B. Regulatory framework OMB classification of metropolitan areas

   The Federal Office of Management and Budget (OMB) has responsibility for developing a variety of statistical programs. See generally 44 U.S.C. §3504 (1994). Pursuant to this statutory directive, OMB devises standards for categorizing urban areas throughout the United States. The OMB criteria for designating metropolitan areas are re-evaluated on a continuing basis and were revised in 1958, 1971, 1975, 1980, and 1990, and are being reviewed now in anticipation of the 2000 census. See Notices, OMB, 63 Fed. Reg. 70526, 70528 (1998). The criteria currently in effect, "Revised Standards for Defining Metropolitan Areas in the 1990's," were published in 1990. 55 Fed. Reg. 12154 (1990).

   OMB's standards for categorizing metropolitan areas are complex. The general concept of a "metropolitan area," as defined by OMB, is that it includes a core area containing a large population nucleus, together with adjacent communities having a high degree of economic and social integration with the core. Thus a metropolitan area typically includes central cities and their outlying but integrated counties. Id.

   Under OMB's 1990 standards, the basic urban unit is the "metropolitan statistical area" (MSA), which typically includes a city or urbanized area of at least 50,000 in population joined by surrounding counties where a sizeable portion of the population commutes to the center. Id., §§1-3. MSAs are classified by population size, from the smallest Level D MSA (generally having a population of 50,000 - 100,000) to the largest Level A MSA (a metropolitan statistical area with population over 1,000,000). Id., §6. If there are two or more adjacent urban areas that independently would qualify as metropolitan statistical areas, they are combined into a single, larger MSA if a substantial percentage of workers commute between the adjacent urban areas. Id., §5. With regard to the Puget Sound region, there are four areas that independently would qualify as MSAs based on their population size and urban core areas: Bremerton, WA (Level C); Olympia, WA (Level C); Tacoma, WA (Level B); and Seattle-Bellevue- Everett, WA (Level A). See, e.g., OMB Bulletin 99-04, "Revised Statistical Definitions of Metropolitan Areas (MAs) and Guidance on Uses of MA Definitions," June 30, 1999, at List II p.23. Under the OMB standards, however, these four urbanized areas are merged into a single Level A metropolitan statistical area.

   OMB's rules recognize that within the large, Level A metropolitan statistical area there may be several of these smaller, identifiable cities or urban areas. If these smaller urban areas have a population of at least 100,000 and meet several other criteria, they are denominated by OMB as Primary Metropolitan Statistical Areas (PMSAs). Some of the urban centers now classified as PMSAs formerly had been designated as "Standard Metropolitan Statistical Areas" (SMSAs) under OMB's pre-1990 guidelines. 55 Fed. Reg. at 12156, §8. Under this aspect of the OMB standard for defining metropolitan areas, the four urban centers in the Puget Sound region Bremerton, Olympia, Tacoma, and Seattle-Bellevue-Everett all are classified as PMSAs. See OMB Bulletin 90-44, supra. That is, each of the four urban centers independently qualifies as a metropolitan area based on its size and economic factors, but they are part

adjusted upward in many metropolitan areas to reflect higher wage rates and living costs in urban locations. 5 U.S.C. §§5301-07; see AR, Tabs Y, Z. By 1996, BLS's OCSP survey of the six-county Seattle CMSA had replaced the earlier BLS surveys restricted to the smaller areas in the Puget Sound region, including the separate wage surveys for the Seattle, Tacoma and Bremerton-Shelton areas. AR, Tabs A-C, W.

D. Impact of the changed BLS survey program on the Wage and Hour Division's Puget Sound wage determinations

   Before 1996 (i.e., when BLS still conducted separate wage surveys for Seattle, Tacoma and Bremerton-Shelton), the Administrator viewed each of these three areas as a "locality" under the Act, and routinely issued separate wage determinations for each location. Because the wage determination rates in each location were based on distinctly separate surveys, the Administrator's wage determination rates varied in each location among the various classes of occupations. AR, Tabs M, N, P, Q.

• Instead of shifting to a single set of wage determinations that would apply to the entire Seattle CMSA, the Administrator continued to issue separate wage determinations for the smaller Seattle, Tacoma and Bremerton-Shelton areas as a transitional measure. Where the CMSA data would have resulted in a large increase above the rates in the predecessor wage determinations, the Administrator capped the increase in the wage determinations at 15% above the earlier wage determination rates.
• For protective service occupations, new wage determination rates were issued based directly on the CMSA data, without applying the 15% cap. Historically, the Administrator would have relied on BLS regional data for calculating such wage rates for Tacoma and Bremerton-Shelton, because the wage surveys formerly conducted in these two areas did not cover local and state government, the usual source of wage data on protective service professions. In setting wage rates for those professions in the 1997 Tacoma and Bremerton-Shelton CMSA-based wage determinations, however, the Administrator relied directly on the Seattle CMSA wage survey because it included data for those professions and because the wage rates for those professions in the prior Tacoma and Bremerton-Shelton wage determinations had been based on data from outside the Tacoma and Bremerton-Shelton areas. See n.7, supra. Because the predecessor wage determination rates had not been based on local data, the Administrator concluded that it was not necessary to ease the transition to reliance on the Seattle CMSA wage data.

AR, Tabs E, F, BB, DD. As relief, the Petitioners requested that the Administrator revise the challenged wage determinations and that the Wage and Hour Division return to the individual locality survey method formerly used for developing SCA wage determinations. AR, Tabs D-F.

   On April 20, 1998, the Administrator issued three final ruling letters, responding to the review requests. AR, Tabs A-C.

    With regard to the challenge to the wage rates for protective service occupations (fire, police), the Administrator considered and rejected the wage data generated by the Navy's survey of Kitsap County fire and police departments. The Administrator concluded that the Navy's survey data could not form the basis for a challenge to the rates in the wage determination because the Navy's documentation did not indicate whether the survey results represented average rates, or wages paid to entry level or experienced employees. Id. The Administrator acknowledged that the 1997 CMSA-based Bremerton-Shelton wage determinations represented a "significant increase in the wage rates for the protective service occupations," and therefore reviewed additional BLS data to confirm that the 1997 CMSA-based wage determination figures were reasonable. Specifically, the Administrator looked to regional BLS wage surveys of protective service workers. In this case, the Administrator found that the BLS average weekly pay rates for Level I Firefighters and Police Officers in metropolitan areas in the western United States region ($882 and $854, respectively) were "remarkably consistent" with the corresponding Seattle CMSA pay rates of $918 and $854. Thus, BLS occupational wage data for protective service employees in the western portion of the United States also confirmed generally the appropriateness of the Seattle CMSA wage data for these occupations. Id. at 4.

