so than super- visors who believed they had the authority. Of the estimated 39,002 supervisors who believed they lacked authority, about 21 percent were unlikely to propose action against poor performers.6 Of the supervisors who believed they had authority, only about 7 percent were unlikely to propose action. ?he sampling error for this estimate is greater than 6 percent. See appendix VI. ‘The sampling error for this estimate is greater than 6 percent. See appendix VI. %he sampling error for this estimate is greater than 6 percent. !%e appendix VI. Page 34 GAO/GGD91-7 Performance Management Chapter 8 The Process for Identifying and Dealing With Poor Performem Can Be Difficult to Implement Over 40 percent of the personnel officials, managers, and supervisors Supervisors Are we interviewed said that the potential for an employee using the appeal Reluctant to Go or arbitration process would affect a manager’s or supervisor’s willing- Through the Appeal ness to pursue a performance action. and Arbitration Some of the comments we received are as follows. Process . “The appeal/arbitration process is a very definite deterrent due to the fear of the unknown and the fact that you are out on a limb.” . “The supervisor knows that it will be a time-consuming, uphill battle because the burden of proof is on the supervisor.” . “It is a painful process, but it is needed.” . “The balance with respect to documenting and dealing with poor performers is on the side of the employee. The burden of proof rests almost exclusively with the supervisor, requiring much time and an incredible amount of documentation before appropriate action can be taken by the supervisor. Many first line supervisors are reluctant to take action for this reason. The grievance/appeals/arbitration process often assumes the supervisor to be ‘guilty’ and the problem employee ‘innocent.’ ” Our questionnaire analysis showed that about 35 percent of the esti- mated 18,354 supervisors who would be unlikely to propose a perform- ance action in the future did not want to go through the appeal or arbitration process.9 The comments from supervisors responding to our questionnaire indi- Conclusions cate that identifying and dealing with poor performers can be a difficult and time-consuming task. We agree with this observation. It is difficult to inform employees of performance problems. In addition, the interac- tion associated with providing employees with the opportunity to improve their performance can result in the preparation of a certain amount of documentation and require a great deal of a supervisor’s time. It would be unrealistic to expect to eliminate all of the difficulty associ- ated with this process. However, it is unfair to expect managers and supervisors to operate in an environment where identifying and dealing with poor performers tends to be much more difficult than it needs to be. The problems experienced by supervisors indicate that this has fre- quently been the case. OThesamplingerror for this estimate is greater than 6 percent. !3eeappendix VI. Page 37 GAO/GGD-91-7 Perf’ormance Management chapter 3 llw Process for Identifying and Dealing With Poor Performers Can Be DifTkult to Implement described within the scope of the performance standards, the marginal level of performance necessary to retain the position be accurately described, etc.). Jack’s performance did not improve during the opportunity period; therefore, his supervisor felt that a proposal to remove him was appro- priate. Jack then presented a medical letter and used an excerpt from that letter in an attempt to be reassigned. The excerpt stated that the employee may benefit from a change in assignment. No further explana- tion or information was requested, and, based on that one statement, Jack was reassigned. The supervisor said that this was at least the second time that Jack has been dealt with through reassignment. This sense of inadequate support was higher among supervisors who had experience in dealing with poor performers than it was among those who had not. About 24 percent of the supervisors with poor performers perceived a lack of management ~upport.~ Only about 13 percent of the supervisors without poor performers perceived this problem. According to our questionnaire analysis, of the approximately 18,354 supervisors who would be unlikely to propose a performance action in the future, about 68 percent would be unlikely to take action because they believed they did not have management’s support.7 Our analysis also showed that supervisors who believed they did not have manage- ment support to propose actions were less likely to propose actions against unacceptable performers than those supervisors who believed they had support. Of the supervisors who believed they did not have management support, about 31 percent said they were unlikely to pro- pose actions.8 Only about 6 percent of the supervisors who believed they had management support were unlikely to propose actions. ‘?he sampling error for this estimate is greater than 6 percent. See appendix VI. 7The sampling error for this estimate is greater than 6 percent. See appendix VI. %w sampling error for this estimate is greater than 5 percent. See appendix VI. Page 36 GAO/GGD-91-7 Performance Management Chap&r 4 Maagement CommitmentIs Neededto EnhanceEfforts to Deal With Poor Performers There are no easy answers to the questions of how to better motivate managers and supervisors to identify poor performers or how to make it easier to deal with them. An essential element, however, is that top man- agement needs to be committed to assisting its managers and supervi- sors in carrying out this important task and assuring them that they will be supported throughout the process. We believe that agencies could best demonstrate this commitment and alleviate the burden on the employees’ supervisors by improving super- visory training, ensuring that supervisors receive a greater degree of technical assistance from agency personnel officials, and providing more definitive and visible management oversight. Enhanced training and technical assistance could help supervisors to clearly understand their authority and responsibilities and make them aware of the assistance and support that is available to them when working to resolve perform- ance-related problems. Managerial oversight could complement this effort by reinforcing management’s interest in the performance manage- ment process and identifying situations where supervisors may need assistance. Supervisors responding to our questionnaire cited a lack of guidance as Supervisors Need one of the reasons they would be unlikely to deal with poor performers Training and in the future. Of those supervisors who indicated that they would be Technical Assistance unlikely to take performance actions, about 3,212, or 18 percent, have not been given sufficient guidance on using the pr0cess.l Supervisors need training and technical assistance to assist them in identifying and dealing with poor performers. The Federal Personnel Manual states that no effort should be spared to ensure that supervisors are fully prepared to carry out their supervisory responsibilities. The Manual states that at a minimum, agencies need a formal training pro- gram with periodic refresher training for all supervisors. The program should, at a minimum, include training to help supervisors develop and use realistic performance standards and distinguish among various levels of performance to justify rewards or initiate performance action. The Manual also states that agency personnel offices should furnish supervisors with the technical information they need to implement and manage the performance appraisal program. ‘The sampling error for this estimate is greater than 6 percent. See appendii VI. Page 39 GAO/GGDW-7 Performamx Management chapter 3 The Procelle for Identl,fyhg and IkaQbg With Poor Performem Can Be Diffkult to Implement There are no easy answers that would remedy this situation. We note, however, that supervisors need to believe that management is concerned about poor performers, that there is an interest in getting poor per- formers to improve, and that assistance will be provided if it is needed. Chapter 4 provides additional insight into the actions that could be taken to help demonstrate such a commitment. Page 33 GAO/GGDol-7 Perfommnce Management Chapter 4 Management Ckmmitment Is Needed to Enhance Efforta to Deal with Poor Performera One essential aspect of the performance appraisal process is a provision Additional Managerial for continuous review, assessment, and-when needed-revision. Fed- Oversight Needsto Be eral personnel instructions require agencies to establish methods and Provided procedures for periodically evaluating the effectiveness of their per- formance appraisal systems. At 16 of the 20 locations we visited, however, personnel officials said that they did not receive or maintain statistics on the number of poor performers given opportunities to improve, how poor performance cases were being resolved, or how long it took to deal with performance problems. Moreover, most of the program managers we contacted did not believe they had any responsibilities for managing and monitoring the process for dealing with poor performers. Management oversight would be particularly useful in helping address two of the problems cited by supervisors in identifying and dealing with poor performers-( 1) the difficulty in using standards to measure per- formance and (2) the concern that working with poor performers takes too much time. On their questionnaires, supervisors provided the following examples of specific standards they were having problems with and why they were a problem. . A numeric standard required “three to six successful industrial engi- neering studies during the year.” The supervisor commented that the problem with this standard is that studies may vary from relatively short duration (Le., a single issue economic analysis) to a multiyear plant layout. The supervisor also commented that the plant layout may not prove successful or unsuccessful until the facility is occupied. l A non-numeric standard required that an employee “lead and/or con- duct inspections/investigations of a complex and/or specialized nature.” The supervisor said that the problem with this standard is that the actual quality of individual inspections is often unknown because the employee does them away from the office. . A non-numeric standard required an employee to “process routine and most complex claims timely and accurately, utilizing the appropriate adjudicative technique.” The supervisor said that the problem with this standard is that “timely” and “accurately” are not defined. The super- visor also commented that performance standards need to be well- Page 41 GAO/GGD-91-7 Performance Management , Chapter 4 Management Commitment Is Needed to Enhance Efforts to Deal With Poor Performers According to personnel officials at each of the 20 locations we visited, initial performance management training is provided for its supervisors; however, 7 of the locations did not provide refresher training. Many of the supervisors responding to our questionnaire said they had not received training or believed their training was inadequate in the fol- lowing areas: l using performance standards, l assisting poor performers in improving their performance, and/or l initiating performance action against unacceptable performers. Table 4.1 shows the supervisors who said they did not receive training in these areas. Table 4.2 shows, for the supervisors who received training, their opinion of the adequacy of training in the three areas. Table 4.1: Supervisor8 Who Did Not Receive Training Type of training Number Percent Using performance standards 25,695 15 Providing assistance to poor performers 44,967 25 Initiating/proposing performance actions 51,391 29 Table 4.2: Supervisor@ Opinions on the Adequacy of Training Adequate Neither Inadequate Type of training Number Percent Number Percent Number Percent Usting performance .