United States General Accounting Office Testimony GAO For Release Federal White-Collar Employee Salary Reform on Delivery Expected at 9:30 a.m. EST Wednesday March 21, 1990 Statement of Richard L. Fogel, Assistant Comptroller General for General Government Programs Before the Committee on Governmental Affairs United States Senate FEDERAL WHITE-COLLAR EMPLOYEE SALARY RJZFOFU'l Summary of Statement by Richard L. Fogel Assistant Comptroller General for General Government Programs GAO strongly endorses actions to reform federal white-collar employee pay-setting principles and processes. The salary comparability principle adopted in 1962 has been ineffective because (1) increases necessary to maintain comparability with rates paid by nonfederal employers have not been granted for many years, and (2) national average nonfederal salary rates used to determine comparability have little relevance to nonfederal rates in many localities where federal employees work. Also, the comparability principle does not consider the wide variances in the cost-of-living in different parts of the country. Noncompetitive salary rates are the major reason for federal recruitment and retention difficulties, especially in high-cost, high-paying localities. GAO believes that federal salary rates must be restored to competitive levels in a fiscally responsible manner. S.2214 accomplishes this objective by initially calling for adjustments to basic pay rates in localities where nonfederal pay rates are above the national average. The bill does not require the direct comparison of federal and nonfederal salaries in the localities, but is designed to make federal salaries more competitive in those localities where the federal pay disparities are the most severe. GAO suggests that direct comparisons of federal and nonfederal salaries would be more appropriate and believes that cost-of-living levels should also be considered in determining locality adjustments. GAO also points out that private sector companies often grant salary increases to individual employees based on their job performance and suggests that requiring greater accountability from federal employees in return for higher salary levels is an issue that needs serious consideration. Mr. Chairman and members of the subcommittee: Thank you for the opportunity to appear here today and provide our views on the issue of federal white-collar employee pay reform. This is an extremely important subject on which we have done considerable work and have several reviews currently in process. Our work all points to one basic conclusion. The federal government must recognize that its pay disparities with the . nonfederal sector can no longer be tolerated. Federal recruitment and retention difficulties are real. The government must be able to compete for and hire quality employees if it is to effectively deliver services to the American people. WEIAT IS TEE PAY PROBLEM AND WBAT CAUSED IT? The Federal Salary Reform Act of 1962, as refined by the Federal Salary Reform Act of 1970, established the principle that federal salary rates for white-collar employees should be comparable with average rates paid by private sector employers for similar jobs. The comparability principle holds that the private sector will determine the "going rates" for jobs like those found in the government, and by paying national average rates, the government, federal employees, and the taxpayers are assured that federal salary rates are reasonable and competitive. Two factors have prevented the comparab ility pr inc pie from operating as intended. First, the process established to maintain salary comparability with the private sector has not been followed for many years. Every year, beginning in 1978, Presidents have decided to grant federal pay raises at lesser amounts than needed to maintain comparability. As a result, a large gap between national average federal and private sector salaries gradually developed over the years. The federal salary increase recently granted in January provides a good illustration of how the gap continues to grow. To stay even with the average salary increase in the private sector during the previous year, federal salaries should have been increased about 6.4 percent. Instead, a 3.6 percent increase was granted, thereby widening the gap even further. The other factor that caused the pay comparability principle not to accomplish its objective was its assumption, implicit in the use of national averages, that nonfederal rates are similar in the different parts of the country where federa 1 employees work. While very 1 imited 2 information is now publicly available on how federal rates for particular jobs in specific locations compare to nonfederal rates in the same localities, we are finding that substantial variances exist. A national average often has little relevance to nonfederal rates in a specific locality. For example, our work in connection with the National Commission on Law Enforcement showed that state and local government salaries for law enforcement personnel clearly varied by region, with the lowest salaries reported in the South, Southwest, and rural areas of the Midwest. The highest salaries were in California, New York, Washington State, New Jersey, and Massachusetts. In another analysis, we are comparing private sector and federal salaries by job level and location for certain occupations primarily in the technical and clerical job categories. Our analysis was limited to about 60 of the over 250 metropolitan areas in the country. of the locality/job level comparisons we were able to make, we found that the private sector paid more than the federal government about 90 percent of the time. The private sector pay advantages ranged from minor amounts to over 100 percent. In over 80 percent of the comparisons, the private sector advantage was at least 10 percent. Overall, 3 the median private sector pay advantage was over four times greater than the median federal pay advantage. Among the areas in our analysis where the federal rate was most behind the private sector rates were Detroit, San Francisco, Los Angeles, Newark, Boston, and New York. The comparability principle also does not consider the wide variance in the cost-of-living rn different parts of the country. When two federal employees in different locations are paid the same salary, their purchasing power can be quite different depending on the cost-of-living in their locations. Among the 59 metropolitan areas with at least 5,000 federal white-collar employees, living costs vary more than 65 percent. Housing costs differ widely by area and are the primary determinant of the living cost variances. For example, the current market value for a house in San Francisco is more than three times the market value of the same house in San Antoniol. Yet, federal employees with the same job in these two cities receive the same base pay under the federal General Schedule salary system. In general, the areas with the highest nonfederal salary levels also have the highest cost of living. However, lThe standard house used for comparison purposes contained 1,400 square feet, 3 bedrooms and 1.5 baths. 