   The Administrator discounted the data from the Washington State surveys of Kitsap, Clallam and Jefferson Counties because no information about the surveys' methodology had been provided and because it was unknown exactly what the data represented (e.g., mean versus median rates, whether rates were entry level, scope of the universe of employers surveyed, etc.). The Administrator nonetheless noted that even if the State survey were used, a comparison of the State survey wage rates with the Seattle CMSA wage rates did not support completely the Navy's challenge to the Bremerton-Shelton area wage determination, because the State survey wage rates were higher for some occupations than the CMSA data. Id.

   The Administrator ultimately concluded in the April 20, 1998 rulings that the arguments and documentation submitted by the Armed Forces did not provide a basis upon which to revise the 1997 CMSA-based wage determinations. The Administrator also emphasized that the capping methodology was properly utilized to limit the increases in wage rates for most occupations to no more than 15% as a means of easing the transition "to full integration of the CMSA data." Id. These appeals followed.

A. Introduction

   The Armed Forces offer a series of arguments challenging the Administrator's ruling upholding the 1997 CMSA-based Tacoma and Bremerton-Shelton wage determinations. In their petitions, the Armed Forces allege that the Administrator erred with regard to issues of law, fact, and policy when issuing the new wage determinations based on the Seattle CMSA data.

   Errors of Law The Petitioners assert that the wage determinations are incorrect legally on several grounds. First, the Armed Forces assert that the wage determinations

are incorrect because they are based on a "locality" that is inconsistent with the statute, regulations, case law and the Wage and Hour Division's internal procedures. Second, the Armed Forces argue that the 1997 wage determinations are defective because the 15% "capping" mechanism used by the Administrator is not explicitly authorized by statute or regulation. Third, the Petitioners contend that the Administrator's linking of the "due consideration" clause of the Service Contract Act with the locality pay adjustment provisions of the Federal Employees Pay Comparability Act when justifying his reliance on the Seattle CMSA wage survey data is not supported by law.

   Errors of Fact The Armed Forces also assert, based on various evidentiary submissions, that the wage rates in the 1997 CMSA-based Tacoma and Bremerton- Shelton wage determinations do not reflect accurately the lower level of wages actually paid in these communities when compared with the higher wage rates paid in Seattle itself (i.e., King County). Stated differently, the Armed Forces contend that the Administrator's published wage determination rates for the Tacoma and Bremerton-Shelton areas simply are out of line with wage rates actually being paid in the vicinity of the Armed Forces' installations.

   Errors of Policy In addition to raising challenges based on the law and facts, the Armed Forces cite the practical effects of the Administrator's decision to rely on CMSA wage survey data, arguing that Federal contracting agencies with service contracts in the Tacoma and Bremerton-Shelton areas (and beyond) will be adversely affected by the Administrator's change in policy. In this connection, the Petitioners argue that the Wage and Hour Division's decision to use the BLS's CMSA data, rather than to commission separate wage surveys for the smaller Tacoma and Bremerton-Shelton areas, is driven primarily by budgetary considerations at the Department of Labor, and that the Seattle CMSA wage survey does not truly reflect "local" wage rates.

   We consider each of the Armed Forces' arguments in our discussion below using this framework. In each instance, we review first the arguments raised by the Petitioners and Intervenors in challenging the April 20, 1998 rulings, then review the Administrator's defense of the rulings, and finally offer our analysis on each point. We review the Administrator's rulings to determine whether they are consistent with the statute and regulations, and are a reasonable exercise of the discretion delegated to the Administrator. See generally ITT Federal Services Corp. (II), ARB Case No. 95-042A (July 25, 1996), slip op. at 4-5 (according Administrator's reasonable interpretation of Section 4(c) of the SCA "great weight"); Service Employees Int'l Union (I), BSCA Case No. 92-01 (Aug. 28, 1992), slip op. at 7-14 (upholding Administrator's methodology for calculating nationally-prevailing fringe benefit rates as reasonable and within his discretion, while remanding on other grounds).

   Based on the record before us, we conclude that the Administrator acted within the limitations set by the SCA and its regulations when issuing the 1997 CMSA-based wage

determinations for the Tacoma and Bremerton-Shelton areas and denying the Armed Forces' requests for review and reconsideration.

B. Whether the challenged wage determinations for the Tacoma and Bremerton- Shelton areas are incorrect as a matter of law because 1) the Administrator relied on wage survey data encompassing the entire Seattle CMSA; 2) the Administrator adopted a 15% "capping" methodology; and 3) the Administrator relied on the "due consideration" clause of the Act to justify using CMSA-based wage survey data.

1. The Seattle Consolidated Metropolitan Statistical Area as a "locality" under the Service Contract Act.

   Armed Forces/Intervenors The Armed Forces contend that the Administrator's reliance on the Seattle CMSA data to develop the Tacoma and Bremerton-Shelton wage determinations that are at issue here is inconsistent with the statutory and regulatory provisions relating to "locality." Under the Service Contract Act, the Labor Department is responsible for issuing wage determinations that reflect the prevailing wage and fringe benefit rates paid to various classifications of service employees "in the locality." 41 U.S.C. §351(a)(1),(2). The regulatory guidelines for identifying a "locality" under the SCA are found at 29 C.F.R. §4.54(a).

