For other versions of this document, see http://wikileaks.org/wiki/CRS-RL32596 ------------------------------------------------------------------------------ Order Code RL32596 Sick Leave: Usage Rates and Leave Balances for Employees in Major Federal Retirement Systems Updated August 26, 2008 Curtis W. Copeland Specialist in American National Government Government and Finance Division Sick Leave: Usage Rates and Leave Balances for Employees in Major Federal Retirement Systems Summary Full-time federal civilian employees receive 13 days of sick leave each year, and may carry over an unlimited amount of unused sick leave from year to year. At the time of retirement, employees in the Civil Service Retirement System (CSRS) (those hired before 1984) receive credit in the computation of their civil service retirement annuities for any unused sick leave they have at the time of retirement. In contrast, employees in the Federal Employees Retirement System (FERS) (those hired after 1983 and others who transferred to FERS) generally receive no credit for unused sick leave at the time of retirement. As a result, some believe that FERS employees will use more of their sick leave as they approach retirement (resulting in productivity losses) than their CSRS counterparts. The conference report for the legislation creating FERS in 1986 urged the Office of Personnel Management (OPM) to examine this issue. FERS employees comprise an increasing percentage of the federal workforce, and by 2014 virtually all federal employees are expected to be in FERS. In 2004, CRS obtained data indicating that FERS retirees and FERS employees in that group who were eligible to retire or approaching eligibility used substantially more sick leave and had much lower sick leave balances than comparable CSRS retirees and employees. In 2006, OPM similarly concluded that FERS employees who were eligible to retire or were nearing eligibility used more sick leave than their CSRS counterparts, but concluded that to make sick leave creditable under FERS would cost several times more than the potential productivity savings. In April 2008, the Treasury Inspector General for Tax Administration concluded that the lack of compensation for unused sick leave was causing FERS employees in the Internal Revenue Service to use more sick leave than their CSRS counterparts. In March 2008, Representative Jim Moran introduced H.R. 5573, which would permit FERS employees who satisfied the age and service requirements for immediate or early retirement to receive a lump-sum payment equal to 15% of the hourly rate of their basic pay for each hour of unused sick leave balance over 500 hours, with the total payment capped at $10,000. Union and federal manager organizations supported the legislation, but OPM said greater sick leave use by FERS employees could be caused by factors unrelated to differences in FERS and CSRS policies. In July 2008, the House of Representatives passed the Family Smoking Prevention and Tobacco Control Act (H.R. 1108), which included a section that would permit FERS employees to receive service credit for unused sick leave. In 2004, all except five state governments provided some type of compensation to employees for unused sick leave at retirement, usually either as cash payments, service credit in the computation of annuities, or payment of health or life insurance premiums. The states generally limited these payments in some way (e.g., capping the number of hours of sick leave payable). This report will be updated when additional or more current data on sick leave use become available, or if other changes in federal sick leave policies are proposed. Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Federal Sick Leave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 CSRS and Sick Leave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 FERS and Sick Leave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Indications of Greater Sick Leave Use by FERS Employees . . . . . . . . . . . . . 8 FERS Employees Use More Sick Leave Than Their CSRS Counterparts . . . . . 9 Differences in Sick Leave Usage and Balances . . . . . . . . . . . . . . . . . . . . . 10 Sick Leave Policies in Other Organizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Unused Sick Leave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Sick Leave Policies in State Governments . . . . . . . . . . . . . . . . . . . . . . . . . 14 Cash Payment Plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Service Credit Plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Insurance Plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Policy Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 More Stringent Enforcement of Sick Leave Policies . . . . . . . . . . . . . . . . . . 17 Providing an Incentive to Conserve Sick Leave . . . . . . . . . . . . . . . . . . . . . 17 A Comparison of Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 OPM's 2006 Study of Sick Leave Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Analysis of the OPM Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Treasury Inspector General Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Legislative Developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Appendix. State Sick Leave Policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 List of Figures Figure 1. Number of FERS Employees Is Increasing . . . . . . . . . . . . . . . . . . . . . . 3 Figure 2. Number of FERS Retirement Eligibles Is Increasing . . . . . . . . . . . . . . 4 List of Tables Table 1. FERS Employees Used More Sick Leave than CSRS Employees as They Approached Retirement Eligibility . . . . . . . . . . . . . . . . . . . . . . . . . 11 Table 2. FERS Employees Had Lower Sick Leave Balances than CSRS Employees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Table 3. FERS Retirees Had Lower Sick Leave Balances than CSRS Retirees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Table 4. OPM Study Shows FERS Employees Generally Used More Sick Leave in 2005 - 2006 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Table 5. Sick Leave Policies in State Governments . . . . . . . . . . . . . . . . . . . . . . 26 Sick Leave: Usage Rates and Leave Balances for Employees in Major Federal Retirement Systems Introduction Pensions for nearly all federal civilian employees are provided through one of two retirement programs -- the Civil Service Retirement System (CSRS) or the Federal Employees Retirement System (FERS).1 Most federal civilian workers hired before January 1, 1984, are covered by CSRS, whereas those hired after that date are covered by FERS (as are employees who voluntarily switched from CSRS to FERS during "open seasons" held in 1987 and 1998). CSRS and FERS are similar in several respects. In both retirement systems, employees' annuities are a function of age, length of federal service, and final average pay.2 The normal retirement age (without reduction of annuity) under both CSRS and FERS is 55 for employees born before 1948 who have at least 30 years of service.3 Both retirement systems also allow normal retirement at age 60 for employees with 20 or more years of service, and 62 for employees with at least five years of service. Also, employees in both systems receive 13 days (104 hours) of sick leave each year, and may carry over an unlimited amount of unused sick leave from year to year. Therefore, an employee who carries over an average of nine days (72 hours) of sick leave each year during a 30-year career would accumulate 2,160 hours of unused sick leave by the time he or she is eligible to retire. One of the many ways in which CSRS and FERS differ is in how accumulated sick leave is accounted for at the time of retirement. In the CSRS system, retiring employees receive service credit in the computation of their annuities for any unused sick leave they had at the time of retirement. For example, a retiring CSRS employee with 30 years of service and one year (2,087 hours) of accumulated sick leave would 1 Although the federal government has many different retirement plans, CSRS and FERS are by far the largest. In this report, "federal civilian employees" refers to those in the CSRS or FERS plans. For a more thorough discussion of these two plans, see CRS Report 98-810 EPW, Federal Employees' Retirement System: Benefits and Financing, by Patrick Purcell. 2 FERS annuities are generally smaller than in CSRS because they are only one component of the system. A more detailed description of the two systems is provided later in this report. 3 The FERS normal retirement age increases for employees born in 1948 or later, and eventually reaches age 57 for employees born in 1970 or later. CRS-2 be considered to have 31 years of service for purposes of annuity computation.4 In sharp contrast, FERS retirees generally receive no service credit or other type of compensation for their accumulated sick leave.5 As a result, some observers suggest that FERS employees will be on sick leave more frequently than CSRS employees, particularly as they approach retirement age, thereby resulting in reduced productivity, increased personnel costs, or both. Reports of FERS employees actually carrying out this "use it or lose it" behavior periodically appear in press articles.6 To discourage this phenomenon from occurring, some observers have advocated changing FERS to provide some kind of value for unused sick leave (e.g., allowing full or partial service credit to be used in annuity calculations, or providing full or partial lump-sum cash payments for such leave).7 For example, in April 2006, a coalition of federal management groups (the Federal Managers Association, the Senior Executives Association, the Federal Aviation Administration Managers Association, the National Council of Social Security Management Associations, and the Professional Managers Association) called for a change in federal sick leave policy (e.g., allowing FERS employees to apply a portion of their unused leave toward post-retirement health care policies).8 Others, however, believe that the CSRS sick leave benefit is excessively generous, and that FERS should not replicate that generosity. Opponents of a policy change (and even some proponents) also point out that use of sick leave for unauthorized reasons is inappropriate, and that FERS employees knew (or should have known) about this policy when they agreed to the terms of employment.9 4 The congressionally mandated number of hours in a work year is 2,087. Only full years and months are counted in the annuity computation. For example, if a full-time employee worked 30 years and had 835 hours (four months and 24 days) of accumulated sick leave at the time of retirement, the employee would be credited with 30 years and four months of service; the 24 additional days would be dropped. 