-.. _. _~.-....--~standards ._.. 110,123 74 20,648 14 18,354 12 --. Providing ..-assistance -_ .-..-.__....”to poor performers ..-.-_~- 104,158 78 19,272 14 10,554 8 InitiatinQ/ proDosina Derformance actions 96,817 76 18,354 14 11,471 9 Generally, supervisors who either did not receive training or received training they considered to be inadequate in these three areas were more negative in their perceptions of (1) their authority to propose a perform- ance action, (2) management support, and (3) their willingness to pro- pose actions against poor performers. For example, about 34 percent of the supervisors who did not receive training in the area of proposing performance actions believed they lacked authority to take action against unacceptable performers.2 Only about 17 percent of the supervi- sors who received training believed they lacked authority. 2The sampling error for this estimate is greater than 6 percent. See appendix VI. Page 40 GAO/GGD-91-7 Performance Management Chaptur 4 Management Commitment b Needed to Enhance Efforb to Deal With Poor Performers instances of reluctance to propose performance actions against unac- ceptable performers because they perceived a lack of authority and inadequate management support. Identifying and dealing with poor performers involves subjective judg- ment. Thus, there are no simple solutions to the questions of how to better motivate managers and supervisors to identify poor performers or how to make it easier to deal with those performers. Some of the obstacles cited by supervisors, such as a perceived lack of authority, may lend themselves to such prescriptive solutions as additional training. But solutions to other obstacles, such as the perceived lack of management support or the amount of time involved, are more complex. In fact, given human nature, well-intended prescriptive actions could provide unintended results. For example, supervisors who find it diffi- cult to confront poor performers because they do not want to engage in adversarial situations might simply stop doing it altogether if additional reporting requirements were imposed that they considered to be unreasonable. Individual motivation is a key factor regarding the extent to which supervisors deal with poor performers. Thus, it is important for man- agement to create an environment within which supervisors are encouraged and motivated to identify poor performers and supported when they try to deal with them. Toward that end, options 3 and 4, which we discussed in chapter 2, offer the potential for eliminating some of the frustration supervisors experience in dealing with their min- imally successful performers. These proposals could provide some incen- tive for such employees to improve their performance and make it easier forsupervisors to deal with them. Other actions that could provide a more positive environment for super- visors include more definitive and visible involvement and oversight on the part of agency top management and increased training and technical assistance. By making the commitment and becoming more actively involved, management could more clearly demonstrate its interest in helping supervisors identify and deal with their poor performers. Over- sight would enable management to identify areas in which further training and support are needed and help to ensure that the process requirements do not result in supervisors spending more time than war- ranted in dealing with poor performers. Finally, through training and technical assistance, management could help supervisors more clearly understand their authority and assist them in developing better stan- dards for assessing performance. Page 43 GAO/GGD-91-7 Performance Management Chapter 4 Management Commitment Is Needed to Enhance Efforta to Deal With Poor Performers defined (clear and specific) and that supervisors need methods of col- lecting and reviewing employee performance that are not labor inten- sive. He pointed out that if the above conditions are no&et, it is most difficult for the supervisor to do his or her job. l A non-numeric standard required an employee to implement Equal Employment Opportunity/Affirmative Action Program objectives at the first line supervisory level. According to the supervisor, the problem with this standard is that it is too vague. Also, the actions required are beyond the authority of the ratee to implement. Through oversight, management could take an active interest in identi- fying such standards and working with supervisors and employees to improve them. Oversight is particularly important in situations where agencies use generic standards that are written to cover all employees who occupy a particular type of job. Because of the general applicability of such standards, it would be difficult for supervisors to initiate action to improve them without management interest and cooperation. Oversight would also be useful in managing the time period over which supervisors deal with their poor performers. As we pointed out previ- ously, the process for dealing with poor performers usually begins with informal counseling and closer supervision and proceeds to a formal written notice of an opportunity to improve if the informal process is not successful in improving performance. Neither the formal nor the informal processes were monitored by management at the locations we visited to assess the progress employees were making or the problems supervisors were having in dealing with them. The lengthy time of some of the poor performance cases identified in this report strongly suggests that increased management oversight is needed and that progress in dealing with poor performers should be periodically reviewed. The process for identifying and dealing with poor performers has been Conclusions used to improve many employees’ performance or to take various other performance actions. However, a significant number of supervisors either have not dealt with their poor performers or said they would not deal with them in the future. In addition, approximately 61 percent of the supervisors reported that they experienced difficulties in imple- menting the process for dealing with poor performers. Performance standards can be difficult to use in identifying poor performance; the process can also be very time consuming. Further, supervisors reported Page 42 GAO/GGD-91-7 Performance Management Chapter 4 Management Commitment L Needed to l3nhani!o EffOrta to D?a.l with Poor Penformer . giving more attention to working with supervisors and employees to develop and maintain good performance standards; and . ensuring that managers and supervisors are aware that they have the authority to take performance actions against unacceptable performers. Additionally, we recommend that OPMrequire agencies to establish methods and procedures for overseeing how well poor performers are being identified and dealt with,,Oversight is important in assessing supervisors’ performance in dealing with these individuals, identifying other problems in need of management attention, and helping ensure that the time involved in dealing with individual poor performance cases does not become unreasonable. In making this assessment, agen- cies should periodically check the progress being made in dealing with persons who have been identified as poor performers. Agency Comments and Our Evaluation Comments From OPM OPMagreed with our recommendations and cited a number of actions that, if fully implemented, should improve the ability of supervisors to identify and deal with their poor performers. OPMsaid that it plans to issue memoranda to heads of departments and agencies and directors of personnel urging them and their top managers to make effective performance management a high priority. It also is revising its guide for managers, entitled “Taking Actions on the Problem Employee,” and said that the revised guide will place additional emphasis on dealing with poor performers. OPMalso agreed that it should assist agencies in providing the training and technical assistance necessary to effectively deal with poor per- formers and expressed the view that the preparation of a specialized course on dealing with poor performers could lead to a marked improve- ment in supervisors’ skills and confidence in this area. OPMalso concurred in the need for effective monitoring of performance management activities within agencies. It said that agencies are required to periodically evaluate the effectiveness of their performance manage- ment systems and that identifying and dealing with poor performers is Page 46 GAO/GGD91-7 Performance Management chapter 4 Management Caxnmitment h Needed to J3nluanceEfforta to Deal With Poor Performers How much time should be spent in dealing with poor performers? This is a difficult question and can vary given individual circumstances. As noted earlier, supervisors usually begin to deal with a poor performer by providing informal counseling and supervision before proceeding to a formal opportunity period. Because employees who are given a formal opportunity period must be told how much time they will be given to improve their performance, management should use this time frame as a basis for assessing progress. However, making such assessments in informal situations will be more difficult. At a minimum, management should stress to supervisors the importance of using milestones even when they are working informally with their employees. A final issue we identified was the supervisors’ concern about the appeal and arbitration processes. Many perceived these processes to be cumbersome, complex, lengthy, and slanted too much in favor of the poor performer. Because an assessment of these processes was outside the scope of our review, we cannot judge the validity of the supervisors’ perceptions on this issue. We can only note that this issue is particularly complex in that it involves balancing employee and management rights. The ultimate answer to the supervisors’ perceptions may rest with a determination by the executive branch and Congress as to whether the present balance is proper. Within the area of performance management, OPMis responsible for pro- Recommendationsto viding personnel leadership to federal agencies. Accordingly, we recom- the Director of OPM mend that the Director of OPMstress the need for a greater commitment and more visible involvement from top managers in identifying and dealing with poor performers. ,$‘op management commitment and involvement will show managers and supervisors that the identification and resolution of performance problems are important, are matters of great concern, and that upper management supports the effort. We’also recommend that OPMassist federal agencies in demonstrating this commitment by ensuring that all agencies provide periodic training and the necessary technical assistance to adequately prepare managers and supervisors to identify and deal with poor performers. The training should focus on l making clear, to managers and supervisors at all levels, the policies and objectives regarding the identification and resolution of employee per- formance problems; Page 44 GAO/GGDSl-7 Performance Management chapter4 Mnmtgemant Chdtment Is Needed k, Enhance Efforts to Deal With Poor Performerr cause as will promote the efficiency of the service,” a requirement not specified in chapter 43. Finally, the agency must be prepared to justify its choice of sanction in light of existing mitigating and aggravating fac- tors. No similar requirement exists for chapter 43 actions. The National Treasury Employees Union also did not believe legislation was needed to deal with minimally successful performers. It suggested requiring agencies to establish written performance standards at all levels of performance as opposed to only the fully successful level. While such action could possibly help supervisors identify poor per- formers, we do not believe it would directly address supervisors’ problems in dealing with poor