   In the Armed Forces' view, the Administrator is responsible for determining wage rates under the Service Contract Act that "reflect, and not interfere with, local labor market conditions." Joint Pet. for Rev. at 2. Considering the large geographical area of the Seattle CMSA, the Armed Forces criticize the Administrator's use of the BLS survey because "[a]s a result, large areas of disparate communities, lumped together in CMSA surveys for purposes unrelated to SCA, now provide a database for average wages to be applied locally. The effect of this consolidation is contrary to the intent of SCA." Id. The Navy characterizes reliance on CMSA wage survey data as the improper use of "large portions of the United States as survey areas for specific localities." Navy Brief at 7-8. The Armed Forces cite Southern Packaging and Storage v. United States, 618 F.2d 1088 (4th Cir. 1980), and Descomp v. Sampson, 377 F. Supp. 254 (D. Del. 1974) as support for their contention that SCA wage determinations must provide wage rates that are tied to comparative economic conditions. Reply Brief at 8-9.

as a pertinent factor. Navy Brief at 8, Encl. 5. The Armed Forces assert that commuting patterns in and around the Kitsap peninsula (including Bremerton) support their claim that the Seattle CMSA is not an appropriate "locality" for purposes of assembling wage data under the SCA, because Puget Sound separates Kitsap and King Counties, adding additional time and expense to the commute between the two counties. Reply Brief at 17-18.

   Intervenors Johnson Controls and CSA advance similar arguments in support of the Armed Forces' challenge to the 1997 CMSA-based wage determinations. For example, Johnson Controls states that 97% of its workforce employed at the Navy's Bangor facility reside on the Kitsap Peninsula in Kitsap, Mason, Jefferson and Pierce Counties, with only 1% residing in the Seattle area (King and Snohomish Counties). Johnson Controls Ltr. Brief at 1. CSA contends that, unlike Federal employees, SCA contract employees are concentrated near the site of performance of current contracts and are not spread out over the Seattle CMSA. CSA Brief at unnumbered pp. 7, 9-10.

   Administrator The Administrator contests the arguments advanced by the Armed Forces and the Intervenors. With regard to the Southern Packaging and Descomp decisions cited by the Armed Forces, the Administrator observes that those cases are factually distinguishable from the three cases at issue here, and urges particularly that contrary to the Petitioners' position the opinion of the Court of Appeals in Southern Packaging actually supports the Administrator's position that the Seattle CMSA is an appropriate locality under the Service Contract Act. Admin. Brief at 19-22. Specifically, the Administrator relies on the appellate court's discussion of "locality" in its Southern Packaging opinion, 618 F.2d at 1092, as support for the proposition that "locality" must not be interpreted in a restrictive manner. Admin. Brief at 19-20, 21-22.

   As further support for his contention that "locality" must not be narrowly construed, the Administrator cites statements from the SCA legislative history on this issue. The Administrator notes that the undefined term "locality" in the SCA, which had been interpreted flexibly under the Walsh-Healey Act, 41 U.S.C. §35 et seq., was chosen by Congress in lieu of the more rigid definition of the Davis-Bacon Act (40 U.S.C. §276a) requiring that prevailing wage rates be determined based on wages paid in the "city, town, village or any other civil subdivision of the State in which the contract work is to be performed." Admin. Brief at 21 (citing Hearings, Subcommittee on Labor, Committee on Labor and Public Welfare, 89th Cong., 1st Sess., on H.R. 10238, Sept. 23, 1965, pp.12-14). The Administrator also cites the regulatory guidelines for identifying a locality under the SCA, which are found at 29 C.F.R. §4.54. The Administrator asserts that the 1997 CMSA-based wage determinations "define locality the same as the geographic area included within the scope of the BLS survey" and are thus in compliance with Section 4.54. Admin. Brief at 22.

   In response to the Armed Forces' contention that use of the Seattle CMSA wage survey data disregards the commuting distances of employees who work on SCA contracts, the

of service contracts that are subject to SCA coverage. 29 C.F.R. §4.54(a), quoted supra. We share the Administrator's view that the guidelines provided by Section 4.54(a) are consistent with the SCA's legislative history, which indicates that the undefined term "locality" intentionally was chosen instead of the more rigid "city, town, village" language contained in the Davis-Bacon Act. See Hearings, Subcommittee on Labor, Committee on Labor and Public Welfare, 89th Cong., 1st Sess., on H.R. 10238, Sept. 23, 1965, p.11 (Statement of Charles Donahue, Solicitor of Labor); see also Descomp, 377 F. Supp. at 265. On the following basis, we also agree with the Administrator that the Descomp and Southern Packaging decisions do not support the Petitioners' claim that designating the Seattle CMSA as a "locality" is inconsistent with the statute.

   The Descomp and Southern Packaging courts, respectively, rejected the application of wage determination rates issued for the Washington, D.C., area to prospective contractors who intended to perform their service contracts in other locations. In Descomp, the place of contract performance was Wilmington, Delaware; the contract under consideration in Southern Packaging designated the place of performance as the entire continental United States (i.e., the services could be performed in any one of various locations around the country). The common thread running through those decisions is that the wage determination applicable to a contract must represent wages prevailing in the locality where the contract is to be performed, and could not simply be based on Washington, D.C., wage rates if the place of performance was outside the Washington area. Southern Packaging, 618 F.2d at 1091-92, 458 F. Supp. at 732-35; Descomp, 377 F. Supp. at 266.

   However, both the Descomp and Southern Packaging decisions clearly provide support for the proposition that "locality" must be given a flexible construction under the SCA. See Southern Packaging, 618 F.2d at 1091-92, 458 F. Supp. at 732-35; Descomp, 377 F. Supp. at 264-66. For example, the District Court in Southern Packaging stated that "the term 'locality' is indefinite and lends itself to varied interpretations . . . ." 458 F. Supp. at 733. That court also noted that the Senate Report that accompanied the 1965 bill that was enacted into law stated that "[t]he Secretary in determining the locality . . . would take a realistic view of the type of service contract intended to be covered by the determination." 458 F. Supp. at 735 (quoting from S. Rep. No. 798, 89th Cong., 1st Sess. 2 (1965), reprinted in 1965 U.S.C.C.A.N. 3737, 3738).

   It is noteworthy that the district court in Southern Packaging endorsed the Administrator's use of the Standard Metropolitan Statistical Area (SMSA) as the preferred "locality" unit for issuing wage determinations under the SCA, based on the testimony of the Director of the Department's SCA wage determination office regarding the practice of that office at that time. 458 F. Supp. at 730, 733-34. When Southern Packaging was before the district court in 1978, the SMSA was, like the CMSA is in the instant case, a multi-jurisdictional unit consisting of an urban core surrounded by adjacent, economically integrated areas, established by OMB for collecting and evaluating data by various Federal agencies. See Notices, OMB, 55 Fed. Reg. 12154 (1990).