5 FERS employees who transferred from CSRS receive credit for sick leave balances at the time of the transfer or at the time of retirement, whichever is smaller. 6 See, for example, "Sick Leave and FERS: Leave Abuse? Justifiable Cheating? Flexible Ethics?," in FedSmith.com, available at [http://www.fedsmith.com/articles/ articles.showarticle.db.php?intArticleID=1143]; and "FERS and sick leave," Federal Times, April 24, 2006, p. 20. 7 For example, in testimony before the Senate Committee on Governmental Affairs on February 4, 2004, the President of the National League of Postmasters said, "We also need to address the issue of sick leave for FERS employees. Currently, they get no credit for unused sick leave at retirement. We need to change this rule so they could sell back sick leave or get credit at retirement." 8 Deborah Funk and Stephen Losey, "Sick Leave abuse: Managers prescribe cure for use-or- lose rules," Federal Times, April 17, 2006, p. 1. 9 For example, in the "Sick Leave and FERS" article mentioned above, 56% of the respondents to an unscientific survey said it was not ethical for a federal employee to use sick leave without having an authorized medical reason for using the leave. One respondent said "If you were hired under FERS, then you should go by the rules, and the CSRS rules are not your concern." CRS-3 This report provides data that CRS obtained on sick leave usage rates by many employees and recent retirees in the CSRS and FERS retirement systems in 2003 and 2004. It also discusses a study of federal sick leave use in 2005 and 2006 by the Office of Personnel Management (OPM), and provides information on unused sick leave policies in private sector and other public sector organizations (e.g., state governments). First, though, the report provides some background information on CSRS and FERS trends and federal sick leave policies in general. Background As Figure 1 below shows, because all new federal employees for more than 20 years have been put in FERS, the number of FERS employees in the federal workforce has increased substantially, whereas the number of CSRS employees has declined. By September 1996, more federal employees were in FERS (818,650) than in CSRS (809,098). As of September 2007, nearly 1.3 million federal employees were in FERS, compared with about 335,000 in CSRS. At the rate of decline in the CSRS population since 1990, by about 2014 virtually all federal employees are expected to be in FERS. Figure 1. Number of FERS Employees Is Increasing 1,400,000 1,279,416 1,173,589 1,200,000 1,000,000 Employees 800,000 600,000 655,750 400,000 335,594 200,000 0 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 1 9 1 9 1 9 19 1 9 1 9 1 9 1 9 19 1 9 2 0 20 2 0 2 0 20 2 0 2 0 20 pt ept ept ept ept ept ept ept ept ept ept ept ept ept ept ept ept ept Se S S S S S S S S S S S S S S S S S CSRS FERS Source: U.S. Office of Personnel Management Central Personnel Data File. Data include full-time, permanent employees. Figure 2 shows the number of employees who were eligible to retire ("retirement eligibles") in both the CSRS and FERS retirement systems from September 2000 until September 2007. Although the number of CSRS retirement eligibles at the end of this period still exceeded the number of FERS eligibles, the difference between the two retirement systems has been steadily narrowing in recent years. CRS-4 Figure 2. Number of FERS Retirement Eligibles Is Increasing 160,000 136,573 140,213 139,214 137,206 136,100 140,000 128,396 108,366 115,296 120,000 100,000 69,933 80,800 80,000 51,255 60,122 60,000 42,913 30,608 36,179 40,000 25,922 20,000 0 0 1 2 3 4 5 6 7 00 00 00 00 00 00 00 00 2 2 2 2 2 2 2 2 pt pt pt pt pt pt pt pt Se Se Se Se Se Se Se Se CSRS FERS Source: U.S. Office of Personnel Management Central Personnel Data File. Data are for full-time, permanent employees. Federal Sick Leave Sick leave benefits provide paid time off while an employee temporarily cannot work due to a non-work-related illness or injury. As noted previously, all federal civilian employees receive four hours of sick leave each biweekly pay period, for a total of 104 hours (13 days) each year. Employees must request sick leave within such time limits as their agencies may require, and agencies may grant sick leave only when supported by evidence they consider administratively acceptable. For absences in excess of three days, or for a lesser period when determined necessary by the agency, an agency may require a medical certificate or other administratively acceptable evidence before approving sick leave requests. The federal sick leave program has been expanded several times, allowing the benefit to be used for purposes other than personal medical attention. For example: ! the Federal Employees Family Friendly Leave Act of 1994 (P.L. 103-388) permitted employees to use up to 13 days of sick leave each year to provide medical or personal care to a family member, or to make arrangements for or to attend the funeral of a family member.10 10 The act guaranteed full-time employees the use of five days (40 hours) of sick leave each year to care for family members or for bereavement purposes. Employees wanting to use more than five days had been required to maintain a sick leave balance of at least 80 hours. However, in 2006, OPM issued final regulations that removed this requirement. See U.S. Office of Personnel Management, "Absence and Leave," 71 Federal Register 47693, August 17, 2006. CRS-5 ! the Treasury-Postal Service Appropriations Act for FY1995 (P.L. 103-329) permitted the use of sick leave to pursue activities related to the adoption of a child. ! in 2000, the federal sick leave program was again expanded, allowing employees to use up to 12 weeks of sick leave per year to care for family members with a serious health condition.11 According to OPM, federal sick-leave usage rates increased during this period -- from an average of 8.59 days of sick leave used per year in 1994 to 9.52 days in 2001. One possible explanation for this overall increase in sick leave use could be the changes in policy that allowed additional uses for sick leave. As explained more fully later in this report, though, another possible explanation could be the increasing number of FERS employees who were approaching retirement eligibility. From time to time, Congress attempts to broaden or otherwise change federal sick leave policy even further. For example, in the 110th Congress, one of the provisions in H.R. 2200 would, if enacted, exempt federal employees who are undergoing medical treatment for a combat-related disability sustained while a member of the armed forces from the requirement in the federal leave sharing program that they exhaust their annual and sick leave before using any transferred leave. S. 1649 in the 110th Congress would (among other things) require OPM to establish a program to allow federal employees serving as caregivers to dependents of members of the armed forces deployed overseas to use their sick leave in the same manner as annual leave is used. (Note: A subsequent "Legislative Developments" section of this report discusses a broader proposed change in federal sick leave policy.) CSRS and Sick Leave The Civil Service Retirement Act of 1920 (P.L. 66-215) created CSRS to provide pension benefits to federal employees. CSRS is a "defined benefit" retirement plan in that the benefit is paid as a lifelong annuity based on years of service and average salary in the last few years of employment. CSRS employees accrue benefits equal to 1.5% of their highest three consecutive years of base pay (often called the employee's "high-three" pay) for their first five years of service, 1.75% of their "high-three" pay for each year in years 6 through 10, and 2% of their "high-three" pay for each subsequent year. Therefore, a CSRS employee who retires with 30 years of service would receive an annuity equal to 56.25% of the employee's "high-three" pay. CSRS annuities are adjusted annually for inflation and are financed 11 Office of Personnel Management, "Sick Leave for Family Care Purposes; Final Rule," 65 Federal Register 37234, June 13, 2000. Until 2006, to use the full 12 weeks of leave, employees had been required to maintain a sick leave balance of at least 80 hours. However, in 2006, OPM issued final regulations that removed this requirement. See U.S. Office of Personnel Management, "Absence and Leave," 71 Federal Register 47693, August 17, 2006. CRS-6 by contributions from both employees and the government (although the CSRS system is currently not fully funded).12 Until 1969, CSRS employees forfeited any unused sick leave at the time of their retirement. At that time, the Civil Service Commission (now OPM) estimated that about half of all retiring federal employees had zero sick leave balances, and the other half had an average of about 44 days (352 hours) of sick leave that were forfeited at retirement. Also, a House Post Office and Civil Service Committee report noted that retiring employees used an average of 40 days (320 hours) of sick leave during their last year of employment.13 In 1969, Congress enacted legislation (P.L. 91-93) to permit CSRS employees to receive service credit for unused sick leave in the computation of their retirement annuities. This change in policy made unused sick leave highly valuable. For example, as noted previously, an employee retiring with 30 years of service and a year of unused sick leave would get credit for 31 years of service for purposes of annuity computation. If this employee's "high-three" salary was $50,000, the additional year of service would increase his or her pre-tax retirement annuity by 2%, or $1,000 per year. Therefore, even before indexing for inflation, if the employee received the annuity for 20 years, the unused sick leave would be worth about $20,000. To pay for this and other changes in the retirement system, the law increased employee and agency contributions to the civil service retirement fund from 6.5% to 7.0% of pay. This change in policy was expected to reduce federal employees' use of sick leave and to grant limited recognition to those who prudently utilized their sick leave. In 1986, the General Accounting Office (GAO, now the Government Accountability Office) examined the implementation of this change in policy, and reported that employees who retired in 1984 and 1985 had significantly higher sick leave balances than employees who retired in 1968, before the law was changed.14 Specifically, the average sick leave balances for employees who retired in 1984 and 1985 were 38% and 46% higher, respectively, than the estimated average balance of 1968 retirees. GAO concluded that the changes made in 1969 had "contributed to reducing the overall usage of sick leave and increasing the unused sick leave balances of retiring federal employees." 