performers after they have been identi- fied. Because this latter problem is the focus of our attention in presenting the third and fourth options, we continue to believe they are appropriate. Page 47 GAO/GGBI)l-7 Performance Management Chapter 4 Management Commitment Ie Needed to Enhance Eff~rta ta Deal With Poor Performers an aspect of performance management. It also identified a number of activities it has undertaken to assist agencies in carrying out this task. A more detailed discussion of OPM'Scomments is included in appendix I. Comments From Federal Three federal employee unions expressed general agreement with the Employee Unions information contained in our report. (See apps. VII, VIII, and IX.) How- ever, the National Federation of Federal Employees expressed concern with our position on minimally successful performers, pointing out that because their performance was not unacceptable, such employees were performing successfully. The union also questioned whether denial of annual pay increases could serve as incentive for performance improve- ment, and said that minimally successful performers could, in fact, be fired. We considered poor performers to be employees who were performing below the fully successful level. Looking at this issue from the tax- payer’s point of view, we believe that the fully successful level is the minimum level of performance federal agencies should accept, and therefore disagree with the National Federation of Federal Employees on the acceptability of minimally successful performance. Our view is supported by the options available for dealing with minimally successful performance under SESand PMRSsystems. With regard to denial of within-grade increases, our concern was that employees who have been in grade for longer periods of time do not receive within-grade increases every year. It was our view that denying within-grade increases would provide little incentive if the employee was not scheduled to receive one. Perhaps the union is right in saying that denial of annual pay increases would be no more effective. But at least under that situation, minimally successful performers would not receive pay that was comparable to those who performed at the fully successful level or above. Finally, while it is true that chapter 75 of title 5 of the U.S. Code, which generally governs disciplinary actions, can be used in some circum- stances to reduce in grade or remove poorly performing employees, the requirements are far more administratively burdensome on agencies than performance-based actions governed by chapter 43 of the US. Code were designed to be. An agency using chapter 76 procedures must meet a higher standard of proof than that required under chapter 43. The agency must also prove that the action is being taken “for such Page 46 GAO/GGD-91-7 Pedormance Management Appendix I C!4mment9 From the Offke of Personnel Management KEY ISSUES RELATED TO RECOMMENDATIONS AND CONCLUSIONS I. Conclusions and Recommendation on Dealing with Poor Performers Clarification of Anencv Flexibilitv in Establishing the Minimallv Successful The report includes a discussion of possible options for dealing with poor performers which culminates in four recommendations for Congressional consideration. (Reference pages 41-W When discuaeing poor performers, the report refers to two categories of employees: minimally successful and unacceptable performers. See comment 1. Moet of the iesuee raised in the report revolve around perceived difficulties in dealing with the minimally eucceseful performer. In thie respect, there appear6 to be a misconception underlying some of the report’s statements and conclusions regarding whether agencies must establish a performance rating level of minimally eucceeaful in their performance appraisal plane. Prior to addressing the report’s recommendations on dealing with poor performers, we want to clarify that OPM regulations, at 6 CFR Part 430.204(h), require only that agencies muet establish k summary performance rating levels, including a level between fully successful and unacceptable for employeee in the General Schedule and Prevailing Rate categories. The regulations aleo give agenciee certain discretion in establiehing the number of rating levels for individual critical performance elements. Although agencies must provide for @ least three rating levels for each critical performance 1 Page 49 GAO/GGD91-7 Performance Management Appendix I CommentsFrom the Office of PersonnelManagement Note: GAO comments supplementing those in the report text appear at the UNITED STATES end of this appendix. OFFICE OF PERSONNEL MANAGEMENT WA~“,NOTOI. D.C. 10.1 B Richard L. Fogel A&&ant Comptroller General General Accounting Office Waehington, DC. 20648 Dear Mr. Fogel: In reeponseto your recent request, I am forwarding the Office of Personnel Management’s (OPM) comme$a on the draft Geyeral Accounting Office (G+O) report entitled, performance anaeement; How Well 18 the Government DeahnP:mth Poor Performers. I would like to commend GAO for undertaking a broad-gauged study on this difficult but important topic. The information contained in the report will be very helpful to agencies and OPM in understanding better the challenging issues and problems faced by managers and eupervieore in dealing with poor performers in the Federal workforce. Also useful wan the discussion of the draft report between Bernard Ungar and Directors of Personnel held recently at OPM. I am confident that the information, analysis and recommendations in the report will stimulate and support a variety of meaeures to strengthen the ability of supervisors to identify and deal with poor performers effectively. Our ~pecillc commenta on the draft report are grouped into two mGor categories: comments relating to the report’s key conclusions and recommendations end comments dealing with technical aspects of the report’s underlying assumptions, characterizations end findinga. These comments are contained in the enclosed material. Questions concerning the comments may be directed to Allan D. Heuerxnan of OPM’s Personnel Syrtsme and Oversight Group on (202) 606-2910. We appreciate the opportunity to comment on the report. Sincerely, LLtiL!zSNb Director Encloeure Page 48 GAO/GGD91-7Performance Management Comments From the OffIce of Personnel Management unacceptable. Further, such activity would not require action by Congress as implied by the later section entitled, “Matters for Congressional Consideration.” OPM would also like to point out that agencies currently have the ability to revise standards to reflect performance requirements more accurately. Further, OPM believes that establishing and revieing fair and defensible performance standards must be done on a case-by-case basis and relate to performance expectations and mission requirements and not as a response to a general admonition to upgrade performance requirements. Accordingly, OPM suggests that option one be reconsidered. See comment 3 The second option recommends eliminating the minimally successful rating category either through law or regulatory action. As pointed out above, agencies, through the design of their performance appraisal systema, already may choose to do this with respect to rating levels on individual performance elements. Only OPM’s regulation requiring agency performance appraisal systems to provide for a minimally successful summary performance rating level would have to be changed to support this option. Although OPM will consider this recommendation, we would prefer a closely related legislative approach, eimilar to that recently enacted for Performance Management and Recognition System (PMRS) employees (P.L. 101. 103, the Performance Management and Recognition System Reauthorization Act), which preserves agency flexibility to rate employeee at three to five different levels on individual performance elemente, and which allows performance-based reduction in grade and removal actions to be initiated solely on the basis of determining that an employee% performance falls below the fully succeseful level for one or more 3 Page 61 GAO/GGD91-7 Performance Management Appendix I Canmentrr From the Off&e of Personnel Menagement element, including a description at the filly successful level (5 CFR 430.204(e), there Is no requirement to establish a minimally successful performance level for individual elements. Many agencies have used this flexibility and have established performance appraisal systems which provide for only three ratings levels on individual performance elements. These syetems do not utilize the minimally successful rating level on individual performance elements. The only level of performance below fully successful in these systems is unacceptable. Thus, for agencies who have three rating level eysteme (for individual performance elements), personnel actione to reduce in grade or remove employees may be initiated when an employee’s performance is determined to be at the unacceptable level on one or more critical performance elements, irrespective of the employee’s summary performance rating. (Note that 5 CFR 0432.104 provides for talcing reduction-in- grade and removal actions based on unacceptable performance in one or more critical elements) See comment 2. With this flexibility in mind, we find that the first option, which would encourage agencies to change employees’ individual performance standards so that current minimally successful performance would be considered unacceptable performance, is not effectively precluded or limited by OPM regulation as the discussion implies. That is, by eliminating the standard of minimally successful, an agency gains the ability to determine that performance below the fully successful level is 2 Page 50 GAO/GGD91-7 Perf’onnance Management . Appendix I Cammentn Fxnm the Office of Personnel Management Option 4 The fourth option addresses the possibility of amending the Civil Service Reform Act of 1978 to allow for reduction-in-grade and removal actions based on minimally successful performance by a GS or WG employee. While this approach directly addresses the minimally aucceaeful issue, we would suggest an alternative approach. Specifically, combine the flexibility agencies now have to eliminate the minimally successful level for individual elements (see discussion on pages 1 and 2) with the authority for agencies to initiate performance based actions on the basis of determining that the employee ie not performing at the fully successful level. Thie approach would give agencies the same flexibility with respect to non-PMRS employees that they now have with respect to PMRS employees, i.e., to eliminate, or retain, the minimally successful performance level and to initiate an action based on a determination that performance is below the fully successful level, without needing to assign either an unacceptable or minimally successful rating. It should be pointed out that this approach should also contain the provision, aa is the case with the PMRS system, that agencies must provide employees performing below the fully successful level with an opportunity to correct their deficiencies through a performance improvement plan (PIP) prior to a determination being made on whether or not to propose a removal or reduction in grade action. This approach should help deal with chronically poor performers as intended by option 4, preserve agency flexibility in deeigning performance appraisal systems, and avoid the difficulty of “selling” a legislative change which would permit management action against an employee who received a rating that indicates “successful” performance (as in “minimally successful”). In further support of this 6 Page 53 GAO/GGD91-7 PerPormance Management Appendix I Comments Prom the Omce of Personnel Management critical elements. This differs from the non-PMRS law and regulations which require a determination that the employee’s performance is unacceptable in one or more critical elements. Thie difference in the kind of determination required - that the performance ie pnacceotable for non-PMRS employees vs, that the performance is below fullv eucceseful for PMRS employees - can be a substantive one where agenciee have five level systems, including minimally satisfactory. OPM supported the revision to the PMRS system and has recently issued regulations, at 6 CFR Parts 430 and 432, implementing the new provisions of law relating to actions based on below fully successful performance. This approach preserve6 flexibility and permits agenciee to take timely action in dealing with poor performers. It also closely relates to GAO’s option number four discussed on page 43 of the report, (See also our comments below on option 4.) Ontion 3 The third option recommends the enactment of legislation which would more closely link pay to performance for GS and WG employees. OPM does, of course, eupport this type of approach and, aa the report recognizes, the Administration has a Pay Reform bill currently pending before Congress which includes provisions for denying baeic pay increases to employees who are performing below the fully euccessful level. However, as the diecussion in the report points out, meaBure8 of thie nature will not completely address the types of problems identified by supervisor-e and managers in dealing with poor performers, especially those employee8 who are chronically marginal performers. 