   In 1978, OMB standards for defining statistical areas already were being revised to reflect the increasing size of metropolitan areas and changing commuting patterns. By the time new OMB standards were promulgated in connection with the 1980 decennial census, OMB had abandoned the SMSA in favor of the general formulation that is in effect today, i.e., recognition of different classes of Metropolitan Statistical Areas (Levels A - D), with Level A MSAs (over 1 million population) eligible to be designated as Consolidated MSAs if they include two or more subareas that independently would qualify for MSA status (PMSAs). See discussion supra at pp. 4-6. See generally 45 Fed. Reg. 956 (1980) (1980's MSA standards); 55 Fed. Reg. 12154 (1990) (1990's MSA standards). And the notion of "metropolitan area" continues to evolve. In a recent notice announcing further modifications to the definitions of metropolitan and non-metropolitan areas issued in anticipation of the 2000 decennial census, OMB noted that commuting patterns continue to be "the most reliable and broadly available measure of functional integration" of communities, but observed that

[t]he spatial patterns of commuting are more complex today than in previous decades, but no less important. The spatial structure of the urban environment is less consistently monocentric than was the case in the early part of the twentieth century . . . . commuting patterns are less likely to resemble a hub-and-spoke model than a polycentric structure of multiple employment nodes serving a region's needs.

Notices, OMB, 63 Fed. Reg. 70526, 70534 (1998).

   The Armed Forces have submitted two types of non-wage data in support of their position that the Tacoma and Bremerton-Shelton areas are economic environments distinct from the Seattle/King County core of the Seattle CMSA. First, the Navy asserts that of the more than 13,000 Federal civilian and contractor employees working at Navy installations in Kitsap County, very few only 2% commute from King County (Seattle). AR, Tab E, Exh. D. However, the number of workers commuting from King County to the Kitsap peninsula is only a partial measure of the economic integration of the Puget Sound region, failing to consider the number of Kitsap County residents who commute in the other direction i.e., out of the county to find employment. A more complete picture of the region is provided by data compiled by the Census Bureau from the 1990 census, and we take notice of this data. See United States v. United Bhd. of Carpenters & Joiners of America, Local 169, 457 F.2d 210, 214 n.7 (7th Cir. 1972). Whatever the geographic barriers of the Puget Sound region, the Census Bureau's home-to-work commuting data demonstrate that there is a high level of economic interaction among the counties, with large numbers of workers moving across county lines as part of their daily commute:

______________________________________________________________________________________________________
                              | Total     |          COUNTY OF WORK                      |
Seattle-Tacoma, WA CMSA       | living in | ____________________________________________ |
Bremerton MSA                 | county of |  King   Snohomish  Pierce   Kitsap  Thurston |   Worked
Olympia MSA                   | residence |  County County     County   County  County   |   elsewhere
______________________________________________________________________________________________________

Total Working in county of      . . . .     912.610  179,864   233,197   83,826   68,923     . . . .
work

COUNTY OF RESIDENCE

King County                     805,782     750,970   28,328    14,452    1,673    1,304     9,055

Snohomish County                231,967      84,722  141,802       368      339      108     4,628  

Pierce County                   270,589      53,657    2,526   203,626    3,626    3,750     3,404

Kitsap County                    88,144       8,459      530     2,960   74,323      109     1,763

Thurston County                  75,364       2,064      172     8,526      108   59,069     5,353

Living elsewhere                 . . . .     12,738    6,506     3,265    3,685    4,583     . . . . 

Source: U.S. Bureau of the Census, "Commuting Flows of American Workers Charted by New 1990 Census Computer File." Press Release No. CB92-267, December 22, 1992, Table 14, derived from computer file STF- S-5, Census of Population 1990: Number of Workers by County of Residence by County of Work.

    Although the Census Bureau data confirm generally the Navy's claim that only 2% of the workers

employed in Kitsap County commute across (or around) Puget Sound from King County, fully 9,053 of Kitsap's 83,826 workers (11.3% of the total) commute into Kitsap County from the other counties in the region. Perhaps more important as a measure of the economic integration of the region is the commute in the opposite direction, with 9.2% of the King County (Seattle) workforce traveling into the urban core from Kitsap County 8,459 workers.

   The second type of non-wage data submitted by the Armed Forces are statistics comparing the sales prices of residential real estate in King County (Seattle) with prices in surrounding counties, including Pierce County (Tacoma) and Kitsap County (Bremerton-Shelton). Air Force Brief, Encl. 4; Navy Brief at 12, Encl. 8; AR, Tab FF. The Armed Forces cite the lower real estate selling prices in the Tacoma and Bremerton-Shelton areas in support of their argument that these outlying areas are economically distinct from the Seattle/King County core of the CMSA. Navy Brief at 12; Reply Brief at 5.

   Similar to the commuting data submitted by the Navy, we do not find these real estate statistics persuasive in view of the degree of economic integration that is demonstrated by OMB's designation of the Seattle- Tacoma-Bremerton CMSA. As already discussed, the criteria for determining whether a Metropolitan Statistical Area qualifies as a CMSA cover a range of factors relevant to population density and housing and work patterns. 55 Fed. Reg. at 12154-56. The CMSA designation provides a comprehensive gauge of the economic integration of the Seattle- Tacoma-Bremerton area that is not overcome by the real estate statistics submitted by the Armed Forces.

   We therefore find the Administrator's reliance on BLS wage survey data from the Seattle CMSA when developing the challenged Tacoma and Bremerton-Shelton wage determinations to be a reasonable exercise of his discretion under the Act and its implementing regulations. It is supported by government data documenting the work- home relationships within the Puget Sound labor market, and also by OMB's criteria for designating CMSA units. The wage data provided by the Armed Forces, which the Administrator dismissed as anecdotal or flawed (and which we discuss below), does not persuade us to the contrary.

2. The Administrator's 15% "capping" methodology.

   Armed Forces/Intervenors The Armed Forces challenge the Wage and Hour Division's use of the capping methodology on various grounds.