12 See CRS Report RL30023, Federal Employee Retirement Programs: Budget and Trust Fund Issues, by Patrick Purcell. 13 Cited in U.S. General Accounting Office, Federal Workforce: Retirement Credit Has Contributed to Reduced Sick Leave Usage, GAO/GGD-86-77BR, June 1986. 14 GAO/GGD-86-77BR. Specifically, GAO reported that retirees in 1968 had an average of 642 hours of unused sick leave. Retirees in 1984 averaged 884 hours and retirees in 1985 averaged 940 hours -- about 38% and 46% higher, respectively, than the average retirees' leave balance in 1968. CRS-7 FERS and Sick Leave The Federal Employees' Retirement System Act of 1986 (P.L. 99-335) created FERS with three basic elements -- (1) Social Security, (2) a FERS basic annuity, and (3) a thrift savings plan (TSP). ! The Social Security component is based on wages earned while a federal employee. Employees may begin receiving reduced Social Security benefits at age 62, but may receive a portion of that amount in the form of a FERS supplement as early as age 55. ! The FERS basic annuity is a defined benefit plan in which employees retiring below age 62 accrue benefits at the rate of 1% of their "high-three" average pay per year of service. Therefore, a FERS employee retiring at age 58 with 30 years of service will accrue a pension benefit equal to 30% of his or her "high-three" average pay.15 Like CSRS, FERS annuities are funded by contributions from employees and the government, but (unlike CSRS) FERS benefits must be pre-funded according to their full actuarial costs. FERS annuitants over age 62 receive annual cost-of- living adjustments, but those adjustments may be less than the amount of inflation.16 ! The TSP is patterned after the "401(k)" savings plans available in the private sector, and encourages employees to save for retirement by making contributions "pre-tax" and by making interest earned on these contributions non-taxable until they are withdrawn. The government contributes 1% of pay for all FERS participants, and provides whole or partial matches for up to 5% of pay contributed by FERS employees. The maximum annual employee contribution is $15,500.17 In contrast to the approach taken with regard to CSRS employees in 1969, Congress did not authorize FERS employees to receive service credit for unused sick leave (except for leave carried into the system by employees who transferred to FERS from CSRS).18 The legislative history for the 1986 act creating FERS does 15 FERS employees who have at least 20 years of service and who work until at least age 62 earn 1.1% for each year of service. Therefore, an employee who retires at age 62 with 30 years of service will accrue a pension benefit equal to 33% of his or her "high-three" average pay. 16 For example, if the amount of inflation is more than 3%, the annuitant would receive an adjustment equal to the amount of inflation minus 1 percentage point. 17 Employees age 50 and older can also make catch-up contributions of up to $5,000 per year. CSRS employees can contribute to the TSP, but they receive no government matching contribution. 18 Retiring employees who transferred to FERS receive service credit for the amount of (continued...) CRS-8 not explain why Congress took this approach. However, in the conference report for the act, the conferees noted that FERS employees' unused sick leave generally could not be used in the computation of their annuities, and urged OPM to examine sick leave usage by FERS employees. The conferees went on to say that they were "concerned that without an incentive to save sick leave, the use of sick leave may substantially increase."19 OPM did not conduct this study until 2006, and the study is described and analyzed later in this report. OPM representatives told CRS that their office had not conducted the study previously because the agency's Central Personnel Data File does not contain data that would allow comparison of sick leave usage rates or sick leave balances by employees in different retirement plans.20 Indications of Greater Sick Leave Use by FERS Employees There have been indications for some time that FERS employees may be using sick leave more often than their CSRS counterparts, particularly as they approach retirement. As noted previously, average sick-leave usage rates for all federal employees increased by more than 10% from 1994 to 2001 (rising from 8.59 days per year in 1994 to 9.52 days per year in 2001). One possible explanation for this increase is the additional uses for sick leave that were allowed during that period (e.g., to care for sick family members or to adopt a child). However, another explanation could be the increasing numbers of FERS employees in the federal workforce and the fact that they receive no benefit for accrued sick leave at retirement. The percentage of full-time, permanent federal civilian employees in FERS increased from about 46% in 1994 to about 62% in 2001.21 Also, a May 2004 survey suggested that FERS employees may be more likely to use sick leave as they approach retirement than their CSRS counterparts. FPMI Solutions, Inc., a human resources services, staffing, and training company primarily for federal agencies, conducted an unscientific online poll of its subscribers (current and retired federal employees) regarding sick leave usage. Of the more than 2,300 participants in the poll, 51% said they were under CSRS and 49% said they were under FERS. Although both groups of participants said they were strongly in favor of sick leave conservation, the groups diverged regarding sick leave use in the last year before retirement. Whereas nearly 85% of CSRS employees and retirees said they would or did conserve as much sick leave as possible, more than 75% of FERS employees and retirees said they planned to use as much sick leave as possible during their last year before retirement. One respondent said that there is "a huge incentive for CSRS and offset employees to conserve sick leave and a very large disincentive 18 (...continued) unused sick leave they had at the time of transfer or the time of retirement, whichever is less. 19 Committee on Post Office and Civil Service, U.S. House of Representatives, Committee Print 99-8, 99th Congress Second Session June 12, 1986, p. 127. 20 Telephone discussions with the Confidential Assistant to the Director and Chief, Administration, Office of Congressional Relations, OPM, May-June 2004. 21 These percentages include employees in federal retirement systems other than FERS or CSRS (e.g., employees in the foreign service retirement system). CRS-9 for FERS employees to do the same."22 Another respondent said that this incentive structure causes "a lot of productivity to go down the drain." However, other respondents said that, this incentive structure notwithstanding, the use of sick leave when not sick was morally "the wrong thing to do." A more scientific study of sick leave use within the Bureau of Prisons from 1994 through 2003 indicated that FERS employees in the agency were, in fact, using more sick leave than those in CSRS, even when controlling other possible factors (e.g., race, gender, education, age, and sick leave balances of employees).23 The authors concluded that the study's findings "strongly suggest that the apparent difference between the incentive systems in the CSRS and FERS systems regarding the use of sick leave is translated into actual differences in behaviors of CSRS and FERS employees."24 FERS Employees Use More Sick Leave Than Their CSRS Counterparts Although OPM's Central Personnel Data File does not contain data on sick leave usage rates by employees in different retirement systems, those data are available from each federal agency or, more centrally, through organizations that administer payroll functions for multiple federal agencies. For example, the Defense Finance and Accounting Service (DFAS) administers payroll functions for nearly 700,000 civilian employees in the Department of Defense and the Executive Office of the President. In 2004, CRS requested data from DFAS on sick leave usage rates and sick leave balances for employees and retirees in the agencies they service. We focused on full-time, permanent employees in General Schedule (GS) and Wage Grade (WG) (or equivalent) pay systems, which included more than 500,000 employees in CSRS and FERS -- about one-third of the full-time permanent workforce outside of the Postal Service, intelligence agencies, and law enforcement.25 Our data request was essentially in two parts. First, for employees who did not retire during the one-year period from June 1, 2003, through May 29, 2004, we requested the average number of sick leave days used during that period and average sick leave balances as of the end of that period for the following categories: 22 CSRS "offset" retirement generally applies to employees who had a break in service that exceeded one year and ended after 1983, and had five years of creditable civilian service on January 1, 1987. When these employees become eligible for Social Security benefits, their annuities will be offset by the value of the Social Security Benefit earned during CSRS offset service. 23 Scott D. Camp, and Eric G. Lambert, "The Influence of Organizational Incentives on Absenteeism: Sick Leave Use Among Correctional Workers," Criminal Justice Policy Review, vol. 17, no. 2 (June 2006), pp. 144-172. 24 Ibid., p. 168. 