4 Page 62 GAO/GGD-91-7 Performance Management Appendix I Commenta From the Office of Pen9onnel Management high priority and to apply the flexibilitiee in the current government-wide syetem in ways that recognize and reward good performers and take appropriate actions to deal with poor performers. Further, we are revising and updating a very ueeful and popular guide for managers (‘Taking Action on the Problem Employee”) dealing with employees with conduct and performance deficiencies. The revised guide will place additional emphasis on dealing with poor performers. Finally, OPM has recently released a draft FPM Chapter 432, “Performance Based Reduction in Grade and Removal Actions,” to agencies and major unions for review and comment. This FPM material includes extensive guidance on identifying and dealing with poor performers. Second Recommendation We aleo support the second GAO recommendation, which calls for OPM to assist agenciee in providing periodic management and supervisory training, aa well as the needed technical assistance in dealing effectively with poor performers. While we believe that the focus of training and technical assistance should encompass the full range of performance management skills and techniques, we can see the possibility that a specialized course on dealing with poor performers could lead to a marked improvement in supervisors’ skills and confidence in this particular area. OPM’s regulations at 5 C.F.R 430.208 require that “agencies must provide appropriate training and information to supervisors and employees on the appraisal process.” The appraisal process, ae described at C.F.R. 430.204 (i) and (i), includes 7 Page 66 GAO/GGD91-7 Performance Management API-- 1 Cwunen~PmmtheOfficeof Personnel Management approach, agencies, to OPM’s knowledge, have not experienced any diiYiculties in implementing the revised job retention provisions of the new PMRS legislation. II. Recommendations to the Director of OPM Chapters 3 and 4 provide a discussion of supervisory and managerial perceptions of the procedural difIiculties involved in dealing with the poor performers as well as the management initiatives which could help ease some of these dif3culties. Included in the discussions are recommendations for OPM action, The following comments address these recommendations. &ret Recommendation The first recommendation is that OPM be more active in stressing the need for top agency management to become more visibly involved and to make a greater commitment to identifying and dealing with poor performers effectively. We agree with this recommendation, particularly in light of the results of the study which indicate that, in many cases, supervisors do not feel that they have management’s support or sui?icient authority to initiate or take performance-based actions. OPM has several current initiatives underway which should contribute to this objective, including commissioning a major study by the National Academy of Sciences to examine successful performance appraisal systems and techniques in the private sector which can be applied in the Federal system to manage and measure employee performance more effectively, In addition, OPM plans to issue memoranda to heads of departments and agencies and Directors of Personnel urging them and their top managers to make effective performance management a 6 Page 64 GAO/GGD-91-7 Performance Management . Appendix 1 Comments From the Offlce of Personnel Management Third Recommendation The final recommendation is that OPM require agencies to eetablish methods and procedures for overseeing how well poor performers are being identified and dealt with. We concur with the need for effective monitoring of performance management activities within agencies. Again, we believe that OPM’s regulations at 6 C.F.R. 430.208 establish such a requirement, i.e., agencies “must establish methods and procedures to evaluate periodically the effectiveness of the system(s) and to implement improvements as needed.” In our view, this requirement necessarily encompasses the evaluation of how well agencies identify and deal with poor performers, including whether there is timely identification of and feedback to employees with performance problems, and whether agency action is tak.en as appropriate to deal with poor performers. Agency systems developed in accordance with statutory and regulatory requirements are reviewed by OPM to ensure that each such system meets all requirements. In addition, OPM has issued FPM guidance providing information and possible methods for evaluating all aspects of performance appraisal systems. Further, OPM has several activities underway aimed at enhancing the evaluation and oversight of Federal personnel management programs, including efforts to assist agencies in improving their internal oversight programs. Among thase activities is a recent project coordinated by CPM’s Agency Compliance and Evaluation (ACE) office which involved a review of agency performance management systems. ACE has produced a draft report entitled. “OPM 9 Page 57 GAO/GGD-91-7 Performance Management Appendix I Comments From the Office of Personnel Management supervieory assistance to improve employees’ substandard performance, and taking appropriate personnel actions when employees’ performance continues to be unacceptable. It is our expectation that agencies understand the importance of complying with OPM’s regulatory requirements, but OPM may need to increase its efforts to encourage and assist agencies in seeing that this type of training is provided to their managers and supervisors. OPM’s nationwide training program includes several courses on dealing with problem employees, e.g., “Performance Management: Coaching and Appraisal,” Performance Management: Individual Performance Planning,” and “Supervising Performance, Conduct and Leave.” Alao, OPM currently is engaged in an extensive review and “overhaul” of its training curriculum in key personnel management areas, including that of performance management. As part of this review, OPM will consider the benefits of adding a course which has as its primary objective the improvement of super-v&ore practical skills in dealing with poor performers. Finally, OPM will consider recommending to agencies that training and information on the appraisal process be offered on a recurring basis to help ensure that supervisors and managers keep their skills up to date and recognize the importance of proper implementation of the performance appraisal system. 8 Page 56 GAO/GGD91-7 Performance Management Appendix I Comment43 From the Offlce of Pernonnel Management The following are GAO'Scomments on OPM'Sletter dated July 23, 1990. 1. We revised the wording on pp. 2, 13, 22, 24, and 28 to more clearly GAO Comments show that OPMrequires agencies to establish the minimally successful rating category for summary performance ratings and that agencies have discretion in the number of rating levels they establish for indi- vidual job tasks. It should also be noted that although agencies do have the flexibility to use only 3 rating levels for individual job tasks, 4 of the 10 agencies we visited used 5. As pointed out by OPM,agencies may initiate personnel actions when an employee’s performance is determined to be at the unacceptable level for one or more critical elements, regardless of the employee’s summary rating. This does not, however, eliminate problems supervisors experi- ence in dealing with employees who are viewed as performing at the minimally successful versus unacceptable level. 2. We have clarified our discussion of option one so as not to imply that changing performance standards would be precluded by OPMregulation. As we pointed out in the report, option one does not appear to offer a complete solution, because even if the standards were made more strin- gent, there would still be employees who would perform at the mini- mally successful level. Like OPM,we view other options as offering better solutions to the problem. 3. As previously noted, we have revised the report to address agencies’ flexibility with respect to establishing rating levels for individual job tasks. This, however, does not eliminate the problem. While the mini- mally successful rating category could be eliminated, the type of per- formance it identifies would remain. Page 69 GAO/GGD-91-7 Performance Management Appendix I Conunenta FroIll the Offlce of PeraoamelManagement Government-Wide Review: Performance Management Systems at Federal Installations.” This study involved many aspects of performance management,such aa training for super&ore and employees, as well aa documentation and eupport for performance-baaed actions taken by agencies. The draft report ie currently undergoing internal OPM clearance. ACE also has undertaken several projects involving the evaluation of agencies’ abilities to monitor the effectiveness of their performance management syetema. In February 1989, ACE published a “Survey of Agency Personnel Management Evaluation Programs,” which contains an assessment of all areas of coverage by agencies’ internal Personnel Management Evaluation (PME) programs, including performance management. As a result of that survey and its recommendations, OPM has established three interagency task forces whoee purpose ie to enhance agency oversight capabilitiee by (1) developing guidelines for agencies’ internal PME programs; (2) developing and exchanging ADP re8ourcea for PME; and (3) developing PME guidance for small agencies. Finally, OPM will continue to produce the “Annual Report to the President and the Congress on the Effectiveness of the Performance Management and Recognition System.” The fir& three reports provide a broad range of information concerning the effectiveness of the system with regard to quality of performance standards and elements, perception of equity, inflation of ratings, training, and agency efforta to improve effectiveness. A fourth report is now in preparation. 