   As discussed previously (at pp. 8-10), when shifting from the earlier BLS survey data based on the separate smaller communities (i.e., Seattle, Tacoma, Bremerton) to the BLS's Seattle CMSA wage data, the Administrator did not simply issue a single new set of wage determinations that would apply to the entire region. Recognizing that immediate adoption of the CMSA data would produce dramatic increases in the wage determination rates for some service occupations in certain areas, the Administrator instead adopted a transitional strategy, continuing to issue separate wage determinations for the Seattle, Tacoma and Bremerton-Shelton subregions. For occupations where the new CMSA data resulted in only a modest wage increase in wage rates when compared with the predecessor wage determinations, the CMSA data was incorporated directly into the new wage determinations. However, for occupations (other than protective service classifications) where the CMSA data would have resulted in a substantial increase in wage rates, the increase was capped at 15% over the wage rates in the predecessor wage schedules.

   The Armed Forces note that there is no provision for the 15% capping mechanism in the Service Contract Act or its implementing regulations, and that the capping methodology allows for increases above the 15% cap to be phased in through additional 15% increments in future wage determination revisions. Reply Brief at 16-17. In view of the anticipated phase-in of higher wage rates in future wage determination revisions, the Navy and the Air Force object to the Administrator's reliance on capped wage rates when comparing the wage rates of the challenged Bremerton- Shelton and Tacoma wage determinations with other data. Navy Brief at 6-7; Air Force Brief at 3-6. In support of their objections, the Navy and the Air Force, respectively, analyzed the number and extent of increased wage rates that would have resulted on the Bremerton-Shelton and Tacoma wage determinations had the 15% cap not been applied. Navy Brief, Encl. 4; Air Force Brief, Encl. 2; AR, Tabs F, BB. The Armed Forces also urge that use of the capping methodology by the Wage and Hour Division is a tacit admission by the Division that the CMSA wage data did not provide a sound foundation for the challenged wage determinations. Reply Brief at 17.

   Administrator The Administrator denies that the capping mechanism constitutes any acknowledgment that using the Seattle CMSA survey data was inappropriate. The Administrator counters instead that using the new CMSA wage data, while simultaneously applying the 15% cap to any increases over the predecessor Tacoma and Bremerton-Shelton wage determinations, was an appropriate transitional mechanism that recognized the validity of both the new CMSA data and the older subregional data that formerly was used for issuing wage determinations. Admin. Brief at 25-26; see AR, Tab E.

   Analysis The Wage and Hour Division's capping methodology is not explicitly authorized by the SCA or its implementing regulations, and neither the Administrator's determination letters to the Armed Forces nor the Administrator's Statement to the Board offers an extended discussion of the origin of the 15% cap and its basis. Cf. Admin. Brief at 25-26; AR, Tabs A-C. Nonetheless, we find that the Administrator's general application of a 15% cap on wage rate increases in the 1997 CMSA-based wage determinations was a reasonable exercise of the Administrator's broad discretion to issue wage determinations under the Act, and we uphold this practice on the facts before us in this case.

   The Administrator's discretion under the Service Contract Act is perhaps at its broadest when the Administrator is issuing prevailing wage schedules. The statute requires, in relevant part, that all Federal service contracts include "[a] provision specifying the minimum monetary wages to be paid various classes of service employees . . . as determined by the Secretary [of Labor] . . . in accordance with prevailing rates for such employees in the locality[.]" 41 U.S.C. §351(a)(1). Like its sister statute, the Davis-Bacon Act, nowhere does the SCA prescribe a specific methodology to be used by the Secretary or her designee, the Administrator, when determining the prevailing wage. Perhaps the clearest indicator of the very great deference owed to the Secretary and the Administrator when determining prevailing wage rates is the clear body of case law holding that the substantive correctness of wage determinations is not subject to judicial review. United States v. Binghamton Construction Co., 347 U.S. 171, 177 (1954) (under the Davis-Bacon Act); Commonwealth of Virginia v. Marshall, 599 F.2d 588, 592 (4th Cir. 1979) (under the Davis-Bacon Act); AFGE v. Donovan, 25 Wage & Hour Cas. (BNA) 500, 1982 WL 2167 at *2 (D.D.C. 1982), aff'd 694 F.2d 280 (D.C. Cir. 1982) (table) (under the Service Contract Act). Judicial review "is limited to due process claims and claims of noncompliance with statutory directives or applicable regulations." Commonwealth of Virginia at 592 (citing Califano v. Sanders, 430 U.S. 99, 109 (1977)).

   Although the 15% cap may not be explicitly authorized under the Act, neither is it prohibited. The Board previously considered the Administrator's similar application of a 15% cap on a wage rate increase in the D.B. Clark III case, ARB Case No. 98-106 (Sept. 8, 1998). While noting in Clark that the 15% cap policy apparently was not memorialized either in the SCA regulations or the Wage and Hour Division's internal operating procedures, we nevertheless approved the capping methodology based on the record that was before us in the case, finding that the approach was reasonable and within the Administrator's broad discretion to devise "program guidelines that are administrable and produce consistent results." Clark, slip op. at 8.

   We reach the same conclusion in this case. Once the Administrator decided to shift to the BLS's Seattle CMSA wage survey as the data source for wage determinations in the Puget Sound region, the Administrator concluded that an abrupt shift from the "old" subregional data to the "new" CMSA data would produce comparatively dramatic shifts in wage rates for some occupations. Instead of implementing the change to the Seattle CMSA data cold turkey, the Administrator chose to temper the impact that would be caused by the change to the new survey instrument by limiting the maximum 1-year change in wage rates to 15%. Recognizing that both the lower "old" and the higher "new" survey results are valid if different expressions of the wage rates prevailing "in the locality," it is our conclusion in this case that the Administrator's 15% cap methodology is a reasonable transitional tool that is consistent with the Administrator's discretion under the Act, and also promotes a useful predictability in the procurement process.

3. The link between the Service Contract Act's "due consideration" clause and the use of CMSA data when computing Federal employee locality pay differentials.