25 The GS and WG pay systems are the major white-collar and blue-collar pay systems, respectively, in the federal government. CRS-10 ! employees in CSRS and (separately) employees in FERS who were "eligible to retire" without penalty as of June 1, 2003 (i.e., at least 55 years of age and with at least 30 years of service, or at least 60 years of age and 20 years of service, or 62 years of age and 10 years of service).26 ! employees in CSRS and (separately) employees in FERS who were eligible to retire and "almost eligible to retire" (defined as being within two years of retirement eligibility in terms of any combination of age or service -- e.g., an employee who is 53 with 30 years of service or 55 with 28 years of service, or an employee who is 58 with 20 years of service or 60 with 18 years of service). ! employees in CSRS and (separately) employees in FERS who were eligible to retire and "nearing eligibility to retire" (defined as being within five years of retirement eligibility in terms of any combination of age or service -- e.g., an employee who is 50 with 30 years of service or 55 with 25 years of service, or an employee who is 55 with 20 years of service or 60 with 15 years of service). Second, for employees who retired during the June 1, 2003, through May 29, 2004, period, we asked for the average number of sick leave days used per pay period when they were working and their average sick leave balance as of their date of retirement. Differences in Sick Leave Usage and Balances The Defense Manpower Data Center acted as DFAS's agent to generate the sick leave information that we requested.27 As Table 1 shows, the data indicated that FERS employees in the agencies that DFAS supports who were eligible to retire or approaching eligibility generally used more sick leave than their CSRS counterparts during the one-year period ending May 29, 2004. The differences between the two groups were greatest for GS employees as they approached or achieved retirement eligibility. For example, whereas FERS employees in the GS pay system who were either eligible to retire or were within five years of retirement eligibility used only slightly more sick leave during the year than comparable CSRS employees (89 hours versus 85 hours, respectively), FERS employees who were eligible to retire or within two years of eligibility used 25% more sick leave than their CSRS counterparts (105 hours versus 84 hours, respectively). FERS employees who were eligible to retire used nearly 35% more sick leave than comparable CSRS employees (119 hours versus 89 hours, respectively). FERS employees in the WG pay system who were eligible to retire or were within two years of eligibility also used somewhat more sick 26 Although the minimum retirement age increases slightly for employees born after 1947, in this study CRS used age 55 consistently because the differences in eligibility were minimal. 27 According to its website ([http://www.dmdc.osd.mil/about.html]), the Defense Manpower Data Center supports the information requirements of the Office of the Under Secretary of Defense for Personnel & Readiness and other members of the Department's manpower, personnel, and training communities. CRS-11 leave than their CSRS counterparts (e.g., 139 hours versus 127 hours in the "eligible to retire" category), but the differences between FERS and CSRS employees were not as significant as in the GS pay system.28 Table 1. FERS Employees Used More Sick Leave than CSRS Employees as They Approached Retirement Eligibility Average Hours of Sick Leave Used in Previous Year GS/Equivalent WG/Equivalent Status as of May 2004 Employees in Employees in CSRS FERS CSRS FERS Eligible to retire 89 119 127 139 Eligible or within two 84 105 114 119 years of retirement eligibility Eligible or within five 85 89 110 102 years of retirement eligibility Source: DFAS. Differences between CSRS and FERS employees were even more pronounced when considering their average sick leave balances as of May 2004. As Table 2 shows, FERS employees' sick leave balances in all three retirement eligibility categories were only about half as large as their CSRS counterparts (and in some cases were less than half as large). The differences between FERS and CSRS employees were about the same in both the GS and WG pay systems. 28 The number of WG employees in each of these categories was also significantly smaller than the number of GS employees. For example, there were more than 33,000 CSRS employees in the GS system who were eligible to retire, compared to about 5,500 WG employees in that category. CRS-12 Table 2. FERS Employees Had Lower Sick Leave Balances than CSRS Employees Average Sick Leave Balance (in Hours) as of May 2004 Status as of May 2004 GS/Equivalent WG/Equivalent Employees Employees CSRS FERS CSRS FERS Eligible to retire 1,364 715 804 438 Eligible or within two 1,222 540 761 323 years of retirement eligibility Eligible or within five 934 463 604 331 years of retirement eligibility Source: DFAS. As Table 3 shows, for employees who retired from June 2003 through May 2004, FERS retirees' sick leave balances at the time of retirement were significantly smaller than the sick leave balances of CSRS retirees. CSRS retirees (particularly within the GS pay system) also had lower average days of sick leave used per pay period prior to their retirement than their FERS counterparts. Table 3. FERS Retirees Had Lower Sick Leave Balances than CSRS Retirees GS/Equivalent Retirees WG/Equivalent Retirees CSRS FERS CSRS FERS Sick Leave 1,029 313 537 119 Balance (in Hours) at Retirement Average Hours of 7.4 9.9 10.4 10.6 Sick Leave Used Per Pay Period Before Retirement Source: DFAS. Some (and perhaps most) of the differences in leave balances between employees in CSRS and FERS may be due to differences in length of service rather than their retirement systems. For example, the DFAS data indicated that CSRS employees who were eligible to retire had an average of nearly 34 years of service, compared with nearly 24 years of service for FERS employees. Similarly, CSRS employees who retired averaged nearly 32 years of service, compared with nearly 24 years of service for employees in FERS. However, differences in longevity between these groups do not explain why FERS employees' use of sick leave and FERS CRS-13 retirees' use of sick leave just before retirement was greater than their CSRS counterparts. Sick Leave Policies in Other Organizations According to the Bureau of Labor Statistics (BLS), 57% of workers in private industry received paid sick leave in March 2007.29 That figure rose to 68% for full- time workers, and to 80% for employees in management, professional, and related occupations. On the other hand, only 39% of employees in service occupations had access to paid sick leave in 2007, and only 23% of part-time workers. Large employers were more likely to offer sick leave; 67% of employers with at least 100 workers offered the benefit, compared with 48% of employers with fewer than 100 workers. Other studies have indicated that the absence of sick leave can have a wide range of negative effects on family members and coworkers.30 Other BLS data indicated that 53% of full-time employees with paid sick leave in medium and large private establishments in 1997 were allowed to carry over unused sick leave from year to year.31 Seventeen percent were allowed to obtain cash in exchange for unused sick days at the end of the year. Where a cash-in provision was provided, roughly half (9%) also had a provision in their plan allowing leave to be carried over from year to year. For 36% of employees, neither carrying over nor cashing in of unused sick leave was allowed; any leave not used in the year it was earned was forfeited. BLS data indicate that all (100%) full-time employees in state and local government were provided sick leave in 1998.32 These employees generally received a fixed number of days per year, most commonly between 10 days and 15 days. State and local employees with sick leave were much more likely than their private sector counterparts to be able to carry over unused sick leave from year to year (94% versus 53% in medium and large private sector establishments), with about 12% of state and local employees able to both carry over sick leave and receive cash payments (compared to 9% in medium-to-large private sector firms). Only 3% of state and local employees forfeited any unused sick leave at the end of the leave year (compared to 36% in the private sector). 29 U.S. Department of Labor, Bureau of Labor Statistics, National Compensation Survey: Employee Benefits in Private Industry in the United States, March 2007, Summary 07-05 (August 2007). 30 Institute for Women's Policy Research, No Time to Be Sick: Why Everyone Suffers When Workers Don't Have Paid Sick Leave, Washington, D.C., June 2004. 31 U.S. Department of Labor, Bureau of Labor Statistics, Employee Benefits in Medium and Large Private Establishments, 1997, Bulletin 2517 (September 1999). Of the employees who were allowed to carry over unused leave, nearly 80% faced some type of limit on the amount of leave that could be carried over, with the precise amount varying from fewer than 10 days to more than 130 days. 32 U.S. Department of Labor, Bureau of Labor Statistics, Employee Benefits in State and Local Governments, 1998, Bulletin 2531 (December 2000). CRS-14 Unused Sick Leave. The BLS data did not include specific information on how unused sick leave was treated at retirement in either the public or private sectors. However, some surveys of public sector organizations' personnel policies did contain that information. For example, according to a 2002 survey of 428 agency members of the International Personnel Management Association, 58% of the agencies cashed out sick leave at retirement, 45% offered cash for unused sick leave, 11% converted sick leave to vacation time, 9% converted sick leave to insurance at retirement, and 3% converted sick leave to disability insurance.33 In some cases, though, the agencies limited the amount of sick leave that employees could carry over from year to year or limited the amount payable at retirement.34 Federal regulations sometimes provide compensation for unused sick leave at retirement for non-federal employees. For example, employees in federal Fishery Management Councils may accumulate unused sick leave without limit (with Council approval), and "distributions of accumulated funds for unused sick leave may be made to employees upon his or her retirement, or to his or her estate upon his or her death, as established by the Council."35 Police officers, firefighters, and teachers in the District of Columbia who retire on an immediate annuity receive service credit for unused sick leave.36 Sick Leave Policies in State Governments In one of the most comprehensive assessments of employee benefits in state governments, Workplace Economics, Inc., reported that as of January 1, 2004, most states provided their employees with a fixed number of sick days per year (most commonly between 12 and 15 days), but in several states the amount of sick leave accrual was tied to years of service, date of hire, or both. For example, in Hawaii, state employees hired before July 2, 2001, accrue 21 days of sick leave per year, but employees hired after that date accrue 15 days per year for the first nine years of employment and 21 days per year thereafter. Most states placed no limit on the amount of sick leave that could be accumulated; those that did (nine states) most often limited accumulated leave to between 90 and 150 days. 33 Reported in "Sick Leave Abuse: A Chronic Workplace Ill?," Public Management, 84 (June 1, 2002), p. 32. 