10 Page 58 GAO/GGD-@l-7Performance Management Appendix III FederalAgenciesRepresentedin Our Questionnaire Department of the Air Force Department of the Army Department of Commerce Department of Defense Department of Education Department of Energy Department of Health and Human Services Department of Interior Department of Justice Department of the Navy Department of the Treasury Department of Transportation Department of Veterans Affairs Federal Communications Commission Federal Home Loan Bank Board Federal Trade Commission General Services Administration Interstate Commerce Commission National Aeronautics and Space Administration National Labor Relations Board National Mediation Board National Science Foundation National Archives and Records Administration Office of Personnel Management Securities and Exchange Commission Small Business Administration Smithsonian Institution United States Information Agency Page 61 GAO/GGD-91-7 Performance Management Appendix II Imations Visited During Our Review Federal agencies Location Forest Service Regional Office Denver, CO Regional Office San Francisco, CA Ogden, UT San Antonio Air Logistics Center Kellv Air Force Base San Antonio. TX Environmental Protection Agency Headquarters Washington, DC Regional Office Atlanta, GA General Services Administration National Capital Region Washington, DC Regional Office Atlanta, GA Department of Housing and Urban Development Headquarters Washington, DC Regional Office Denver, CO Department of Labor Regional Office Atlanta, GA Regional Office Denver, CO National Aeronautics and Space Administration Goddard Space Fli ht Center Greenbelt, MD Langley Research 8 enter Hampton, VA Federal Aviation Administration Headquarters Washington, DC Southern Region Atlanta, GA Internal Revenue Service District Office Atlanta, GA Service Center Atlanta, GA Department of Veterans Affairs Medical Center Richmond, VA Medical Center Augusta, GA State governments Commonwealth of Virginia Richmond, VA State of Georgia Atlanta, GA State of Colorado Denver, CO Local governments City of Virginia Beach Virginia Beach, VA City of Atlanta Atlanta, GA Citv of Denver Denver, CO Private corporations Blue Cross and Blue Shield of Virginia Richmond, VA Philip Morris U.S.A. Richmond, VA Humana, Inc. Virginia Beach, VA Sovran Financial Corporation Norfolk, VA E.I. du Pont de Nemours and Company, Inc. Richmond, VA Norfolk Southern Corporation Atlanta, GA Weyerhaeuser Paper Company Plymouth, NC Martin Marietta Corporation Denver, CO $asa;rsoebuck, and Company Denver, CO Denver, CO Delta Air Lines, Inc. Atlanta, GA Coca-Cola USA Atlanta, GA Page 60 GAO/GGD-91-7 Perfo-ce Management Amwdix N Qlleetlonnatrebed ln our Review 1. - II. eERFORMAnCE..STAnDARDS 1. How long hsvs VU hsld the specific job 6. Do the Porformsnc. plsn. of the l mploysss sssignmsnt thst YOU currsntly held? thst YOU dirrctlv contsin (CHECK ONE.1 ,.I rtsndsrds thst l rs GW@&C& Ci.s., stsndsrdr usins numb.r., porcont...., ste.), WI wSPseific lob ..signm.nt” is e (is*., rt.nd.rds using dsfinsd by your e m.n.g.m.nt/ d..triPtiv. words rather than number. or wp~rvioar~ duti.8 bnd r..pan.ibiliti.s. p.rc.ntag.s). or . w of numeric and non-num.ric stsndsrds? (CHECK ONE.) 1. t-1 Loss thsn 6 months ‘I,, 2. C-1 6 months to loss th.n 1 YSS~ 1. C-1 Numeric only 3. t-1 1 Vsar to 1.8. thsn 2 y..r. 2. [,I Non-numwic only 6. t-1 2 yssrs to lsss than 4 yearn 3. C-1 Combinstion of both tvp.. 5. I-1 4 ysmrs to 10~s thsn 6 Y..ra 6. 1-l 6 y..r. or more 2. How msny sm~lo~sss do YOU e e. thst is. do YOU e IF YOU DO m USE prspsrs partormanc* sppr.is.18 for? NUMERIC STANDARDS (i..., NUMBERS, (NUNBCR OF EMPLOYEES) ,141 PERCENTAGES, ETC.) TO RATE EMPLOYEES YOU DIRECTLY S. Since you b.cams l supsr~i~~r. hsv. YOU SUPERVISE _I_, (SKIP TO QUESTION 10.: rscsivsd trsinino concsrnine th. psrformmcs spprai8sl procssst (CHECK ONE.) IF YOU USE NUMERIC 11OP STANDARDS TO RATE 1. c-1 Y.8 EMPLOYEES YOU DIRECTLY SUPERVISE 3 (CONTINUE WITH 2. t-1 No QUESTION 7. J 4. Sines you bscsm. s supsrvi.or. hsvs YOU rscsivsd trsining on how to dssl with 7. In g.n.r.1, how l d.qu.ta or insd.qu.t. srs poor psrformwml (CHECK ONE.) ,,I, ths e p.rform.nc. ot.ndsrds th.t you u.s to msssur. psrformsncs. specifically 1. t-1 Yss (CONTINUE WITH QUESTION 5.1 in terms of idsntifying sm.loyss. who srs psrforming st 1.~8 th.n th. “fully 2. t-1 No (SKIP TO QUESTION 6.1 succ*..ful” l.v.11 (CHECK ONE.) 4”) 5. Did this trsinin. . ..cific.lly c0v.r th. 1. t-1 Mar. th.n .d.qu.t. followin. s8.sct.T (CWECK ALL THAT APPLY.) au-UI 2. C-1 Adsqust. 1. C-1 How to u.. psrformsncs .t.nd.rds 3. C-1 Neither .d.qust. nor insd8su.t. 2. t, I How to ..sist .m.loy.ss in improvins their psrformsncs 4. c-1 1nsd.qu.t. S. I-1 Ths procs.8 for proposing/ 5. C-1 Very inad.quat. initistin. psrformsncs action8 4. t-1 Coun..lling wn~lov~m 5. t-1 Th. .pP..ls/.rbitr.tion proes.8 Page 33 GAO/GGD-91-7 Performance Management Appendix IV QuestionrAre Used in Our Review U.S. GENERAL ACCOUNTINO OFFICE Washington, D.C. 20548 Y OF m The U.S. Oanaral Accounting Office (GAO). an Because this quastionnaira is being sent to PVO”CY of Congrsss, is conducting e survey to ruparvisors in swwal fadoral agencies. we obtain your opinions and information on YOUI- era providing definitions to describe the axpariences as a managw/mupervi+or of GM, threa levels of performance that we are using GS, and FWS employees. Specifically. we are in this questionnairel interested in issues related to performance man*q*mant functions. Fullv w - This is the expected level The questions deal primarily with supervisory of performance. responsibilities in identifying poor psrform- UI EPA uses the term WS, that is. amplowas who did not meet the “satisfactory”. “fully successful” level of their performance standards during fiscal war 1988. Questions apa also included which ask about the actions raquirad whan dealing with these poor Per- B - This is the level of performance formars. between “fully succersfuln and ‘unscceptable’. Your .q.ncy may Please respond to each of the following qua- use other tarms such as marginal, tions as they relate to the employees that mar9inally successful, minimally YOU m and for whom YOU at-0 l ccwtabla. minimally satisfactory, rosponsiblo for w or Partially successful. m. Your responses will be treated m. They will be combined with others and raportad only in summary form. !&wcePtrhl, - This is the lrvol at which The qurmtiannaire is numbered only to mid us pwformenco doas not meet in our follow-up efforts, since it might be astsblishmd “fully successful” necessary for u* to contact you to clarify a parformanca standards in one Particular r.spons.. Such contacts will not or more critic.31 *laments. affect tha confidwtiality of YOUP reoponre. WI EPA and NASA US. tha tat-m “unsatisfactory”. The quastionnsirm should take only about 20 to JO minutes to complete. If YOU have .ny questions. please call Mr. Jim Bishop on (FTS) 827-6621 or (804) 441-6621. Pleasa return the completed questionnaire in the anclosad pro-addressed ~nvelopa within 10 days of receipt. In the avant thm envelope is misplaced, thm raturn address is1 U.S. GENERAL ACCOUNTING OFFICE Norfolk Regional Office Mr. Jim Eishop 5705 Thurston Avenue Virginia Beach, VA 23455 Thank YOU for your help. Page 62 GAO/GGDSl-7 Performance Management . 0 Appendix IV Qnestionnaire used in onr Review 11 What typo(s) of problems, if any, am 13. If YOU had training on using pwformance you having using m performanea stmndards. how l doqusfm or inadaquatm stendardrt (CHECK ALL THAT APPLY.) "o-11' was that twining? (CHECK ONE.) 1w1 1. t-1 I AM NOT HAVIND PROBLEMS WITH 1. t-1 DID NOT HAVE ANY TRAININQ ON v PERFORMANCE STANDARDS USINQ PERFORMANCE STANDARDS 2. t-1 Thm standards er. uncl.sr 2. t-1 War. than l dwwto 3. I-1 Tha standards arm too easy to meet 3. t-1 Adoquata 4. t-1 The standards are too difficult 4. t-1 Neithw adsquote nor inadequate to moat 5. t-1 Inadoquats 5. I-1 It is difficult to measure the l mplov**'s performance against 6. I-1 Vary inmdaquato tha non-numwic strndards 6. [,I 0th.r (Spacify) _ 12. Please provide an wmplm of el pwformanco standard that YOU ra l having l problem with and l xplsin why it is l problm. Non-numwic performance standard* I u-11, Why it is . problmf Page 66 GAO/GGD@l-7 Perfommnce Management Appendix IV Qnestlonmdre Used in Our Review 8. What tyrm(m) of w-oblomm, if any, at-9 you having wing w o*rformanca l tandwda? (CHECK ALL THAT APPLY.) 11*-#*1 1. t-1 I AM NOT NAVINQ PROBLEMS WITH NUMeRIC PERFORMANCE STANDARDS 2. I-1 The standards are unclar IF YOU USE NON-NUMERIC STANDARDS TO RATE J. t-1 Tha mtmndardr arm too easy to maat EMPLOYEES YOU DIRECTLY SUPERVISE m (CONTINUE WITH 4. t-1 Tha standards at-9 too difficult QUESTION 10.) to meat 5. C-1 It is difficult to measure tha bmployoe'o porformanco against 10. In 44n4r41. how mdaquata or inad4quat4 ar4 the numw-ic standards tha w performance standards that you use to measure perform*nca. 4p4cifically 6. C-1 Othw (Spacifyl in terms of id4ntifying employees who ara performing 4t 1444 than thm “fully successful” 14v4lt (CHECK ONE.) 410 9. Pleas* Provida an example of a nulnaric 1. I-1 More than adequate pwformmnco standard that you are having 4 probl4m with and wphin why it is l 2. t-1 Adaquata problem. 3. t-1 Nmithor adequate nor inadequate Nunoric pwformanca standard, (u-2“ 4. C-1 Inadaqumt. 5. t-1 Vwy inadwuata Page 94 GAO/GGD-91-7 Performance Management Appendix IQ Questlonnah Used in Our Review 1 20. For tho4. .mploy..r th4t you 4tt.mpted to 44sist (so. Pu.otion 18), wh.t results wara 4shi4v.d from th. proc.s4? (CHECK ALL THAT APPLY AND FOR THOSE CHECKED, ENTER THE NUMBER OF EMPLOYEES FOR WHOM THE RESULT WAS ACHIEVED.) WI THE TOTAL NUMBER OF EMPLOYEES SHOULD EQUAL THE NUMBER ENTERED IN PUESTION 18. Employee's p.rformanca improved to "fully suec4saful" .............................. / / (NUMBER) Employ44 is curr4ntly performing et the 14s~ th4n “fully succ4ssful" 10~41 and i4 still b4ino arsisted . . . . . . . . . . ..I............ / / (NUMBER) P.rform.nc. 4ction 4gainst employ44 form411y proPo4.d in writing (include chrng. to low4r prod4 or remov41) ............... / / (NUMBER) Employ44 4gr44d to r4duction in grade . ..I....... / / (NUMBER) Employ44 4gr44d to r444sicinment ................. / / (NUMBER) Employ44 r44ign4d .............................. / / (NUMBER) Employ44 r4tir4d ................................ / / (NUMBER) 0th.r (P1.44. sp.cify) -.. 1 / (NUMBER) .. / / (NUMBER) 21. Con4id.r th. .mount of tin. th.t you op.nd w44kly. on th. werag.. providing 4s4i.t.nc. to .n 4mploy44 whos. p.rform.nc. is 14~4 than "fully succ4rsful". P14444 provid4 4" 4otin4t4 of the tin4 you sp4nd p4r w4ek. (NUMBER OF HOURS PER WEEK PER EMPLOYEE) 1”.911 22. Durinp fi4c41 y44r 1961, w.r. th.r. .ny ~44.5 wh.r. emPloy.es that YOU ~stmervipLd w4r4 performing 4t th4 144s th4n "fully sufcsssful" level .nd YOU did net .tt.mpt to 4s4ist th4m to imrwov4 their parformance? (CHECK ONE.) 1-1 1. C-1 Y.4 (INDICATE THE NUMBER OF CASES BELOW AND CONTINUE WITH QUESTION 23.1 (NUMBER) ‘**I 2. t-1 No (SKIP TO QUESTION 25.1 Page 67 GAO/GGD91-7 Performance Management Questionnaire bed in Our Bevievlr III. -TIN0 w THAN "FULLY SUCCESSFVL n NOTEI For tha ~ur~os.s of this section W. 16. For how long have thpsa employees beon l ra darling with l rnplov~~s that YOU parforming at the “minimally successful" baliwa l ra pat-forming at tha lass than level? (ENTER NUMBER(S) IN APPROPRIATE "fully succ*risful" l.v*l ,t anv UlQg CATEGORIES. TOTAL SHOULD EQUAL ENTRY IN QUESTION 15.) TIME AT MINIMALLY NUMBER OF S!!XFSSF'JL LEVEL EMPLOYEES 14. During fiscal year 19ll8. how mmy 0 TO LESS THAN 6 MONTHS . ..a / /,,,,,, l mployoas that YOU &.~&,&runervisrd war. pwforming at the less than “fully 6 TO LESS THAN 12 MONTHS . . . / I,~-.,, succassful" lwel? (ENTER NUMBER. IF NONE, ENTER ZERO "On.) 12 TO LESS THAN 24 MONTHS . . / /,r.,,, (NUMBER) '*'-*' 24 TO LESS THAN 36 MONTHS . * /411-s,, 36 MONTHS OR MORE a......... //lS+,*, IF ZERO "0" IN GIUESTION 14, SKIP TO QUESTION 25. 17. During fiscal year 1988, did you attempt I to assist tha employees you identified in question number 14 to improve their performance? (CHECK ONE.) tub IS. Wet-a l ny of tha l mployoas that YOU idantifiad in question 14 as performing 1. t-1 'Yes (CONTINUE WITH QUESTION 18.) at the 1.8s than “fully suecarrful" leval pwformins at the "m" 2. t-1 No (SKIP TO QUESTION 22.1 lsvml? (CHECK ONE.) ,.,I 1. t-1 Yas CENTER NUMBER OF EMPLOYEES AND 18. How many of these employees did YOU CONTINUE WITH PUESTION 16.1 attempt to assist? (ENTER NUMBER.) (NUMBER) ,‘,-"I 2. t-1 No (SKIP TO OUESTION 17.1 19. What did YOU do to sssist these omploye~s? (CHECK ALL THAT APPLY.) II.