   Armed Forces/Intervenors In the determination letters to each of the Armed Forces, the Administrator noted that the BLS's Seattle CMSA wage data also is used for computing the locality pay adjustments for Federal employees under the Federal Employees Pay Comparability Act. In the Administrator's view, the fact that the BLS data is used for Federal comparability pay purposes provides further justification for using the same data as the basis for SCA wage determination rates, because of the Service Contract Act's Section 2(a)(5) requirement that "[t]he Secretary shall give due consideration to" the wages paid to Federal employees in comparable job classifications when issuing SCA wage determinations. 41 U.S.C. §351(a)(5).

   The Armed Forces contest the Administrator's reliance on the "due consideration" clause of Section 2(a)(5) to justify using the Seattle CMSA wage survey data, asserting that the goals of the SCA and of Federal employee pay legislation are different. Reply Brief at 10-12. The Petitioners point out that the wage standards for white collar Federal employees (i.e., the "GS" wage schedule) establish uniform pay rates for work at similar grade levels, and that these pay levels are largely uniform throughout the country. In contrast, the Service Contract Act requires that SCA wage rates reflect locally prevailing wages. Id. The Petitioners acknowledge, however, that locality pay adjustments for Federal employees, which have been extended since their inception to cover a significant number of metropolitan areas, have diluted the effect of the Federal pay schedules that apply nationally. Id.

   Administrator Before the Board, the Administrator reiterates his argument that reliance on the CMSA data is justified under the "due consideration" provision of Section 2(a)(5) of the SCA, because the same BLS survey is used for computing locality pay adjustments for Federal employees in the Seattle area.

See Notices, Dept. of Labor, BLS, 57 Fed. Reg. 30982, 30983 (1992). See generally 59 Fed. Reg. 62549 (1994) (Presidential memorandum for the Secretary of Labor, the Director of OMB and the Director of the Office of Personnel Management regarding their joint report containing locality pay recommendations); 5 U.S.C. §§5301-07 (1994). In addition, the Administrator points to the observation found in the SCA regulations at 29 C.F.R. §4.51(d) that the purpose of the "due consideration" provision is to narrow the gap between the wage rates paid to Federal employees and those paid to service employees. Admin. Brief at 16-17.

   Analysis Section 2(a)(5) provides that "[t]he Secretary shall give due consideration" to the rates "that would be paid by the Federal agency to the various classes of service employees if section 5341 or section 5332 of title 5, United States Code, were applicable" in setting the minimum wage rates payable to employees covered by the Act. 41 U.S.C. §351(a)(5). The plain language of Section 2(a)(5) thus indicates that Congress intended that SCA wage rates be linked in some way to Federal employee pay rates; that intention is also reflected in the legislative history of the 1972 SCA amendments, which demonstrates concern that the gap between Federal employee salaries and those paid to SCA-covered employees be narrowed. See 118 Cong. Rec. 27136 (Statement of Rep. Thompson), 31282 (Statement of Sen. Gurney) (1972); see also AFGE v. Donovan, supra (examining Section 2(a)(5) legislative history in rejecting argument that "due consideration" provision required SCA wage rates to be no less than Federal Wage Board prevailing rates). Section 4.51(d) of the Title 29 regulations implements the Section 2(a)(5) due consideration provision by requiring the Wage and Hour Division to consider "those wage rates and fringe benefits which would be paid under Federal pay systems" when calculating the SCA wage rates. 29 C.F.R. §4.51(d).

   In these cases, the Administrator does not claim that the wage rates in the challenged SCA wage determinations reflect a direct comparison with wage rates paid to specific comparable classifications of Federal employees. Instead, the Administrator argues that because the geographic area used for computing Federal employee "comparability pay" i.e., the Seattle CMSA is the same geographic area used for the BLS survey that is the basis of the challenged wage determinations, the Section 2(a)(5) "due consideration" clause of the Service Contract Act lends legal validity to the Administrator's conclusion that the Seattle CMSA is an appropriate locality under the SCA.

   On its face, Section 2(a)(5) does not address the use of local pay areas established under Federal employee pay systems as the localities for which prevailing wages will be set under the SCA. To determine whether the Administrator's argument holds, we turn to the Act's legislative history for guidance. See Fort Hood Barbers Assoc. v. Herman, 137 F.3d 302, 307-08 (5th Cir. 1998). See generally AFL-CIO v. Donovan, 757 F.2d 330, 344 (D.C. Cir. 1985) (holding that arguments regarding general congressional intent "cannot overturn the clear language of a specific provision"). An examination of the intention of the lawmakers who incorporated the references to Sections 5332 and 5341 into Section 2(a)(5) must focus on the substance and application of those provisions at the time of the 1972 and 1976 amendments. See Merrill Lynch, Pierce, Fenner & Smith v. Curran, 456 U.S. 353, 378-79 (1982), and cases there cited. Our examination reveals that the pertinent SCA legislative history and the antecedent Federal pay systems that are referenced by Section 2(a)(5) do not support the Administrator's contention.

   Section 2(a)(5) was added to the Act in 1972 and amended in 1976. Pub. L. No. 92-473, §2, 86 Stat. 789 (1972); Pub. L. No. 94-489, §2, 90 Stat. 2358 (1976). The 1972 amendment directed

• Collectively bargained wage rates compiled by the Navy after surveying several Kitsap County employers showing that wages in the Bremerton-Shelton area are lower than rates for many classifications in the 1997 CMSA-based Seattle and Bremerton-Shelton wage determinations, and a statement by Johnson Controls accompanied by copies of collective bargaining agreements that its collectively bargained wage rates are based on wage levels in the Kitsap County area rather than the "higher scales" paid in the Seattle area. Navy Brief at 10-11, Encl. 6, 7; AR, Tab E, Exh. B; Tabs DD, EE.

• Copies of 1996 Washington State Employment Security wage surveys for Kitsap County and for Clallam and Jefferson Counties, and a comparison of those rates with specific wage rates provided by the 1997 CMSA-based Seattle wage determination. Navy Brief at 15; Reply Brief at 14-15; AR, Tab E, Exh. A; Tab GG.

• Comparisons of wage rates from the 1997 CMSA-based Tacoma and Bremerton-Shelton wage determinations with wage rates from the predecessor wage determinations for the same locations. Air Force Brief at 5-6, Encl. 2; Navy Brief, Encl. 2; AR, Tab E, Exh. A; Tab F, Attachment 1.

