October 21, 2005

Federal Salary Council
1900 E Street NW
Washington, DC 20415-8200

MEMORANDUM FOR:	THE PRESIDENT'S PAY AGENT HONORABLE ELAINE L.  CHAO HONORABLE JOSHUA B.  BOLTEN HONORABLE LINDA M. SPRINGER
SUBJECT:	Level of Comparability Payments for January 2007 and Other Matters Pertaining to the Locality Pay Program

As authorized by the Federal Employees Pay Comparability Act of 1990 (FEPCA), we present our recommendations for the establishment or modification of pay localities, the coverage of salary surveys conducted by the Bureau of Labor Statistics (BLS) for use in the locality pay program, the process of comparing General Schedule (GS) pay to non-Federal pay, and the level of comparability payments for January 2007.

Bureau of Labor Statistics Surveys and Pay Gap Methodology

We reviewed comparisons of Federal and non-Federal pay using both BLS' old survey data collected under the Occupational Compensation Survey Program (OCSP) from 1994-1996 and newer BLS survey data collected under the National Compensation Survey (NCS) program in 2003 and 2004. All BLS data were updated to March 2005 and compared to Federal pay data as of the same date. The change in non-Federal pay as measured by the nationwide Employment Cost Index was used to update the BLS data. All of the pay gaps we reviewed were calculated using the same general weighting and aggregation methods in use since 1994. Three of the five improvements designed for the NCS program are included in the surveys we reviewed this year.

BLS has implemented the remaining two NCS improvements, the four-factor grade leveling system and better methods for grading supervisory jobs, in surveys now being conducted. However, we will not see the full effect of the new four-factor grading system until BLS completes its 5-year survey replacement cycle. BLS conducts full-scale survey interviews only when it first adds an establishment to the survey sample, and it replaces only 1/5 of the survey sample each year. Surveys delivered to the Pay Agent in 2006 will include the new procedures for grading supervisory jobs and will be the first to contain some data which reflect the four-factor grade leveling system.

The NCS pay gaps continue to be more stable now than before some of the improvements were first implemented in 2002. Overall, the absolute value of change, the minimum and maximum changes, and the range of change in measured pay gaps are less extreme than before the improvements were implemented, as illustrated by the following table:

(Note: Values are not shown for 1997 to 1998 or 1998 to 1999 because the Office of Personnel Management (OPM) did not calculate pay gaps with NCS in 1998. Values are not shown for 2001 to 2002 because three of the five improvements were introduced in 2002, causing a sizable amount of change.)

The pay gap based on NCS data increased since 2004 in 13 locality pay areas and decreased in 19 locality pay areas. Overall, the average NCS pay gap dropped by 1.54 points since 2004. Changes in pay gaps are due to changes in non-Federal pay, sampling error, changes in Federal average salaries, and changes in Federal employment distributions.

In 2002, we agreed to begin phasing in the NCS survey results, with a 50/50 split between OCSP and NCS results for each area. In 2003, we recommended a 75 percent NCS/25 percent OCSP split, and in 2004 we recommended a 90 percent NCS/10 percent OCSP split. Since BLS has continued its implementation schedule for the NCS improvements and the OCSP surveys are now 9 to 11 years old, we recommend that 100 percent NCS survey results be implemented this year and that OCSP data no longer be used.

Establishments with Fewer than 50 Employees

BLS has informed us that it has begun collecting salary data from establishments with fewer than 50 employees and that it will begin to include the data in its publications in September 2006. BLS could include the data in its salary survey deliverables to the Pay Agent in 2007, which would affect pay recommendations for 2009.

Historically, the pay comparability process has not included data from small firms, although there have been a number of attempts to do so in response to criticism that the surveys did not reflect the pay practices of all employers. With the pre-1990 nationwide salary surveys and with OCSP surveys, both of which used detailed job definitions, it was difficult to find jobs matching the survey jobs in smaller establishments that tended not to use detailed job definitions, and it was never determined to everyone's satisfaction whether the results would be substantially different.