-*II 1. t-1 Counseled and worked with the *mployea informally 2. I-1 Increased the smount of supervision for th* amplowe 3. C-1 Provided on-the-job training 4. C-1 Arranged for the employem to receive training 5. C-1 Placed the employee on a performance impravem*nt Plan 6. C-1 Referred the employee to a caunsrling service 7. t-1 Other action (Specify) Page 60 GAO/GGD@l-7 Performance Management Appendix TV Questionnahe Used in Our Review 25. Tsking into consideration your a(lenCY’S 27. Minimslly successful l mployats cannot procsdurss for dealing with poor be removed or dmmotad. Doao this factor parformerm. if you have a less than “fully increpsa, decraasa or hsvm no l ffmct on suce*9sful’ employma in the future, how the likalihood of daaling with . likely or unlikely would it be that YOU "minimally successful” l mployas? would attempt to desl with that employee (CHECK ONE.) (“0 using your l 9mcy’s fwocaosl (CHECK ONE.) l,Dl 1. t-1 Greatly incraose likalihaod 1. t-1 Very 2. t-1 Increase lik-lihood 2. c, I Likely (SKIP TO ‘2. 27.) 3. C-1 Neither incrpsso nor dacrearp likelihood 3. t-1 Not sure 4. t-1 Dacrosso likelihood 4. t-1 Unlikely (CONTINUE WITH 5. C-1 Greatly dacreasa likelihood 5. t-1 Vary QUESTION 26,) 26. Why would it be unlikely that YOU would 28. If you hsd training on how to assist uss your spmncy’s process in tha future? employees improve their performance. (CHECK ALL THAT APPLY.) how adequate or inadequate was that training? (CHECK ONE.) 8-B 1. [,I Ths procsss takss too long 1. t-1 Did not hsva any training on 2. tT,l The process uses UP too much of how to sssist employees ths supervisor’s tine improve their performance ------------------------------------ 3. CT,1 The process upes up too much of 2. t-1 Mot-a than adoquata other l mployaar’ timo 3. t-1 Adoquat. 4. [=I The process disrupts working relationships among staff 4. t-1 Naithar sdsquptm nor inadequate S. Ix,1 Employens make sn effort to 5. t-1 Inmdaquatm improve without the process 6. t-1 Vw-v inadesumte 6. [=,I Tha parformanca standards are not sufficient to support or damonstrata poor pmrformmnce 7. [=,I] I hsvr not beon given sufficient guidmncm on using the procmps 8. t-1 Different units within my anancy #“I have provided conflictinp guidsncm on using thm process 9. I=,1 I mm aware of problems encountered by other supervisors who usad tha process 10. 1 I Th*ra is s lack of m~n~nsnmt #r7,, support 11. I-1 Other (Spacifyl a...,, J Page 69 GAO/GGDOl-7 Perf’ormance Management Appendix N Quef~tlonnaire Used in Our Review 2s. For those cmsau wharm you decided & to assist the l mployao, which of the tallowing rmsons antwad into your docirion? (CHECK ALL THAT APPLY.) 1. t-1 Providing l ssistanca tak*s too long 1.1 2. Cm1 Providing assistance usas too much of the supwvisor's timo S. C-1 Providing assistance usw too much of other employems’ time 4.I 4. C-1 Providing assistance disrupts working relationships among steff 8.3 5. [=,I Employees make an effort to improve without assistarea 6. c-1 The performance standards are not sufficient to support 1111 or damonstrata lass than "fully successful" performance 7. [,I I have not been given guidance on providing arsirtmncm (111 to improve employe4s1 perfarmanca 8. t-1 Employee(r) was "ninim~lly successful” and I did not 1111 beliwo assistwwe would improve performance 9. C-1 Emglowo's parformanca was “unacceptable” and I didn't 11.1 baliwe l ssirtanca would improve performance 10. C I I believed that the l mployae(s) did not want msrirtanca eici6, 11. C-1 Assistance provided to employees in the past has bean 4L,.UI ineffective in improving performsnc* 12. t& Employae(s) demonstrated ability to perform adaquataly in tha Pmst so I felt assistance was not necessary 13. t I Other r..sonr (Specify) lIZi,, 24. For those casms whwo YOU decided ti to provide assistance, how did you dam1 with the employma( (CHECK ALL THAT APPLY.) 1. [=I Oava the l mployea easier work 2. lx,1 Raduced the l mployor'r workload 3. [ =I Let the situation work itself out 4. [=,;;I Reassigned tha l mployoa to another work area within your unit 5. CT,1 Reassignad the employee to anothw unit 6. [=,I Advised the employaa to rarign 7. c=,I Other (Specify) Page 69 GAO/GGDBl-7 Performance Management . Que!atlonnalre u9ed in our ReYiew 32. Plaa~a uam the fallowing chart to show tha number of pwformanca actions you hwo ProPomad/initiated in fiscal year 1988 against WY l mtaloyma that you m who ~a8 performing at the %nacceptablmw lwal, and tha finml ramulta of thorn* actiona. [FOR PART (A) ENTER THE NUMBER OF PRCPOSED/INITIATED CHANGES TO LOWER GRADE OR REMOVALS IN FISCAL YEAR 1988. IF NONE. ENTER ZERO “0”. FOR PART (8) ENTER THE FINAL RESULTS. w NOTF THAT FOR Fw THE ENTRY PART (A) THE TOTAL IN PART (&.I PROPOSED/INITIATED PROPOSED/INITIATED TO LOWER (NUMBER) (NUMBER) -z!zF FYzr 1. No action taken . . . . ..a. (NUMBER) lrl-)l) (NUMBER) (.LIuI 2. Employ@* chmnwd to lowor orada ............ Il-yI (NUMBER) (NUMBER) ,“.%I 3. Employa* was roammignod ............. (NUMBER) (NUMBER) T I”-,LI 4. Lmploy** km* romovod .,.............. (NUMBER) (NUMBER) Ilt-YI 117-1.1 1. IZmployoa ratirad ..a.**. ,,,", (NUMBER) (NUMBER) -YEiT- I 6. CmPlovmo roaignrd ..*..* (NUMBER) T 7. Action in program ..... (NUMBER) (NUMBER) 7zz-- T 6. Other action tmken [Plwme l p*cify) *. I(I.L1I (NUMBER) *. (,, u, (NUMBER) I Page 71 GAO/GGD-91-7 PerIormance Management Appendix IV Questionnaire Used in Our Review IV. PERFORMAWCE,L?$LLlIONS JO. In your opinion, to what extant, It at l 11, would the next highw lw*l of WI Pwfarmancm actions includrl msnsrsmont above you support actions l gminat l mployws who l re parforming at s. Change to a lower grad* the “unacceptable” lw*lt (CHECK ONE.) b. Removal 1.71 1. I-1 To a very groat extent Pwformanca actions do & includol 2. C-1 To l groat extant a. Denial of within-grade salary increases or the loss of one-half 3. t-1 To l modwrte extent of comparability PBY b. Performance improvement plans 4. t-1 To soma oxtent 5. C-1 To little or no extent 29. In your opinion, to what extent, if at all. does your position include 31. During fiscal year 1988. did YOU either mufficiant authority to propose/initiate propos*/initiat* any performance pwformanco actions against *mployew action against any employem that YOU who l t-a performing at tha “unaccaptmbla” &ractlv w who was pot-forming l*v*l? (CHECK ONE.) 8UI at the “unaccaptablo” lava11 (CHECK ONE.) IUI 1. t-1 To a vary or-at extent (w do not w any emplovew that 2. t-1 To l ormat axtent 1. t-1 Yor (CONTINUE TO QUESTION 32.1 J. r-1 To a moderata l xtant 2. r-1 No (SKIP TO ‘WESTION 36.1 4. t-1 To some extent 5. C-1 To little or no extant Y Page 70 GAO/GGDOl-7 Performance Management . Appendix Iv Que8tio~ used in our Review 36. In the future, if you had l n “unacceptable” 38. If YOU had training related to proposing/ pwformw. how likely or unlikely would it initiating porformancc, actions. how be thmt you would woposm/initiata a adewato or incldoquota was this training? pwformnnso action to daal with that (CHECK ONE.) 1111 l mployooT (CHECK ONE.) 1111 1. t-1 Did not have any training 1. t-1 Very lik*ly related to Proposing/initiating (SKIP TO porformwwo actions 2. t-1 Likely QUESTION 38.1 ----_-------------__________________ 2. I-1 More then adequate J. I-1 Not sure I- 3. t-1 Adequeta 4. t-1 Unlikely 4. t-1 Neither adequate nor inadequate (CONTINUE WITH 5. t-1 Very unlikely I- QUESTION 37.1 5. t-1 Inadequate 6. t-1 Very inadequate 37. Why would it ba unlikely that YOU would propoo*/initimto l porformanc* action 39. If YOU hwo any comments regarding any in tha futuro? (CHECK ALL THAT APPLY.) Previous quartion or wnoral comments concwning pwormanco management The process takes too long functions. plaara use the space provided~below. If naccssary, attach Thm process takes too much of additional shoots. ,111 the wpwviaor’o time Do not wont to go through tho l ppwl/*rbitrmtion process Thwo im l leek of manwmmmt support Tho porfornanco standards are not wfficiant to support pwformmca actions I am aware of Problems l ncountorad by othw supervisors who proporod/initiatmd a performance 8ction I hwa not bow given suffieiant puidancm on taking a pwformanca action Diffwent units within my l amcy hwo provided conflicting guidance on taking l pwform8nco action Other romson(s) (Specify) / THANK YOU FOR YOUR HELP. -1 PLEASE RETURN YOUR COMPLETED SURVEY IN THE ENCLOSED PRE-ADDRESSED ENVELOPE. J Page 73 GAO/GGD91-7 Performance Management - Appendix IV x Questionnaire Used in Our Review 53. For the propomed/initimted portornunca $4. Of all the gorfornunoo wtionm thet you l ctiona indicated in quwtion J2 progomod/initiatod in fiscal war 19611. (“FINAL RESULTS”1 whw. NO ACTION how many wara not suPPorted by your w., takmn, why w.s this th. cwa? next lwml wpwvisor? (CHECK ALL THAT APPLY.) (ENTER NUMBER AND CONTINUE WITH QUESTION 35. IF NONE, ENTER ZERO “0’ A. CHANOE AND SKIP TO QUESTION 36.1 (NUMBER) “‘-“I 1. tTiil There wws l lack of documentation to support tha action 35. In those cases whara YOU did not rocaiw support from your next lwal supwvisor, what wer. the reasons givm for not supporting your proposal1 S. [=I Othw (Specify) (CHECK ALL THAT APPLY.) 1. I;;;;] No wason provided 2. [=,;I Did not have tha tin* wailablo 3. I-1 Did not want to go through thm ,,.I appeal/arbitration prowess 1. C-1 Thera was a lack of documontgtion 41) to support tha action 4. trir,l Your documentation was not sufficiwt to support the action 5. C-1 Fait that higher lwals of agency 1UI management would not support the J. t-1 Other (Spwify) 1111 action 6. [=I Tha employma was due to ratire 7. [=,I Other alternatives such as roarsi9nmsnt etc. wrg not considered 8. [=,I Other reason(s) Page 7 2 GAO/GGD91-7 Performance Management ! ‘“u: Appendix VI QuestionnaireObjective,Scope, and Methodology The purpose of our questionnaire was to obtain information on how supervisors in the federal government identify and deal with employees whose performance is less than fully successful. Using a mail question- naire, we asked supervisors about performance standards, the process for dealing with poor performers, and performance actions during fiscal year 1988. We defined a supervisor as one who directly supervised at least one employee and was responsible for preparing an employee’s summary appraisal. To validate the questionnaire, we pretested it by administering it to a Instrument Validation number of supervisors employed by two federal agencies in our final and Verification sample. We selected a headquarters site for one of the agencies and a regional office for the other to test the instrument’s universality. During the pretests, we observed respondents while they were completing the questionnaires and asked them to point out any aspect of the instrument that was unclear or otherwise problematic. We also asked for their com- ments and opinions on the questionnaire and discussed their answers with them to see if they understood the questions. Many of their sugges- tions were incorporated into the final questionnaire, which we then mailed to the full sample of supervisors, We reviewed and edited each returned questionnaire for completeness and consistency and entered the responses into a computer database. The resulting database was then verified for its accuracy, and logic checks were performed on the data. We developed two samples for our questionnaire. The governmentwide Sampling Methodology sample was selected from a universe of supervisors in the Office of Per- sonnel Management’s (OPM)database of civilian supervisors and was designed to provide us with a general view of how federal supervisors were identifying and dealing with poor performers. The other sample was drawn from a universe of supervisors at the 20 locations we visited to expand upon information we received at those sites. Governmentwide Sample OPMprovided us with a data tape listing names and office mailing addresses of 1,53 1 government supervisors of General Schedule (GS), General Management (GM), and Wage Grade (WG) employees. The 1,53 1 supervisors were randomly selected from the Central Personnel Data