• Comparisons of 1997 CMSA-based Seattle wage determination wage rates with rates from predecessor wage determinations for Tacoma and Bremerton-Shelton, Air Force Brief at 4-5, Encl. 2; Navy Brief at 6-7, Encl. 2; AR, Tab E, Exh. A; Tab F, Attachment 1. The Navy summarizes the comparative data for Bremerton-Shelton as demonstrating that, when wage rates are ultimately "uncapped," wage rates for 84% of the 294 classifications will increase by more than 25% (31% of the classifications would increase by up to 25%, 44% of the classifications by between 25% and 49%, and 9% of the classifications by more than 50%). Navy Brief at 6-7, Encl. 4. The Air Force characterizes the comparative data for Tacoma as demonstrating a "substantial inflation" of wage rates resulting from application of the Seattle CMSA wage data. Air Force Brief at 4-5.

• Federal wage grade equivalents and wage rates for specific classifications included in the 1997 CMSA-based Seattle wage determination. Navy Brief, Encl. 2; AR, Tab E, Exh. A.

• Comparison of wage rates for Tacoma with rates for Seattle, taken from the respective wage determinations for these areas issued prior to the CMSA-based wage determinations in 1997. Air Force Brief, Encl. 2; AR, Tab F, Attachment 2. The Air Force urges that this comparative data demonstrates a wage structure in Tacoma that is based on "substantially lower wage rates" than those in Seattle. Air Force Brief at 5-6.

    In addition, the Air Force notes the failure of the Administrator to specifically address the evidence and argument that the Air Force presented to the Administrator in its initial challenge to the 1997 CMSA-based Tacoma wage determination. Air Force Brief at 2. The Air Force also questions the Administrator's rationale for rejecting the data submitted below by the Navy, particularly the Washington State wage survey data. Id. at 2-4.

   In sum, the Armed Forces argue that the 1997 CMSA-based wage determinations do not reflect the wage rates in the communities where the Petitioners' SCA contracts will be performed, i.e., the Tacoma and Bremerton-Shelton areas, but instead more closely reflect wages prevailing in Seattle/King County. Joint Pet. for Rev. at 2.

   Administrator The Administrator discounts the wage rate data submitted by the Petitioners, arguing that the Armed Force's statistics either are of questionable reliability or are misleading.

   The Administrator criticizes the Washington State Employment Security Department wage survey data as lacking reliability because, unlike the BLS survey data (which is gathered through on-site visits by BLS field economists), the State survey depends on employers to classify their employees and then submit their responses by mail. The Administrator also notes the relatively high non-response rate among employers participating in the Washington State surveys, i.e., 48% on the Kitsap County survey and 43% on the Jefferson/Clallam Counties survey; and that this high non-response rate contrasts with the 16% non-response rate for the BLS-conducted Seattle CMSA survey. Admin. Brief at 23-24; see AR, Tabs GG, W. The Administrator also notes that because the Washington State survey data does not distinguish between rates paid to different levels within employee classifications, it is "impossible to make valid comparisons between the State and BLS data." Admin. Brief at 24.

   Similarly, the Administrator questions the usefulness of the Navy's survey of police and fire departments, because the documentation (a) does not specify whether the wage rates represent an average wage rate or an entry level wage rate for each profession, (b) does not indicate how many employees are actually employed under the collective bargaining agreements covering Kitsap area protective service personnel, and (c) does not provide job descriptions for the classifications listed in the agreements. Id.; see AR, Tab E, Exhs. B - C-2; Tab DD. The Administrator further contends that even if the Navy's data were complete and reliable, this evidence does not demonstrate that rates based on the Seattle CMSA data are excessive because some of the Navy's survey wage rates are higher than corresponding wage rates in the challenged wage determinations. Admin. Brief at 24; see AR, Tab DD.

   The Administrator also disputes the Armed Forces' reliance on a comparison of predecessor wage determination rates with the rates in the 1997 CMSA-based wage determinations. The wage rates in the predecessor wage determinations although issued within a matter of weeks prior to the issuance of the CMSA-based wage determinations were based on BLS data from surveys conducted in 1993-94. Admin. Brief at 25; see AR, Tab E, Exhs. B-C-2; Tab F, Attachments 1, 2. The Administrator urges that because there had been no interim increases in the wage determinations issued between 1994 and 1997, it is not possible to draw any useful conclusions about the relative economic conditions in different parts of the Puget Sound region by comparing these wage rates based on "old" data with the rates in the 1997 Tacoma and Bremerton-Shelton wage determinations that are based on the Seattle CMSA data. Admin. Brief at 25; see AR, Tabs N-Q.

   The Administrator also takes issue with the comparative data submitted by the Armed Forces that relies on "uncapped" wage rates, i.e., wage rates taken directly from the 1997 Seattle CMSA-based wage determination which, in many cases, were not fully implemented in the 1997 CMSA-based wage determinations for Tacoma and Bremerton-Shelton because of the capping methodology. Admin. Brief at 25. The Administrator urges that the Armed Forces' use of the uncapped Seattle wage rates distorts the effect of the Administrator's reliance on the CMSA wage data for the challenged Tacoma and Bremerton-Shelton wage determinations. Id.

   Analysis At the outset of our consideration of the Armed Forces' wage data, it is useful to acknowledge first the overriding significance of our earlier discussion of "locality." Having found that the sprawling Seattle CMSA is an acceptable "locality" for developing SCA prevailing wage rates, based on OMB's criteria for designating metropolitan areas (supra), it follows that a comparison of wage rates between different subareas of the Puget Sound region adds little to the discussion. Stated differently, once we conclude that a geographic area manifests sufficient economic integration that the Administrator reasonably may deem it a "locality" under the Act, it is largely irrelevant that wage rates may be higher in some parts of the community than in others. For example, wages for clerical employees working at downtown offices in some cities may be higher than the rates paid for comparable positions in outlying suburban areas, but by itself the disparity in wage rates would not mean that the downtown and suburban locations are different "localities" for SCA purposes. Labor market and commuting patterns are far more meaningful in this regard.