Current NCS survey specifications exclude establishments with fewer than 50 employees. While detailed survey job definitions are not an issue with NCS surveys, some members of our Working Group had misgivings about whether jobs in smaller establishments would match very well to Federal jobs. Some members also questioned whether the Government should restrict comparisons to larger establishments, since the Government is such a large employer.

BLS estimates that the number of workers in white-collar jobs matched to GS jobs represented by survey data would increase by about 37 percent if these smaller establishments were included. BLS has also indicated that, resources permitting, we could analyze the NCS data using both the old and new minimum establishment sizes. Since no final decision is needed this year, we recommend that the Council and the Pay Agent further study this matter before any decisions are made about whether to use data from establishments with fewer than 50 employees.

Locality Rates for 2007

Based on calculations provided by Office of Personnel Management staff in taking a weighted average of the locality pay gaps, the overall gap between base General Schedule (GS) average salaries (excluding any add-ons such as special rates and existing locality payments) and non-Federal average salaries surveyed by BLS was 30.36 percent as of March 2005. The amount needed to reduce the pay disparity to 5 percent (the target gap) averages 24.15 percent.

We calculate the pay gaps excluding existing locality payments because locality pay is paid on top of the base General Schedule rates. The overall average pay gap in 2005, including the current average locality rate of 14.99 percent, is 13.37 percent. The calculation is (130.36/114.99-1) X 100.

Under 5 U.S.C. 5304(a)(3)(I), the percentage of comparability payments due in January 2007 may not be less than the full amount of the target gap. Therefore, the overall average locality rate would be 24.15 percent for 2007. We cannot calculate the percentage increase over the average of the rates authorized for 2006 at this time because the 2006 rates have not yet been set. However, we point out that these rates are 1.69 points below the 25.84 percent average rate recommended by the Council for 2006. The overall figure has dropped as we phase in the NCS program because NCS pay gaps are generally lower than pay gaps based on the old OCSP surveys. The recommended comparability payments for 2007 for each locality pay area are shown in Attachment 1.

These locality rates would be in addition to the increase in General Schedule base rates under 5 U.S.C. 5303(a). This provision calls for increases in basic pay equal to one-half of one percentage point less than the percentage by which the Employment Cost Index (ECI), wages and salaries, private industry workers, increased between September 2004 and September 2005. The ECI for September 2005 will not be published until October 28, 2005, so the amount is not known at this time.

Surveys in New Cities

In 2003, we recommended that surveys be expanded in six metropolitan areas currently in the RUS locality pay area where limited BLS data indicated relatively high pay. While BLS has not yet redesigned its sample in these areas, BLS has indicated it will do so and included them in its pay model again this year. The pay gaps are shown in the following table:

While these surveys were not designed as full-scale locality pay surveys, and the proportion of modeled data tends to be above that found in existing locality pay areas, we recommended and you tentatively approved making Buffalo, Phoenix, and Raleigh locality pay areas in 2006. This year the pay gaps in these three locations continue to be well above that for the RUS locality pay area, while the pay gap for Austin continues to be below that for the RUS locality pay area. Last year Memphis was slightly above the RUS locality pay area, while Louisville was slightly below the RUS locality pay area. This year, Louisville is slightly above the RUS locality pay area, while Memphis is slightly below the RUS locality pay area.

We continue to recommend that Austin, Louisville, and Memphis should not be made separate locality pay areas at this time and that you should complete implementation of new locality pay areas for Buffalo, Phoenix, and Raleigh in 2006. BLS informed us that it continues to plan its geographic redesign and will begin implementing the new surveys in 2007 and 2008. (This includes reviving the Raleigh survey, which was cancelled last year as part of a budget reduction.)

Definitions of Locality Pay Areas

On June 20, 2005, OPM published a proposed rule in the Federal Register on behalf of the Pay Agent. The proposed rule would implement the changes in locality pay areas we recommended including merging Kansas City, St. Louis, and Orlando with the RUS locality pay area and creating new locality pay areas for Buffalo, Phoenix, and Raleigh. The notice also adds Fannin County to the Dallas locality pay area because the Office of Management and Budget revised the Dallas Combined Statistical Area (CSA), makes a minor change in the description of the Los Angeles locality pay area, and notes that Culpeper County, VA, is now part of the Washington-Baltimore CSA, so it is also now part of the pay area proper and no longer an area of application.