34 Surveys conducted by other organizations confirm these findings. For example, a 1995 survey of more than 2,600 cities and nearly 900 counties conducted by the International City-County Management Association indicated that nearly half of the local governments (49.6%) paid employees for unused sick leave, and 13.3% permitted conversion of sick leave to vacation leave. See E.R. Moulder and G. Hall, Employee Benefits in Local Government, Special Data Issue (Washington, D.C: International City and County Management Association, 1995), reported by Soonhee Kim, "Administering Family Leave Benefits and New Challenges for Public Personnel Management: The New York State Experience," Review of Public Personnel Administration, 18 (Summer 1998), pp. 42-57. 35 50 CFR 600.120(d). Fishery Management Councils were created by the Magnuson- Stevens Fishery Conservation and Management Act to manage living marine resources up to 200 miles offshore. 36 Department of the Treasury, "Federal Benefit Payments Under Certain District of Columbia Retirement Plans," 65 Federal Register 77500, 77503, December 12, 2000. CRS-15 The study also indicated that 45 states provided some form of compensation for unused sick leave at retirement.37 These states' compensation plans generally fell into three categories: ! states that provided employees with a cash payment for some portion of their unused sick leave, ! states that provided service credit for the leave in the computation of retirement annuity, or ! states that used the unused sick leave to fund some type of health or life insurance for retirees. The type and amount of compensation provided varied substantially within these three categories, and in some cases the states provided employees with options across the categories (e.g., allowing employees to choose either service credit for annuities or health insurance payments), or provided more than one form of compensation. For example, in Nevada, employees were paid for unused sick leave over 30 days to a maximum of $8,000. In addition, Nevada state employees could receive compensation for 50% of the hours in a special sick leave account as either a lump sum payment, advanced payment of an insurance premium, or to purchase additional service credit.38 Cash Payment Plans. In 28 states, retirees were provided a cash payment for a portion of their unused sick leave. In each of these states, the size of the payment was limited in some way -- for example, paying employees for a percentage of their final sick leave balance (most commonly 25% or 50%), capping the number of days or hours of sick leave payable or the size of the payment itself, or calculating the size of the payment based on a formula. In many cases, the states used a combination of these or other factors in determining the amount payable. For example: ! In Virginia, state employees were paid for 25% of their unused sick leave, with the amount of the payment capped at $5,000. ! In New Hampshire, employees were paid for 33% of their unused sick leave, up to 40 days. ! Arkansas used a combination of a formula and a cap on the size of the payment: employees with less than 50 days of unused sick leave on the date at retirement received no payment, employees with 50 to 59 days were paid for 50% of their sick days at 50% of their salary, 37 The five states that generally provided no compensation for unused sick leave were Alaska, Indiana, Maine, Oregon, and Vermont. In Vermont, though, a limited number of employees are covered by a retirement plan in which 50% of unused sick leave is added to an employee's average final compensation. 38 After accruing 90 days of sick leave, Nevada state employees could carry over 50% of unused sick leave each year in a special account to be used for long-term illnesses. CRS-16 with the scale gradually rising to where employees with 80 to 120 days of unused sick leave received payment for 80% of their sick days at 80% of their salary. The maximum payment for unused sick leave was $7,500. In some states, other limits were placed on whether and, if so, how much compensation would be provided to their employees for unused sick leave. For example, in Illinois, unused sick leave earned by state employees from January 1, 1984, through December 31, 1997, was paid at 50% of its value at separation. In Michigan, only retiring employees hired before October 1, 1980, were compensated; they received payment for 50% of their unused sick leave based on their last rate of pay. Service Credit Plans. In 14 states, employees received service credit for unused sick leave in computing retirement benefits (as is currently done for CSRS employees in the federal government), but several states placed conditions or limits on the use of sick leave in this manner. For example, in Georgia, unused sick leave could be used to compute retirement benefits only if the combined balance of sick and annual leave at retirement totaled 120 hours. In Oklahoma, a maximum of one year service credit was allowed, but any credited service of six months or more was rounded up to a full year. In South Carolina, employees could convert up to 90 days of sick leave to retirement service credits, thereby adding up to 4.5 months of service time in calculating the size of their annuity. Insurance Plans. In seven states, the value of some portion of unused sick leave could be credited toward retiree health or life insurance premiums. For example, in Utah, retirees could use unused sick leave to purchase health insurance, with eight hours of sick leave equal to one month of individual coverage. As in the previous categories of compensation, states often placed limits on the use of sick leave in this manner. For example, in Idaho, half of the monetary value of unused sick leave earned since July 1, 1976, or 600 hours (whichever is smaller), could be used to pay the retiree's premiums for group health programs. Policy Options The data presented in this report indicate that FERS employees have been using more sick leave than their CSRS counterparts, particularly as they approached retirement. Although higher sick leave balances for CSRS retirees, retirement eligibles, and near eligibles may be partially a function of longevity, the differences in average usage rates between employees in the two retirement systems lends credence to a longstanding hypothesis -- that FERS employees, who get no value for their unused sick leave, will use that leave rather than simply forfeiting it back to the government at retirement. Various policy options are available to reduce FERS employees' sick leave use, with some designed to entice the employees to conserve their leave balances (i.e., a "carrot" approach), and some designed to more strictly enforce the conditions under which federal employees are allowed to use sick leave (i.e., a "stick" approach). CRS-17 More Stringent Enforcement of Sick Leave Policies As noted previously, federal agencies may grant sick leave only when supported by evidence that they consider administratively acceptable (e.g., a medical certificate). Therefore, if policy makers decide to address this issue, one approach to the apparent greater use of sick leave by FERS employees as they approach retirement would be to enforce existing regulations more stringently -- requiring some type of evidence to support the use of sick leave for more than a certain number of days within a particular period of time. Alternatively, or additionally, agencies could use training and other management practices to reduce sick leave use.39 The advantage of this overall approach is that it could reduce sick leave productivity losses without incurring other costs. However, the success of this approach in encouraging the use of sick leave only when needed would depend on how rigorously it was implemented by the agencies. Also, if employees are able to produce the required documentation easily (e.g., convincing their doctors to provide some evidence of medical treatment or condition), its success in reducing sick leave use may be limited. Providing an Incentive to Conserve Sick Leave The other general approach is to provide FERS employees an incentive to conserve their sick leave. The previous discussion about sick leave policies in state governments suggests that a wide variety of options are available within this approach. One such option would be to allow FERS employees to use unused sick leave in the computation of their annuity. The experience of CSRS employees when they were offered service credit suggests that doing so would have the desired effect. However, FERS annuities are generally much smaller than their CSRS counterparts (because they are only one part of the FERS retirement package), and each additional year of service has less value in FERS than in CSRS (1% instead of 2%). Also, this alternative could prove to be much more expensive than other options over the long term, as larger annuities are payable for as long as the retiree receives a pension. One way to lessen this expense would be to provide partial service credit. For example, if an employee had a year (2,087 hours) of unused sick leave, the employee could be provided six months of service credit, thereby increasing his or her annuity by one- half of one percent. It is unclear, however, whether this relatively small increase in their annuities would be enough to encourage FERS employees to retain their unused sick leave until retirement. For example, a FERS employee with a "high three" salary of $80,000 per year would, under