4 there are many exceptions. For example, salary rates are quite high in Houston, but Houston's cost-of-living iS below the national average. Conversely, San Diego has relatively average salary rates, but its cost of living ranks 10th among the 59 areas with 5,000 or more employees. While we believe it is appropriate to continue a national salary structure for white-collar employees, we are convinced that locality-based pay adjustments are needed because of the significant differences in nonfederal salary rates and cost-of-living by locality. WFlAT ARE TBE EFFECTS OF TEE PAY PROBLEH? If federal salary rates are not competitive, the implications are clear -- the government will be unable to attract and keep the kinds of employees it needs to carry out the Nation's business. Unfortunately, our work and studies by other groups, such as the Law Enforcement Commission and the National Commission on the public Service, have shown that recruiting and retention problems are occurring and can only get worse if corrective action is not taken. Some examples illustrate these problems: According to federal law enforcement personnel, lack of competitive pay deters qualified people from 5 applying. More than half of all managers and employees the Law Enforcement Commission surveyed felt this to be true and many law enforcement officials believe it is the main reason law enforcement personnel leave federal service. In Boston, entry-level school bus drivers receive higher pay than journeyman federal aircraft mechanics. The Internal Revenue Services' Newark, New Jersey office found the applicants for revenue agent positions to be of such poor quality that it chose to leave 50 positions vacant; and The Navy reported it is losing electronics technicians who work on sophisticated radar systems at the Alameda Naval Air Station because they can make $5 more an hour working on stoplights for the local city government. We are finding that the greatest reason employees quit federal jobs is low salary levels. The pay disparities are occurring in virtually all locations, but the recruitment and retention problems are most pronounced in areas where nonfederal salary rates and the cost-of-living are the 6 highest and in areas where other employment opportunities are readily available. Agencies tell us that high turnover rates cause them numerous operational problems, including reduced service delivery, increased recruiting and training costs, more overtime pay, and upper-level employees having to do lower-level work. Of potentially greater long-term significance, we are finding in a series of meetings with college students that they have very little interest in seeking a federal job when they graduate. They have been subjected to so much anti-government rhetoric that they appear convinced that more challenging careers are available in the private sector. They particularly find federal pay rates to be unsatisfactory or "exploitive" as one student described them. Of 64 students in our various discussion groups, only 2 indicated a willingness to accept a starting salary at the federal entry-level rate. Interestingly, the students thought the threshold for an appropriate salary is a function of what it costs to live in a given location. They saw little logic to national salary rates when the cost-of-living was so different across the country. Hopefully, positive recruiting efforts can help to counteract the negative perceptions of federal employment that college students hold. However, it is apparent that 7 low entry-level salary rates will continue to be a problem unless corrective actions are taken. PRIVATE SECTOR PAY PRACTICES As part of our assessment of federal pay and other employment programs, we are surveying the practices followed by large companies that, like the government, have employees in numerous locations. We sent questionnaires to all 148 companies in the country that our data source indicated had at least 25,000 employees each and had employees in at least 10 locations. The survey is not yet complete, but to date we have received usable responses from 62 companies that met our selection criteria. Responses from 14 other companies indicated that they either did not meet our selection criteria or were so decentralized and diverse that they could not complete the questionnaire from a company-wide standpoint. Since the survey is still on-going, we cannot at this time say anything definitive about private sector practices in general, but we are seeing some very interesting information from the responses received to date. The questionnaire asked the companies to indicate their wage and salary objective in relation to market rates. 8 Almost without exception, the 62 companies said their objective is to have pay rates no lower than the 50th percentile of rates paid by other employers, thereby indicating that maintaining competitive salary levels iS a very important aspect of the compensation philosophy of the Nation's leading private companies. Some of the companies said their objective is to pay above the 50th percentile. Since the government's pay rates are around the 40th percentile of national average salaries, these findings are further evidence that the government is not a competitive employer. We also asked the companies whether they paid different rates by locality. The 62 respondents indicated a wide variety of pay practices, including local schedules for all job categories, national schedules for all job categories, national schedules for some categories (most often top officials and managers and professional employees) and local schedules for others, and some local and some national schedules for the same employee groups. In some cases, the companies had local variations for their employees on national schedules, such as different pay raise amounts for employees in different locations and cost-of-living allowances for employees in high-cost areas. 