The notice indicates that no adjacent areas pass our area-of-application criteria for the new Buffalo and Phoenix locality pay areas. Based the area-of-application criteria, the Fayetteville, NC Metropolitan Statistical Area (MSA), consisting of Hoke and Cumberland Counties, NC, and the Goldsboro, NC MSA, consisting of Wayne County, NC, will be included in the new Raleigh locality pay area. We support these additions to the new Raleigh locality pay area.

Federal Correctional Complex, Butner, NC

You also asked for our views on what should be done about the Federal Correctional Complex, Butner, NC, which is partially in Durham County, within the Raleigh CSA, and partially in Granville County, outside of the Raleigh CSA. In fact, the county line runs right through several buildings in the complex, but the Low Security Institute is on the Granville County side of the line by a few hundred yards and would not be in the Raleigh locality pay area.

Granville County passes the commuting criterion for inclusion in the pay area, but fails the GS employment criterion of 400 or more GS employees. The Correctional Complex also fails the current criteria for evaluating Federal facilities that cross county lines. However, you indicated in the Federal Register notice that you believe it would not be administratively feasible or desirable to include only part of the Correctional Complex in the new Raleigh locality pay area and proposed to include the entire facility.

We share your view that the facility should not be split and recommend the following modification to the existing criteria for evaluating Federal facilities that cross county lines in order to ensure that the entire Correctional Complex will be included in the Raleigh locality pay area:

For Federal facilities that cross locality pay area boundaries: To be included in an adjacent locality pay area, the whole facility must have at least 500 GS employees, with the majority of those employees in the higher-paying locality pay area, or that portion of a Federal facility outside of a higher-paying locality pay area must have at least 750 GS employees, the duty stations of the majority of those employees must be within 10 miles of the separate locality pay area, and a significant number of those employees must commute to work from the higher-paying locality pay area.

Contacts with employees

OPM staff had contacts from the following areas since the fall of 2004 requesting higher locality pay:

The contacts also included a request to merge the Sacramento and San Jose-San Francisco locality pay areas, a petition from the Federal Executive Association of Western Massachusetts asking that we include Berkshire County, MA, in the Hartford locality pay area, a letter from Representative Hefley (R-CO) requesting that Colorado Springs be added to the Denver locality pay area, and a request from the Federal Executive Boards of Los Angeles and San Jose-San Francisco to provide higher pay for Federal employees in the core cities.

None of these areas pass the criteria previously agreed to by the Council. Although we have asked our Working Group to further review the situation in some of these areas, we are not recommending changes in locality pay area boundaries at this time as a result of these contacts.

Locality Pay Areas for 2007

We recommend the same locality pay areas for 2007 as those we anticipate you will approve for 2006, as follows:

1. Atlanta-Sandy Springs-Gainesville, GA-AL Combined Statistical Area
2. Boston-Worcester-Manchester, MA-NH Combined Statistical Area, plus the Providence-New Bedford-Fall River, RI-MA Metropolitan Statistical Area, Barnstable County, MA, and Berwick, Eliot, Kittery, South Berwick, and York towns in York County, ME
3. Buffalo-Niagara-Cattaraugus, NY Combined Statistical Area
4. Chicago-Naperville-Michigan City, IL-IN-WI Combined Statistical Area
5. Cincinnati-Middletown-Wilmington, OH-KY-IN Combined Statistical Area
6. Cleveland-Akron-Elyria, OH Combined Statistical Area
7. Columbus-Marion-Chillicothe, OH Combined Statistical Area
8. Dallas-Fort Worth, TX Combined Statistical Area
9. Dayton-Springfield-Greenville, OH Combined Statistical Area
10. Denver-Aurora-Boulder, CO Combined Statistical Area, plus the Ft. Collins Loveland, CO Metropolitan Statistical Area and Weld County, CO
11. Detroit-Warren-Flint, MI Combined Statistical Area, plus Lenawee County, MI
12. Hartford-West Hartford-Willimantic, CT Combined Statistical Area, plus the Springfield, MA Metropolitan Statistical Area and New London County, CT
13. Houston-Baytown-Huntsville, TX Combined Statistical Area
14. Huntsville-Decatur, AL Combined Statistical Area
15. Indianapolis-Anderson-Columbus, IN Combined Statistical Area, plus Grant County, IN
16. Los Angeles-Long Beach-Riverside, CA Combined Statistical Area, plus the Santa Barbara-Santa Maria, CA Metropolitan Statistical Area and all of Edwards Air Force Base, CA
17. Miami-Fort Lauderdale-Miami Beach, FL Metropolitan Statistical Area, plus Monroe County, FL
18. Milwaukee-Racine-Waukesha, WI Combined Statistical Area
19. Minneapolis-St. Paul-St. Cloud, MN-WI Combined Statistical Area
20. New York-Newark-Bridgeport, NY-NJ-CT-PA Combined Statistical Area, plus Monroe County, PA, and Warren County, NJ
21. Philadelphia-Camden-Vineland, PA-NJ-DE-MD Combined Statistical Area, plus Kent County, DE, Atlantic County, NJ, and Cape May County, NJ
22. Phoenix-Mesa-Scottsdale, AZ Metropolitan Statistical Area
23. Pittsburgh-New Castle, PA Combined Statistical Area
24. Portland-Vancouver-Beaverton, OR-WA Metropolitan Statistical Area, plus Marion County, OR, and Polk County, OR
25. Raleigh-Durham-Cary, NC Combined Statistical Area plus the Fayetteville, NC Metropolitan Statistical Area, the Goldsboro, NC Metropolitan Statistical Area, and the Federal Correctional Complex, Butner, NC
26. Richmond, VA Metropolitan Statistical Area
27. Sacramento--Arden-Arcade--Truckee, CA-NV Combined Statistical Area, plus Carson City, NV
28. San Diego-Carlsbad-San Marcos, CA Metropolitan Statistical Area
29. San Jose-San Francisco-Oakland, CA Combined Statistical Area, plus the Salinas, CA Metropolitan Statistical Area and San Joaquin County, CA
30. Seattle-Tacoma-Olympia, WA Combined Statistical Area
31. Washington-Baltimore-Northern Virginia, DC-MD-VA-WV Combined Statistical Area, plus the Hagerstown-Martinsburg, MD-WV Metropolitan Statistical Area, the York-Hanover-Gettysburg, PA Combined Statistical Area, and King George County, VA; and
32. Rest of U.S.--consisting of those portions of the continental United States not located within another locality pay area.

Allocating Locality Pay in 2006

At this point, we do not know what pay raise will be provided in 2006. Congress may provide a total increase of 3.1 percent. If that amount is approved, we recommend a 2.1 percent across-the-board pay increase (the same as would occur under existing law) and an increase of 1.0 percent of payroll for locality pay, distributed so that locations with the largest pay gaps receive the largest increases. Attachment 2 shows how such amounts could be allocated based on our method of allocating funds used in the past.

We note that there are two complications to the process this year, since we have three new locality pay areas being implemented and three others being merged with the RUS locality pay area. For Buffalo and Raleigh, we have used the average phase-in rate of 58.84 percent (based on the other locality pay areas in 2005) applied to the target pay gap in each area as the starting point for allocating their share of funds available for locality pay in 2006. In Phoenix, that resulted in a starting rate below the RUS locality pay rate, so we used the RUS rate as the starting point for Phoenix in the calculation. We have also set the locality rates in Kansas City, St. Louis, and Orlando equal to the rates that would apply in the RUS locality pay area. This allocation provides a small locality pay increase for employees in those three areas, even as they are merged with the RUS area.

By direction of the Council:

Terri Lacy
Chairman

Attachments:

• Attachment 1
• Attachment 2