this scenario, increase his or her pre-tax annuity by about $400 per year. Another option would be to provide employees with a cash payment for their unused sick leave. The experience of the states suggests that this payment be limited in some way -- for example, paying employees for 25% or 50% of their final sick leave balance, capping the number of days or hours of sick leave payable or the size of the payment itself, or calculating the size of the payment based on some type of 39 See, for example, Bob Gilson, "Preventing Sick Leave Problems -- Steps an Agency Can Take to Reduce Problems," Fedsmith, January 3, 2007, available at [http://www.fedsmith.com/articles/articles.showarticle.db.php?intArticleID=1121]. CRS-18 formula. From the employees' standpoint, cash payments are likely to have the added appeal of immediacy when compared to the annuity-based incentive, allowing employees to see a relatively fast, tangible benefit to reserving their sick leave. Formula-based systems can be complicated, but can have the advantage of building in additional incentives to save sick leave. For example, a system that pays for 50% of sick leave over a certain level (e.g., 500 hours) avoids de minimus payments while encouraging employees to maintain a roughly three-month cushion of sick leave for longer-term disabilities. On the other hand, employees with smaller sick leave balances may see little or no benefit as a result of this policy, and may continue to use sick leave at a relatively rapid rate. Finally, the experiences of the states suggest that federal policymakers could permit FERS employees to use their unused sick leave to help pay for health, life, or other insurance in retirement. Again, the amount of the payment could be capped either in terms of the proportion of sick leave convertible to this purpose or by a fixed dollar amount. This option was least common in state governments, and may have the least appeal to retiring employees (who could use a cash payment to pay for insurance or any other purpose). A Comparison of Costs Any of these options to provide value for unused sick leave would likely be viewed by FERS employees as more advantageous than the current policy of providing no value for unused sick leave. Less clear, though, is whether the amount of the perceived advantage would be sufficient to prevent the additional sick leave use. Should Congress decide to adopt one of these approaches, the Federal Employees' Retirement System Act requires that FERS benefits be pre-funded according to their full actuarial costs. Therefore, some increase would likely be needed in the employees' contribution to FERS, the federal government's contribution, or both, with the size of the increase dependent upon the approach taken. Given the hidden lost productivity costs associated with the current policy, though, it appears that a change in federal sick leave policy could (depending on how it is constructed) provide a net benefit to the government. For example, the data CRS obtained from DFAS indicated that FERS employees in the GS and WG pay systems who were eligible to retire used an average of about 28 more hours of sick leave in the one-year period ending in May 2004 than their CSRS counterparts.40 OPM data for 2004 indicate that full-time, permanent federal employees who were ages 55 to 59 and had 30 to 34 years of service (indicative of those "eligible to retire") were paid an average of nearly $80,000 per year, or about $38 per hour. Therefore, the 28 additional hours of sick leave used by FERS retirement eligibles may have cost the federal government about $1,064 per employee 40 The FERS-CSRS difference for GS and equivalent workers was 30 hours per year, and the difference for WG and equivalent workers was 12 hours. Because there were about seven times as many GS workers as WG workers in 2004, the weighted average FERS- CSRS difference is 28 hours. CRS-19 in lost productivity (28 hours times $38 per hour).41 OPM data indicate that there were 51,255 FERS retirement eligibles as of September 2004. Assuming that the sick leave patterns revealed in the DFAS data can be extrapolated to the rest of the government, the total lost-productivity cost in that year of the additional sick leave for all FERS eligibles would therefore be about $55 million ($1,064 times 51,255 employees). Greater sick leave use by other FERS employees, particularly those nearing retirement eligibility, would push this lost-productivity cost even higher. These costs occur year after year, and are likely to grow larger, given the increasing number of FERS employees in the government as a whole, and particularly those nearing retirement. Conceivably, the cost to the federal government of providing FERS retirees with some type of value for their unused sick leave could be substantially less than the cost of the current policy -- even if the full cost were paid by the federal government. For example, in 2004 there were about 42,000 voluntary retirements governmentwide. If we assume that 10,000 of these retirees were in FERS, that they had an average "high three" salary of $80,000, and that they had an average leave balance of 1,050 hours,42 giving these FERS retirees full service credit (at 1% per year for six months) would increase their annual annuities by about $400 ($80,000 times 0.01 times 0.5). Over a 20-year retirement, the value of this annuity increase (before any inflation adjustments) would be about $8,000. However, if the employees were given 25% of that value as a lump-sum payment at the time of retirement, each retiree in this example would get $2,000. Therefore, the total cost of this incentive would be about $20 million ($2,000 times 10,000 FERS retirees). While not as generous as the CSRS incentive,43 this lump sum payment could be enough to encourage reduced sick leave use by FERS employees and, if successful, would be substantially less than the estimated lost productivity costs. OPM's 2006 Study of Sick Leave Usage In May 2006, CRS contacted DFAS in an attempt to obtain more recent data on federal employees' sick leave usage rates. OPM responded to CRS, indicating that it was conducting its own study in response to a request from a Member of Congress (who had reminded OPM that Congress had urged the agency to conduct such a study in 1986). In October 2006, OPM reported to the Member of Congress on the results of the study, and the Member's office shared those results with CRS. The overall design of the study was similar to the design that CRS used in its 2004 study, but the 41 These costs are presented as productivity losses because agencies pay their employees the same amount whether they are on sick leave or in regular duty status. 42 This 1,050 hour estimated sick leave balance is higher than the actual average balance of FERS retirees in 2004 because CRS assumed that the incentive will encourage employees to retain their sick leave. 43 In this example, CSRS retirees with a high-three salary of $80,000 and with a leave balance of 1,050 hours would receive an additional $800 per year in additional pension. Therefore, over the course of a 20-year retirement (before any inflation adjustments), the sick leave would be worth an additional $16,000. CRS-20 data covered more employees. Specifically, OPM obtained sick leave use data on full-time permanent, non-Postal Service employees from the four major federal payroll providers: DFAS (the organization providing data to CRS in 2004), the Department of the Interior's National Business Center, the Department of Agriculture's National Finance Center, and the General Services Administration's National Payroll System. The study focused on the amount of sick leave used by three groups of FERS and CSRS employees between April 2005 and March 2006: (1) those already eligible to retire without penalty; (2) those "nearly eligible" to retire (i.e., within two years of being eligible); and (3) all other employees. OPM also examined sick leave use by FERS and CSRS employees who had retired during this period. Some of the OPM study's results appear similar to the results of the earlier CRS study. As Table 4 below shows, the OPM study indicated that FERS employees who were eligible to retire used an average of 20.2 more hours of sick leave per year from April 2005 through March 2006 than their CSRS counterparts. (As described in Table 1 above, the CRS study indicated that white-collar FERS employees who were eligible to retire used 30 hours more sick leave per year in 2003 and 2004 than their CSRS counterparts; blue-collar FERS employees who were retirement eligible used about 12 hours more sick leave per year than similar blue-collar CSRS employees.) Among employees who were within two years of retirement eligibility, OPM reported that FERS employees used an average of 13.5 more hours of sick leave per year than CSRS employees. (In the CRS study, the FERS-CSRS difference for employees eligible and nearly eligible to retire was 21 hours for white-collar employees, and five hours for blue-collar employees.) OPM reported that there were only minor differences in leave usage rates between other FERS and CSRS employees. Table 4. OPM Study Shows FERS Employees Generally Used More Sick Leave in 2005 - 2006 Average Hours of Average Hours of Sick Leave Used Sick Leave Used FERS minus Employee Group per Year by per Year by CSRS Difference CSRS Employees FERS Employees Eligible to retire 84.7 104.9 20.2 Nearly eligible to 80.9 94.4 13.5 retire Other 75.1 73.2 -1.9 Source: OPM. However, in one area the results of the OPM study differed substantially from the CRS study. Among employees who had retired between April 2005 and March 2006, the OPM study indicated that FERS employees used an average of 3.3 hours less sick leave per year than employees who were in CSRS (22.8 hours per pay CRS-21 period for FERS compared with 26.1 hours per pay period for CSRS).44 (In contrast, the CRS study indicated that white-collar employees in FERS who retired in 2003 and 2004 used 2.5 more hours of sick leave per pay period than their CSRS counterparts. For blue-collar workers the FERS-CSRS difference was 0.2 hours more per pay period.) OPM said it did not believe that the higher rate of sick leave use by FERS employees who were eligible and nearly eligible to retire was solely attributable to the fact that they do not receive retirement credit for unused sick leave. The agency said other possible explanations included