File (CPDF),an OPMdatabase which contained a total of 252,366 civilian supervisors in September 1988. A GAOstatistician reviewed OPM'S Page 76 GAO/GGD-91-7 Performance Management Appendix V . Time SupervisorsSpentAssisting Poor Petiormers Number of employees who Voluntarily Had actions were still proposed poor Number of months Improved agsi:~:: against them performers 0.0 - 2.9 14 13 5 1 3.0 _ 5.9 27 12 12 11 6.0 - 8.9 22 13 10 14 9.0 - 11.9 21 21 4 33 12.0-14.9 15 8 8 8 15.0 - 17.9 9 3 1 19 18.0 - 20.9 1 3 1 9 21 .o - 23.9 3 2 1 4 24.0 - 26.9 1 0 1 2 27.0 - 29.9 1 0 2 'is 30.0 - 32.9 0 1 1 1 33.0 - 35.9 0 0 0 0 si.0-38.9 0 0 0 0 39.0 - 41.9 0 0 0 0 42.0 - 44.9 2 0 0 0 Total 116 76 46 102 Source: GAO analysis of data from the 340 DCls completed during the review. Page 74 GAO/GGBBl-7 Performance Management . Appendh VI Questionnaire Objective, Scope, and Methodology important to note that in instances where only a subset of the popula- tion responded to a question, sampling errors could be greater than + 5 percent because of the decrease in sample size. Supervisors in the governmentwide sample completed questionnaires Survey Response between April and June 1989. From the 650 supervisors who were sent Rates questionnaires, we obtained a 72-percent usable return rate (percentage usable of total mailed) and a 82.3-percent adjusted usable return rate (usable returns as a percentage of total mailed less ineligibles and unde- liverables). The final respondent group consisted of 396 supervisors. Supervisors in the 20 locations sample completed questionnaires between February and May 1989. Of the 1,635 supervisors who were mailed questionnaires, 1,188 returned usable ones to us, a rate of 77.4 percent. The adjusted usable return rate was 84.9 percent. Table VI. 1 summarizes the questionnaire returns. Table VI.1: Questionnaire Return Governmentwide 20 locations Tvoe of return Number Percent Number Percent Usable returns 396 72.0 1188 77.4 Ineligibles: _.---.--- Not supervisors ______-___ 35 6.4 113 7.4 Left aaencv/location 26 4.7 9 0.6 Questionnaires returned as undeliverable 8 1.5 14 0.9 Questionnaires delivered but not returned ~--~ 85 15.5 211 13.7 Total 550 100 1635 100 Note: Figures may not add to 100 percent because of rounding. The results from our governmentwide sample are limited to the 396 Universe Estimates usable returns and represent an estimated 181,704 supervisors, or 72 From the Sample percent, of the total universe of 252,366. These 181,704 supervisors Results directly supervised about 1.6 million employees in fiscal year 1988. The 1,188 usable returns of the 20 locations sample represent an estimated 9,551 supervisors, or 76 percent, of the total universe of 12,511 who e supervised an estimated 78,000 employees. Page 77 GAOJGGD-W-7 PerPormance Management , Appendix VI Questionnaire Objective, Scope, and Methodology random selection process, and we tested the accuracy of the information provided in the file through a telephone survey to a random sample of 60 supervisors. We did not, however, verify the CPDF for completeness. From the list of 1,531 names, we selected a random sample of 550. We determined that this size would be sufficient to permit generalizations to the universe of 262,366 supervisors at a confidence level of 95 percent and a sampling error, or precision, of + 5 percent. We drew 550 random cases that did not overlap with our 20 locations sample. Our sample included cases from 39 of the total 104 federal agencies. 20 Locations Sample From each of the 20 locations that we visited, we requested a list of names and office mailing addresses for supervisors of GS, GM, and WG employees as of December 1988. A total of 12,511 supervisors were identified through the agencies’ computerized personnel data files. We did not independently verify the accuracy and completeness of the list- ings or computer tapes that the agencies provided us. We used standard statistical techniques to select a stratified random sample of supervisors. Each stratum represented 1 of the 20 locations. The total sample consisted of 1,535 supervisors. We selected specific sample sizes to ensure that the sampling error for each stratum would not be greater than 10 percent, and the sampling error for the entire sample (i.e., the 20 strata combined) would not be greater than 5 per- cent, at the 95-percent confidence level. Explanation of Sampling Because we selected a sample instead of surveying all the members of Error the universe, the results we obtained are subject to some degree of uncertainty, or sampling error. The sampling error represents the expected difference between our sample results, or estimates, and the “true” results that would have been obtained from surveying the entire universe of supervisors. Sampling errors are always stated at a specific confidence level,’ which is the degree of assurance that can be placed in estimates derived from the sample. In designing our samples we speci- fied that the sampling error should be no greater than + 5 percent at a confidence level of 95 percent. In other words, we expect the chances are 19 in 20, or the probability is 95 percent, that our sample estimates will be within + 5 percentage points of the “true” universe values. It is ‘Using Statistical Sampling, Methodology Transfer Paper 6, GAO, Program Evaluation and Method- ology Division (Washington, DC.: Apr. 1986). Page 76 GAO/GGD-91-7 Performance Management Appendix Vl Qnentlo~A~ Objective, Scope, end Methodology Table Vl.2: Universe Estimator and Sampling Errors (95Percent Confidence Variable Estimate Lower limit Upper limit Interval) Number of employees directly supervised by the supervisors represented in our survey (in millions) 1.6 1.4 1.7 Proportion of employees who performed below the fullv successful level 5.7% 3.7% 7.7% Proportion of employees performing below the fully successful level who were identified as minimallv successful durina fiscal vear 1988 02.6% 70.0% 86.4% Proportion of supervisors who indicated that they assisted poor performers 98.2% 93.7% 99.8% Average time spent per week with each poor performer 4.9 hrs. 4.0 hrs. 5.8 hrs. Supervisors’ willingness to deal with poor performers in the future Likely 76.1% 71.9% 80.3% Not sure 13.0% 9.7% 16.3% Unlikelv 10.9% 7.8% 14.0% Proportion of supervisors having poor performers that they did not assist 8.3% 3.1% 13.5% Supervisors’ opinions concerning the quality of non-numeric performance standards Adequate 74.6% 69.8% 79.4% Neither adequate nor inadequate 14.0% 10.2% 17.8% lnadeauate 11.4% 7.9% 14.9% Supervisors’ opinions concerning the quality of numeric performance standards Adequate 64.0% 57.9% 70.1% Neither adeauate nor inadeauate 16.9% 12.1% 21.7% Inadequate 19.1% 14.1% 24.1% Proportion of supervisors indicating that they are unlikely to propose a performance action in the future 10.2% 6.3% 14.1% Say the standards are not sufficient to suooort the actions 32.5% 17.8% 47.2% Believe that they did not have management’s support 57.5% 42.0% 73.0% Say the process takes too much of their time 65.0% 50.1% 79.9% Do not want to go through the appeals arbitration orocess 35.0% 20.1% 49.9% Proportion of supervisors who perceive a lack of authority to take performance actions 21.7% 17.6% 25.8% Proportion of supervisors with poor performers who perceive a lack of authority 32.1% 23.3% 40.9% Proportion of supervisors without poor performers who perceive a lack of authority 17.7% 13.3% 22.1% (continued) Page 79 GAO/GGD-91-7 Performance Management s Appendix VI Questionnaire Objective, Scope, and Methodology To show the size of our sampling error, upper and lower limits for all estimates that appear in this report are shown in table VI.2. Page 78 GAO/GGD91-7 Perl’onnance Management Appendix VI Questionnaire Objective, Scope, and Methodology Variable Eetlmrte Lower limit Upper limit Proportion of supervisors who did not receive trainin for proposing performance actrons TEA w o e reved they lacked authority to take action aaainst poor performers 34.2% 25.4% 43.0% Proportion of supervisors who received trainin for proposing performamons -mA w o e reved thev lacked authoritv 16.7% 12.3% 21.1% Proportion of supervisors who are unlikely to take performance actions who say they have not been given sufficient guidance on using the process 17.5% 8.0% 31 .O% Proportion of employees indicated by supervisors as performing at the minimally successful level during fiscal year 1988 4.7% 2.5% 6.9% Number of months employees were minimally successful performers 0 to less than 6 months 46.0% 35.8% 56.2% 6 to less than 12 months 26.7% 17.7% 35.7% 12 or more months 27.3% 18.3% 36.3% Proportion of supervisors indicating that the fact that minimally successful employees cannot be removed or demoted decreases the likelihood that they will deal with these employees 19.0% 15.1% 22.9% Proportion of supervisors expressing that they are having any problems in dealing with poor performers 50.5% 45.6% 55.4% Page 81 GAO/GGD@l-7 Performance Management Appendix VI Questionnaire Objective, Scope, and Methodology Variable Ertimate Lower limit Upper limit Proportion of supervisors who believe they lack authority who say they are unlikely to rnttrate actron against poor performers. 21.4% 12.6% 30.2% Prooortion of suoervisors who believe thev have authorit &ho say they are unlikely to iiiRmx& 7.2% 4.3% 10.1% Proportion of supervisors who believe they lack adequate management support for taking action against unacceptable performers 16.1% 12.5% 19.7% Proportion of supervisors with poor performers who expressed a lack of management support 23.9% 15.9% 31.9% Proportion of supervisors without poor performers who expressed a lack of management support 13.1% 9.2% 17.0% Proportion of supervisors who believe they do not have management support and who say they are unlikely to propose actions 30.6% 19.1% 42.1% Proportion of supervisors who believe they have management support and who say they are unlrkely to propose actions 6.4% 3.7% 9.1% Proportion of supervisors who did not receive trainina for usina performance standards 14.7% 11.1% 18.3% Proportion of supervisors who did not receive training for providing assistance to poor performers 25.1% 20.8% 29.4% Proportion of supervisors who did not receive training for initiating/ proposing performance actions 28.9% 24.4% 33.4% Supervisors’ opinions regarding the quality of training for using performance standards Adeauate 73.8% 69.0% 78.6% Neither adequate nor inadequate 13.8% 10.0% 17.6% Inadequate 12.3% 8.7% 15.9% Supervisors’ opinions regarding the quality of trarning for providing assistance to poor performers - Adequate 77.7% 72.9% 82.5% Neither adequate nor inadeauate 14.4% 10.4% 18.4% Inadequate 7.9% 4.8% 11 .O% Supervisors’ opinions regarding the quality of training for initiating/proposing actions Adeauate 76.4% 71.4% 81.4% Neither adequate nor inadequate 14.5% 10.4% 18.6% Inadequate 9.1% 5.7% 12.5% (continued) Page 80 GAO/GGD-91-7 Performance Management Comments From the National Federation of Federal Employees GAO-SMK-001327 -2- June 12, 1990 performance is Ned unacceptable. We fail to see any problem with the employee continuing as a Federal employee. As you point out, managers currently have statutory authority to withhold within grade increases in such cases. Seepp. 