9 The responses received to date do not show any predominant pattern in the private sector on locality pay practices for white-collar employees. It is evident, however, that companies give a great deal of thought to how they set salary levels to meet their own needs and that there is often considerable variation in pay practices by locality. One company, for example, reported that it has divided the country into three regions based on analyses of the cost- of-living in its various locations. Base salaries are set for the lowest cost-of-living region, and employees in the other two regions receive 10 percent and 20 percent add-ons to the base salaries for their pay grades. This practice is followed for employees in all job categories. Another company reported it uses local salary schedules for all employee categories except top executives. On the other hand, one company said it paid all employees at national rates, and these national rates were based on prevailing salary levels in New York City where the company's headquarters is located. This company acknowledged that its pay rates in many parts of the country are higher than needed to be competitive with other employers. 10 =AT KINDS OF CEANGES ARE NEEDED? Our work strongly suggests that the following basic principles should be followed in determining federal salary amounts: 1. Federal salaries must be competitive by locality to attract and retain high-quality employees. 2. In determining prevailing salaries in a locality, rates paid by state and local governments as well as private companies should be taken into account. (The law now limits salary comparisons to the private sector.) 3. The restoration of federal pay to competitive amounts must be done in a fiscally responsible manner. 4. Action should be taken to avoid any further widening of the national pay gap. S.2274 is generally consistent with the above principles. It uses nonfederal pay relatives to determine locality pay adjustments and calls for a plan to be developed later to correct any remaining pay disparities. To stop the overall pay gap from widening, the bill also requires that General Schedule rates be adjusted each year to reflect increases 11 in the Employment Cost Index (ECI) compiled by the Bureau of Labor Statistics. We agree with the bill's incremental approach to resolving the government's salary problems. The vast gulf that has been allowed to develop between federal salary levels and market rates is simply too large to correct at one time. However, we would point-out that the pay relatives the bill uses to determine locality adjustments measure local nonfederal pay rates against national average nonfederal pay rates. They do not compare local federal and nonfederal salaries. Accordingly, these pay relatives will not show the extent to which federal salaries are behind (or ahead of) prevailing nonfederal rates in any locality. They do, however, afford a means by which an initial step can be taken to provide some relief in those localities where the pay disparities are the most severe. We think it would be more appropriate over the long term to directly compare federal and nonfederal salaries in each locality. Moreover, we believe cost-of-living differences among localities should be considered along with the salary comparisons in determining the locality adjustments. Since the primary objective of the bill is to make federal pay rates more competitive, not necessarily comparable with 12 nonfederal rates, including cost-of-living considerations in deriving locality pay adjustments would better accomplish that objective. Also, as mentioned previously, nonfederal pay rates do not always mirror cost of living differences by locality. By adding cost of living differences to the salary comparisons, federal salary rates can be equalized in terms of spending power. Adjustments, however, should not result in federal salary rates exceeding the nonfederal rates in any area. If our suggestions are adopted, a decision would then have to be made on how much of the pay disparities by locality could be corrected in the short run. One possible approach is to fill some percentage, maybe one-half, of the pay disparities. As the bill provides, a reading could be taken later of the budget situation and a plan developed for correcting any remaining disparities. A further alternative is to begin to reduce the disparities immediately only in cities with the greatest differences between federal and nonfederal rates. Also, though it may take some time to reduce the disparities for all levels of the federal workforce, it is critical that the disparity at 13 the entry level be dealt with quickly to better insure the government's competitiveness. The notion of affordability also has another aspect. We do not believe agencies, across the board, should be expected to absorb all or large parts of the cost of pay increases within their existing budgets. If Presidents and Congress want good government, they must pay for it. Once it has been decided what federal programs are appropriate to be funded, the costs of properly carrying out the programs must also be funded. These costs include equitable compensation levels for the employees who manage and operate the programs. CONSIDERATION OF PERFOFUUNCE LEVELS IN UAKING SALARY ADJUSTUENTS In our opinion, the bill is on the right track in calling for a study of the feasibility of linking pay adjustment amounts and individual employee performance. According to the questionnaires we have received thus far from private companies, very few companies grant automatic pay increases to all employees when salary schedules are adjusted. They generally report that job performance is the primary criteria used in determining the amount of increase to be 14 received by individual employees. We think the concept has merit and should be explored. * * * * * * That concludes my statement. My colleagues and I would be pleased to answer any questions you may have. 15
Federal White-Collar Employee Salary Reform
Published by the Government Accountability Office on 1990-03-21.
Below is a raw (and likely hideous) rendition of the original report. (PDF)