demographic differences between employees in the two retirement systems and the increased number of ways that sick leave can be used. Nevertheless, using what it characterized as "the unproven assumption that the entire difference in sick leave usage ... is solely attributable to the fact that FERS employees do not receive retirement credit for unused sick leave," OPM estimated that the lost productivity caused by increased use of sick leave by FERS employees who were eligible or nearly eligible to retire cost the federal government $68 million during the April 2005 through March 2006 period. OPM estimated that providing FERS employees with service credit for unused sick leave would increase costs to the retirement fund by about $180 million annually, and would increase the retirement fund liability by about $2 billion, amortization of which would require annual payments of about $137 million over 40 years. In total, OPM said "to pay for future costs and to amortize the cost of benefits based upon past service would have an annual cost of about $317 million." Therefore, OPM concluded that "to make sick leave creditable under FERS would cost several times more than the potential savings." Analysis of the OPM Study Several elements of the OPM study could have affected OPM's conclusions.45 First, the reasons that OPM suggested as alternative explanations to why FERS employees are using more sick leave than CSRS employees can be questioned. OPM said one such reason could be the differences in the demographic characteristics of employees in the two retirement systems (e.g., age differences between FERS and CSRS employees). However, these hypothesized effects can be tested statistically, and the results of such tests in one federal organization do not appear to support OPM's conclusions. As noted earlier in this report, a 2005 study of employees at the Federal Bureau of Prisons concluded that demographic differences between FERS and CSRS employees could not explain differences in sick leave use rates between employees in the two retirement systems.46 Instead, the authors concluded that 44 It is not clear how OPM determined an annual rate of sick leave usage for employees who retired in the first few months of this one-year period. 45 CRS raised these and other issues with officials in both the OPM pay group that did the study and OPM's office of congressional relations in October 2006. CRS also requested the leave usage data that OPM used in its study to verify the results of the study. Despite this and several other subsequent inquiries, as of March 2008, CRS has not received a response from OPM. 46 Scott D. Camp, and Eric G. Lambert, "The Influence of Organizational Incentives on (continued...) CRS-22 differences in sick leave use were directly attributable to differences in how unused sick leave is treated in the two retirement systems. Another reason that OPM said could cause the FERS-CSRS difference in sick leave use was the increased number of ways that sick leave can be used (e.g., for adoption of a child or to care for a sick family member, up to specified limits). However, this seems an unlikely explanation for FERS-CSRS differences in leave usage within the same time period, particularly since both FERS and CSRS employees could use their sick leave for all of these purposes. Second, at least some of the data that OPM used in the study appear to be of questionable validity. For example, OPM said that CSRS employees who had retired between April 2005 and March 2006 used 26.1 hours of sick leave per pay period, while FERS employees who retired during this period used 22.8 hours per pay period. Because each federal pay period is 80 hours, OPM's data suggest that CSRS retirees were on sick leave nearly one-third of the time that they worked during their last year of service, and FERS retirees were on sick leave nearly as much. While these levels of sick leave use are possible, the levels seem unusually high (i.e., nearly 600 to 700 hours per year on an annual basis), and are about three times higher than the DFAS data that were provided to CRS just two years earlier. Also, the OPM data indicated that CSRS employees (who receive significant value for their unused sick leave) used more sick leave per pay period than FERS employees (who receive nothing for unused sick leave) -- again, counter to what the DFAS data indicated in the CRS study, and counter to what one would expect given the current incentive structure. Also, OPM did not explain in any detail how it arrived at its estimates of the cost of lost productivity, or its estimate of the cost of providing FERS employees with service credit for their unused sick leave balances. For the lost productivity estimate, OPM said that it was a function of the number of FERS employees eligible or nearly eligible to retire times the employees' average salary times the difference in the number of hours of sick leave used. OPM identified the number of employees in each retirement system who were eligible or nearly eligible to retire, as well as the average differences in sick leave use for each group, but did not indicate what value(s) it used for the average salary of these employees. Based on its $68 million productivity cost estimate and knowing the other variables, though, it appears OPM assumed an average salary of nearly $54 per hour, or about $112,000 per year. However, OPM data on the federal workforce as a whole in March 2006 indicate that the average salary for full-time, permanent employees was less than $66,000 per year (or less than $32 per hour). For such employees who were ages 50 to 54 with 25 to 29 years of service (i.e., those close to retirement eligibility), the average salary was less than $77,000 per year (or less than $37 per hour). Therefore, assuming all other factors in the equation stayed the same (number of FERS employees eligible and nearly eligible to retire, and leave usage patterns for employees in those groups), OPM's estimate of the annual cost of lost productivity for these FERS employees may be accurate. 46 (...continued) Absenteeism: Sick Leave Use Among Correctional Workers." CRS-23 Finally, and most notably, the only policy option that OPM appears to have considered in its study was to give FERS employees full service credit for unused sick leave -- an option that was characterized by one retirement expert as the "most obvious -- and most expensive" option.47 However, as the discussion of non-federal sick leave policies earlier in this report makes clear, a number of other, less expensive policy options are potentially available to encourage FERS employees to save their sick leave, including partial service credit, lump sum payments for a portion of unused sick leave (based on either the cash value of the sick leave or the value of the service credit), or applying all or a portion of the value of an employee's sick leave balance to the cost of the employee's retirement health care premium. Had OPM considered these other options, its conclusions regarding the economic feasibility of providing value for unused sick leave may have been different. Different approaches and formulas might be tested with focus groups of federal employees to determine which method yields the greatest incentive at the least cost to the government. On the other hand, and as suggested earlier, a completely different way to reduce sick leave use by FERS (and CSRS) employees is to impose more stringent management policies and controls on the use of sick leave. Or, perhaps some combination of "carrot" and "stick" approaches could be tried. The first step, though, may be to have OPM or some other entity conduct a comprehensive, transparent, and methodologically defensible study of federal sick leave usage trends, with defensible estimates of the costs associated with various policy options that Congress could consider. Treasury Inspector General Report On April 24, 2008, the Treasury Inspector General for Tax Administration (TIGTA) issued a report concluding that Internal Revenue Service (IRS) employees in FERS were more likely to use sick leave than CSRS employees, especially as they approached retirement.48 TIGTA said it believed that "the lack of compensation for unused sick leave at retirement has contributed to the higher amount of sick leave used by FERS employees," and recommended that the IRS Chief Human Capital Officer ensure that all IRS managers receive training on leave policies to increase awareness of sick leave abuse and improve sick leave administration. The Chief Human Capital Officer agreed with the findings, but questioned how training would resolve the issue of increased sick leave use by FERS employees. Overall, TIGTA reported that 97,000 IRS employees took more than 15 million hours of sick leave in 2005 and 2006, costing IRS $450 million in salary plus lost productivity. 47 Reg Jones, "Study compares use of sick leave under CSRS, FERS," Federal Times, February 21, 2005. 48 Treasury Inspector General For Tax Administration, "Lack of Compensation for Unused Sick Leave at Retirement Has Contributed to Higher Use by Employees in the Federal Employees Retirement System," Report Number 2008-30-093, April 24, 2008, available at [http://www.treas.gov/tigta/auditreports/2008reports/200830093fr.html]. CRS-24 Legislative Developments On March 10, 2008, Representative Jim Moran introduced H.R. 5573, which would compensate employees under FERS and the Foreign Service Retirement System for a portion of their unused sick leave at the time of retirement. Specifically, the legislation would permit employees49 who satisfied the age and service requirements for immediate or early retirement in those systems to receive a lump- sum payment equal to 15% of the hourly rate of their basic pay for each hour of unused sick leave balance over 500 hours, with the total payment capped at $10,000. For example, an employee with a salary of $75,000 (i.e., $35.90 per hour) and a sick leave balance of 1,250 hours at the time of retirement could receive about $4,000 for her unused sick leave in excess of 500 hours (i.e., 750 hours times $35.90 per hour times 0.15 = $4,039). The legislation indicates that the payment "shall be payable by the agency from which the employee was separated," and "shall be considered pay for taxation purposes only" (e.g., not for purposes of calculating an employee's "high three" average salary in determining her pension). H.R. 5573 was referred to the House Committee on Oversight and Government Reform on the date it was introduced. Representatives from the Federal Managers Association reportedly supported the measure, saying it would discourage many federal employees from taking unnecessary sick days.50 The bill was also supported by the National Treasury Employees Union as "a reasonable and welcome step to correct some of the disparity in treatment for FERS retirees."51 However, one FMA local president said the bill did not go far enough, and that "People are going to keep burning their leave until [Congress] provides a better program."52 An OPM official reportedly remained cautious about whether differences in sick leave usage rates between FERS and CSRS employees would be addressed by the legislation.53 On July 30, 2008, the House of Representatives passed the Family Smoking Prevention and Tobacco Control Act (H.R. 1108), and the legislation was referred 49 The bill defines "employee" to mean an employee as defined in 5 U.S.C. 2105, and also includes employees of the U.S. Postal Service and the Postal Regulatory Commission, but does not include a congressional employee as defined in 5 U.S.C. 2107. 