46-47 In your report, however, you state that within grade increase denial does not serve as a sufficient incentive to promote better performance. Yet, paradoxically, the report suggests that denying the annual pay increase to minimally successful workers could serve as an incentive for performance improvement. Unfortunately, the report does not contain any evidence that such a change will be a more effective incentive than the within grade increase denial which YOU deem to be ineffective. We believe it is inappropriate to make public policy suggestions without factual foundation. See pp. 4647 Of more importance, however, is the report’s implicit adoption of the concept that minimally successful workers are “poor performers” that cannot be fired. This is blatantly untrue. As stated above, we do not believe it is proper to categorize a minimally successful employee as a poor performer. Moreover, while a minimal successful rating cannot lead to removal under Chapter 43 of Title 5 of the U.S. Code. such a rating, over a period of time, can lead to a removal action under Chapter 75 of Title 5 of the U.S. Code for the efficiency of the service. To suggest otherwise is a disservice to the readers of the report. We do not believe that the criticisms above should detract from what is otherwise a fine analysis of the performance management system. We look forward to seeing a copy of the final report. Sincerely, Steven Kreisberg Executive Direct0 Page 83 GAO/GGD91-7 Performance Management Appendix VII comments From the National Federationof * Federal Employees National Federation of Federal Employees James M. Peirce l President Ronald W. Kipke l Secretary lkeasurer In reply refer to: GAO-SMK-001327 June 12, 1990 Mr. Richard Page1 Assistant Comptroller General General Accounting Office Washington, DC 20540 Dear Mr. Pogel: We appreciate your submitting to us for comment your Draft Report concerning Performance Management. We ara in general agreement with much of the content of the report but we have a few brief comments that we wish to share with you. Pirat, GAO’s focus on the lack of managerial commitment to the performance management process is clearly the most salient finding of the report. Use of generic performance standards. vague standards, subjective standards, non- measurable standards, etc. have created a situation that undermines effective performance management. Accordingly, supervisors lack confidence in the process since they realize that once the process is exposed in a third party forum they will find it difficult to support their actions. As the report points out the failing of the performance management system in this regard is primarily a management failure to properly implement the system. We believe a focus on training of individual managers and supervisors is necessary in order to effectuate the purposes of the Civil Service Reform Act. If the current program is not working it is because management has failed to make it work. Seepp. 46-47. We are, frankly, confused by your suggestions to resolve the minimally successful category “problem.” First, we do not concur with your analysis that such a “problem” indeed exists. After all. a minimally successful rating indicates that an employee is performing successfully since his/her 1016 16th Street, NW; Washington, D.C. 20036; Phone: (202) 662-4400; FAX (202) 862-4432 tme lwbMl vlw Prorldmrte fteglon5, Mhur R. GuanMo, Jr., SanlaFe,NM Regbn 1, Qoofgh Kwhun, Hunl!iybn, NV Re9lon6, Jameo0. Md)onald,Bhhop,CA Fbg!m 2, R&art E. Simmofm,Newark,NJ Reglw 7, DauglasD. Wrbhl, Mboula, MT I*gkn 3. Row Mfi!y While,PsnamaUly, FL Reglw 8, BrendaK. Otslad.Muorhead,MN lU#l01I4, Vlvkn W. Grw, Shnvcport,LA Regh 9, ShellaVSIUOO,Munck, IN Page 82 GAO/GGD91-7 Perfomance Maheeement . Appendix IX CommentsFrom the National Treasury EmployeesUnion The National ‘keasury EmployeesUnion June 26, 1990 Mr. Richard L. Fogel Assistant Comptroller General General Accounting Office General Government Division Washington, D.C. 20548 Dear Mr. Fogel; Thank you for the opportunity to comment on the draft GAD Report concerning: "Performance Management: How Well Is The Government Dealing With Poor Performers?" The issue of agency performance appraisal systems is of primary importance to NTEU and the employees we represent. Clearly, the system envisioned by the Civil Service Reform Act of 1978, (CSRA) I has not fully materialized. This observation appears to be supported by the preliminary findings of the GAO study. As an example, as stated in your draft report, "about half of the supervisors said they had experienced difficulty in implementing a process for dealing with poor performers. Among other things, supervisors cited the significant amount of time that can be involved, a perceived lack of management support, difficulitee in using the performance standards, a perceived lack of authority to propose performance actions, and a reluctance to go through appeal or arbitration." These perceptions on the part of eupervieors must be addressed in any attempt to formulate proposed solutions to the problems identified by this report. Supervisors must be motivated, and supported in their efforts to deal with poor performers. NTEU fully supports GAO'0 recommendation that federal agencies demonstrate their commitment to their supervisors and their employees by, "ensuring that all managers and supervisors receive periodic training and the necessary technical assistance to adequately prepare them to deal with Poor performers..." Although 5 CFR 430.204 (i) requires that “each appraisal system shall provide for assisting employees in improving performance rated at a level below fully successful...," the means of providing such assistance is typically left to the imagination of individual supervisors who are ill prepared to deal with this responsibility. Specific training must be provided to supervisora and to employees. 1730 K Street, N.W. * Suite 1101 l Washington, D.C. 201X6 . (202) 785411 @-u&- Page 85 GAO/GGDSl-‘7 Performance Management . Appendix VIII Comments From the American Federation of * ’ GovernmentEmployees,LWLCIO AMERlOlN FEDEIWTICN OFGWERNMENr EMPLOYEES, AFL-CIO John N. Sturdlvant Allen H. Kaplan Joan C. Welsh Nallonal PresldenI NatIonal SeoreIary.Trearurer Dlrmnor, Women’* 0epulmw-d 6 /GAO June 25, 1990 Mr. Richard L. Fogel Assistant Comptroller General U.S. General Accounting Office 441 G Street, N.W., Room 3a5ac Washington, DC 20548 Dear Mr. Fogel: This is in response to your letter dated June 5, 1990, addressed to President John Sturdivant regarding GAO's report on performance management. We find your report to be most interesting and the conclusions reached sup ort beliefs of AFGE that the federal government does effectively dea P with poor performance. We also agree that the procedures are unduly complicated and not understood, not on1 by supervisors, but also workers who need to improve their wor E to increase their performance ratings. We have found that many supervisors confuse disciplinary matters with performances~d&cause of this, they have difficulty supporting their positions they be challenged in an appeals process. Disciplinary matters such as tardiness, laziness, and absenteeism should be properly dealt with as dieci linary matters. However, supervisors have not been trained in t K is process either. As a result, their views of such workers show up in the performance evaluation which should narrowly be applied to how the employees perform their duties outlined in their position descriptions and their performance standards. Ae you know, the burden of proof in taking discipline is different under both disciplinary actions and performance actions. When supervisors do not clearly distinguish between the two and save disciplinary complaints against a poor performer for a performance appraisal, they then have difficulty sustaining any action should it be challenged. If supervisors could be trained to clearly delineate between the two and separate personnel actions into disciplinary and erfonnance matters instead of mixing the two, the system would work %etter. Thank you for the opportunity to submit our comments. Sincerely, 02%-J J n%!6%~&ar& Director Y ield Services Depi&tment JWM/dao 80 FStreet,N.WI44shington,DC 20001 (202) 737-8700 *-,a Page 84 GAO/GGD-91-7 Performance Management Gary B. Brumback Senior Policy Advisor to the Assistant Secretary for Personnel, U.S. Department of Health and Human Services, Washington, DC Alan K. Campbell Executive Vice President and Vice Chairman of the Board, ARA Ser- vices, Inc., Philadelphia, PA (Vice Chairman of the Center for Excellence in Government) DennisM. Devaney Board Member, National Labor Relations Board, Washington, DC (Former MSPB member and FLEA General Counsel) John F. Hillins Vice President, Corporate Compensation, Honeywell, Inc., Minneapolis, MN (Vice President, American Compensation Association, and Chairman of ACA Productivity Task Force) Patricia W. Ingraharn Associate Professor of Public Administration and Director of the Mas- ters Program in Public Administration, State University of New York, Binghamton, NY (Served as Project Director, Task Force on Recruitment and Retention, National Commission on the Public Service) Paul D. Mahoney Director, Office of Management Analysis, Merit Systems Protection Board, Washington, DC JamesL. Perry Professor of Public and Environmental Affairs, School of Public and Environmental Affairs, Indiana University, Bloomington, IN (Past Presi- dent, Section on Personnel Administration and Labor Relations, Amer- ican Society for Public Administration; Past Chairman, Public Sector Division, Academy of Management) Timothy M. Dirks Chief, Labor Relations Division of the Office of Employee and Labor Relations, Office of Personnel Management, Washington, DC Page 07 GAO/GGD91-7 Performance Management AppendixM Comments From the National Trensury Employees Union Federal government agencies must also create a work environment which supports and encourages positive aupervisory efforts to deal with poor performers. Supervisors must believe that this is an important part of their work responsibility. Without such a positive motivation, the many competing priorities placed upon individual supervisors will effectively negate any efforts to address this problem. NTEU does not agree with GAO's recommendations for possible congressional action. Simply "eliminating the minimally successful rating category," or "changing performance standards so that what is now considered to be minimally successful becomes unacceptable" is an overly simplistic solution to a complex problem. NTEU would recommend revising the regulations to require agencies to establish written performance standards at all levels of performance as opposed to only the "fully successful" level. (See 5 pCrFoRVTtj:.204(e)). It is our experience that in agencies that only written performance standards at the "fully successful" level, supervisors and employees have great difficulty clearly articulating performance expectations above or below the "fully successful" level. Written performance standards at each level of performance would add a degree of objectivity to the performance evaluation process. It would also address the issue raised by supervisors regarding the difficulty in using performance standards. Again, thank you for the opportunity to comment on this draft report. If I can answer any additional questions, or provide greater details concerning NTEU's comments or recommendations, please feel free to contact Mark L. Gray, Assistant Director of Negotiations, at 202-785-4411. Sincerely, Robert-M. Tobias National President Page 86 GAO/GGD91-7 Performance Management ,_/.A- 88, ../ I,.,. 88, 8, s/s Ord~~riug Irrformation The first fivch copies of each GAO report are free. Additional copies iii-t* .$2 each. Orckrs should be sent to t.he following address, acco~n- parlid by a ch~k or money order made out to the Supclriuteudeut of Ih~wn~twt.s, wl~n utmmary. Orders for 100 or more copies to be ruailtvl 1.0a single address are discmuted 28 p6m:ent. 1l.S. (;rutval Accounting Office I’.(). Hex 6015 Gaithm-shurg, MD 20877 Ortkrs may also be placed by calling (202) 2756241. Norman A. Stubenhofer, Assistant Director, Federal Human Resource Ywwd Government Management Issues B Washington, Valerie A. Miller, Technical Advisor James G. Bishop, Issue Area Manager YE’Iolk Regional Bradley D. Simpson, Project Manager iii#kG Robert V. Arcenia, Evaluator Sandra F. Bell, Evaluator Clyde E. James, Evaluator Rudolf0 G. Payan, m ;YItiver Regional ;yltrcr
Performance Management: How Well Is the Government Dealing With Poor Performers?
Published by the Government Accountability Office on 1990-10-02.
Below is a raw (and likely hideous) rendition of the original report. (PDF)