50 Stephen Losey, "FMA praises bill to credit unused sick leave at retirement," Federal Times, Mar. 11, 2008, available at [http://www.federaltimes.com/index.php?S=3417468]. 51 See [http://www.nteu.org/PressKits/PressRelease/PressRelease.aspx?ID=1231] for the NTEU press release. 52 Brittany Ballenstedt, "Legislation would allow more feds to cash out sick leave at retirement," Government Executive, Mar. 10, 2008, available at [http://www.governmentexecutive.com/story_page.cfm?articleid=39484&dcn=todays_m ost_popular]. 53 For example, she said FERS has more women than CSRS, and women may use more sick leave than men because of family responsibilities. Stephen Barr, "Bill Would Give Retirees Partial Pay for Unused Sick Leave," Washington Post, Mar. 11, p. D04, available at [http://www.washingtonpost.com/wp-dyn/content/article/2008/03/10/AR2008031002708 .html]. CRS-25 to the Senate on August 1, 2008. Section 407 of the bill, as amended, would permit FERS employees to receive service credit in the computation of their annuities for any unused sick leave they had at the time of retirement -- just as CSRS employees are currently able to do (although FERS employees receive only a 1% increase in their annuities for each additional year of service, not the 2% that CSRS employees receive). To phase in the program, those who retire in the first three years after the bill is enacted would receive credit for only 75% of their unused sick leave. Representative Moran reportedly said that the bill would cost $70 million in the first five years, and $337 million in the first 10 years.54 The Federal Managers Association and the National Treasury Employees Union have expressed support for the bill.55 The Senate version of the Family Smoking Prevention and Tobacco Control Act (S. 625) does not contain a comparable section pertaining to FERS sick leave. 54 Stephen Losey, "Credit for Unused Sick Leave May Be On the Way," Federal Times, August 3, 2008. 55 Louis C. LaBrecque, "Federal Employees Would Receive Credit for Unused Sick Leave Under House-Passed Bill," Government Employee Relations Report, August 5, 2008, p. 875. CRS-26 Appendix. State Sick Leave Policies The following information on sick leave policies in state government are drawn from the 2004 State Employee Benefits Survey, published by Workplace Economics, Inc. The data are as of January 1, 2004. Where more than one accrual rate or maximum accumulation is listed, the rate depends on length of service or other factors. An asterisk (*) indicates that there are additional or qualifying details in the survey. For example, in Colorado, employees are generally limited to 45 days maximum leave accumulation, but employees hired before July 1, 1988, are limited to their accrual as of that date plus 45 days. Table 5. Sick Leave Policies in State Governments Annual Maximum Payment for Unused Sick State Accrual Accumulation Leave at Retirement Rate Alabama 13 days 150 days Cash payment for 50% of leave balance. Alaska 15 days No limit None. Arizona 12 days No limit Retiree may select cash payment or deposit funds into fund to pay future health premiums as follows: 25% of hourly rate for 500 to 749 hours; 33% of hourly rate for 750 to 999 hours; 50% hourly rate for 1,000 to 1,500 hours. Arkansas 12 days 120 days Cash payment, to a maximum of $7,500, as follows: less than 50 days, no payment; 50-59 days, 50% of days at 50% of salary; 60-69 days, 60% of days at 60% of salary; 70-79 days, 70% of days at 70% of salary; 80-120 days, 80% of days at 80% of salary. California 12 days No limit Service credit used to determine retirement benefits. Colorado 80 hours 45 days* Cash payment for 25% of leave balance. Connecticut 15 days No limit Cash payment for 25% of leave balance, not to exceed 60 days' pay. Delaware 15 days No limit Cash payment for 50% of leave balance, up to 90 days. CRS-27 Annual Maximum Payment for Unused Sick State Accrual Accumulation Leave at Retirement Rate Florida 13 days No limit Cash payment for 25% of leave balance. Georgia 15 days 90 days Service credit used to determine retirement benefits, but only if 120 days of combined unused sick leave and forfeited sick and annual leave. Hawaii 15 days/ No limit Service credit used to determine 21 days* retirement benefits. Idaho 12 days No limit Value of 50% of sick leave used to pay retiree's premiums for group health insurance. Illinois 12 days No limit Cash payment for 50% of sick leave, but only if earned from 1/1/84 through 12/31/97. Indiana 9 days No limit None. Iowa 18 days No limit Cash payment to maximum of $2,000. Kansas 12 days No limit Cash payment as follows: 8 years of service and 100 days accumulated, 30 days pay; 15 years and 125 days, 45 days pay; 25 years and 150 days, 60 days pay. Kentucky 12 days/ No limit Service credit used to determine 22 days/ retirement benefits. 32 days Louisiana No limit Partial payment based on actuarial calculation. Maine 12 days 120 days None. Maryland 15 days No limit Service credit used to determine retirement benefits. Massachusetts 15 days No limit Cash payment for 20% of sick leave. Michigan 13 days No limit Cash payment for 50% of sick leave, but only if hired before 10/1/80. CRS-28 Annual Maximum Payment for Unused Sick State Accrual Accumulation Leave at Retirement Rate Minnesota 13 days No limit Cash payment for 40%of sick leave, but only for mandatory retirement or employees with 10 years of service or age 65 or older. Mississippi 12 days/ No limit Service credit used to determine 10.5 days/ retirement benefits. 7.5 days Missouri 15 days No limit Service credit used to determine retirement benefits (with every 168 hours credited as one additional month of service). Montana 12 days No limit Cash payment for 25%of sick leave. Nebraska 12days/ No limit Cash payment for 25%of sick 14 days/ leave. 18 days Nevada 15 days No limit* Cash payment for excess over 30 days to a maximum of $8,000. Also, any leave in special leave account may be taken as cash payment, insurance payment, or to purchase service credit. New Hampshire 15 days 90 days/ 105 Cash payment for 33% of sick days/ 120 days leave, up to 40 days. New Jersey 15 days No limit Cash payment for 50%of sick leave, up to $15,000. New Mexico 12 days No limit Cash payment for up to 400 hours of sick leave if hours previously cashed in on annual basis exceeds 600 hours. New York 8 days/ 200 days/ 1,500 Up to 165 days may be used as 10 days/ days service credit to determine 13 days retirement benefits, and up to 200 days may be used to pay for health insurance during retirement. North Carolina 12 days No limit Service credit used to determine retirement benefits. CRS-29 Annual Maximum Payment for Unused Sick State Accrual Accumulation Leave at Retirement Rate North Dakota 12 days No limit Cash payment for 10% of sick leave after 10 continuous years of service. Ohio 10 days No limit Cash payment for 50% of sick leave. Oklahoma 15 days No limit Service credit used to determine retirement benefits. A maximum of one year credit is allowed, with six months or more rounded up to one year. Oregon 12 days No limit None. Pennsylvania 13 days 300 days Cash payment for up to 163 days of sick leave, paid under certain full retirement situations. Rhode Island 104 hours 125 days Cash payment for 40-hour employees as follows: 50% of total from 468 hours to 720 hours; 75% of total from 721 hours to 1,000 hours. For 35- hour or non-standard employees, 50% of total from 390 hours to 630 hours; 75% of total for 631 hours to 875 hours South Carolina 15 days 195 days Service credit used to determine retirement benefits. A maximum of 90 days (4.5 months service time) credit is allowed. South Dakota 14 days No limit Cash payment for 25% of sick leave (for those with 7 years of service) to a maximum of 480 hours. Tennessee 12 days No limit Service credit used to determine retirement benefits, with one month of service credit for every 20 days of sick leave. Texas 12 days No limit Service credit used to determine retirement benefits, with one month of service credit for every 160 hours (or fraction thereof) of sick leave. CRS-30 Annual Maximum Payment for Unused Sick State Accrual Accumulation Leave at Retirement Rate Utah 13 days No limit Retiree may use 25% of sick leave to purchase health insurance (eight hours sick leave buys one month of individual coverage) Vermont 6 days/ No limit Generally none, although a 12 days/ limited number of employees 18 days/ are covered by a contributory 21 days retirement plan in which 50% of unused sick leave is added to their average final compensation. Virginia 8 days/ No limit Cash payment for 25% of sick 10 days/ leave (for those with 5 years of 15 days service) to a maximum of $5,000. Washington 12 days No limit Cash payment for sick leave hours over 480 hours at 25% of current salary. West Virginia 18 days No limit Service credit used to determine retirement benefits or converted to credit for health insurance. Wisconsin 16.25 No limit Sick leave converted to credits days to pay group health insurance. Wyoming 12 days No limit Cash payment for 50% of sick leave, to a maximum of 480 hours. ------------------------------------------------------------------------------ For other versions of this document, see http://wikileaks.org/wiki/CRS-RL32596