   The regulatory scheme under which SCA wage determinations are developed directs the Administrator to exercise discretion when determining the specific methodology to be employed in calculating particular wage rates. The Board "will upset a decision of the Administrator only when the Administrator fails to articulate a reasonable basis for the decision, taking into account the applicable law and the facts of the case." Court Security Officers [of Austin, Texas], ARB Case No. 98-001 (Sept. 23, 1998), slip op. at 4, aff'd sub nom. Richison v. Herman, No. W-97-CA-385 (W.D. Tex. Feb. 1, 1999); see also D. B. Clark III, slip op. at 6. Thus, the central question on appeal in these consolidated cases is not whether a different methodology from the one chosen by the Administrator might have been more reasonable, but simply whether the Administrator's chosen methodology is consistent with the law and the facts before us. See COBRO Corp., ARB Case No. 97-104 (July 30, 1999), slip op. at 23. The quality of the evidence in the record can be a significant consideration in determining whether to uphold the Administrator; as the Deputy Secretary noted in Tri-States Service Co., an analogous case involving a challenge to SCA wage determination rates, "the basic issue to be decided is whether the wage information supplied by Petitioner represents more accurate and probative evidence of the prevailing wages in the locality than the data and methods utilized by the Wage and Hour Division." Case No. 85- SCA-WD-12, Dep. Sec. Dec. (Sept. 18, 1990), slip op. at 5.

   Although we recognize the substantial effort expended by the Armed Forces in developing their documentary exhibits, we agree with the Administrator's conclusion that none of the exhibits demonstrates that the Administrator erred in relying on the BLS's Seattle Occupational Compensation Survey, and thereby using data from the entire Seattle Consolidated Metropolitan Statistical Area. Based on OMB's regulatory criteria for designation of statistical areas, reinforced with commuting statistics developed by the Census Bureau, we found above that the CMSA is an acceptable "locality" under the Service Contract Act.

   Neither the Armed Forces nor the Intervenors have challenged the underlying methodology or accuracy of the BLS's Occupational Compensation Survey, which is an extensive cross-industry survey of employers using sophisticated statistical techniques. See AR, Tab W Appendix. For all the reasons cited by the Administrator, supra, we are persuaded that the wage data submitted by the Armed Forces is less complete and less reliable than the BLS-conducted Seattle CMSA survey, and that the Administrator's decision not to reconsider the wage rates in the 1997 Tacoma and Bremerton-Shelton wage determinations therefore was reasonable.

D. Whether the Administrator's shift to the Seattle CMSA survey data as the basis for the challenged 1997 wage determinations reflects an error in policy.

   In addition to challenging the Administrator's conclusion that the CMSA is an appropriate "locality" under the SCA, and submitting evidence in support of their claim that the 1997 Tacoma and Bremerton-Shelton wage determination rates are inconsistent with wages actually paid in these communities, the Armed Forces and the Intervenors also criticize the Administrator's decision to use the BLS Seattle CMSA data as a poor policy choice. This criticism is approached from several directions.

   Citing the tremendous volume of SCA contracts that are entered into by the Department of Defense (DOD), the Armed Forces point out that they are a major constituent of the SCA wage determination process. They note various cooperative efforts between the Wage and Hour Division and the DOD in recent years that resulted in a streamlined process by which blanket requests for wage determinations are submitted to the Department; additional progress has been made in developing computerized access to wage determinations. Reply Brief at 20. The Armed Forces complain that DOD had expected that the cost savings achieved by the Wage and Hour Division through these streamlined procedures would be invested "in developing appropriate SCA WDs." Id.

   In addition, the Armed Forces disparage the Administrator's shift to the regional BLS data by arguing that the decision is more motivated by Department of Labor fiscal concerns than by concerns regarding effective implementation of the Service Contract Act. They characterize the shift to Seattle CMSA data as "budget- friendly" from the standpoint of the Wage and Hour Division, but inadequate for developing the "locality"-based wage determinations required under the Act. Air Force Brief at 11; see Joint Pet. for Rev. at 2, 6; Reply Brief at 10. Intervenor CSA cites the significant fiscal impact that an increase in SCA contract wages will have on Federal procurement costs, and urges that "concern with creating an efficient survey process should be balanced with its resulting impact on the Federal government as a whole." CSA Brief at unnumbered p. 9.

   The Armed Forces contend further that the shift to using a new BLS survey instrument as the foundation for the wage determination program makes the resulting wage schedules legally suspect. In their view, the Administrator has chosen to use the CMSA survey data simply because that data is readily available; they posit that the use of "perpetually changing methodologies" by the Wage and Hour Division has resulted in a wage determination process that is "so arbitrary and inconsistent as to be without any credible foundation." Reply Brief at 13, 15, 21.

   The Armed Forces also suggest that the shift to the CMSA-based data may lead to undesirable swings in SCA wage rates, both higher and lower. Although the Armed Forces specifically challenge the 1997 CMSA- based Seattle wage determinations in this case because, in their view, wage rates have increased too much, they also warn that the shift to the CMSA-based data could result in future decreases in wage rates. They observe that "the extensive combination of large, disparate communities into one single wage survey effort" could result in a significant decline in wages, which would have an equally disruptive effect on the SCA labor market. Joint Pet. for Rev. at 2; see Air Force Brief at 9.

   Analysis In Service Contract Act cases, the Administrative Review Board's role is to provide appellate review of the Administrator's decisions, having "jurisdiction to hear and decide . . . appeals concerning questions of law and fact from final decisions of the Administrator of the Wage and Hour Division or authorized representative . . . ." 29 C.F.R. §8.1(b). Although the Board is "delegated authority and assigned responsibility to act for the Secretary of Labor in issuing final agency decisions" in matters arising under the Service Contract Act (Secretary's Order 2-96, 61 Fed. Reg. 19982 (1996)), it is the Administrator, not the Board, who has primary responsibility for implementing and enforcing the SCA. To that end, the Board and its predecessor agencies extend broad deference to the Administrator's interpretations of the Act and its implementing regulations, so long as the Administrator's policies and determinations are legally sound and otherwise reasonable. For example, in a 1996 case involving conflicting interpretations of Section 4(c) of the Act, we observed that:

[The] Petitioners contend that there are good policy reasons for interpreting the statute differently [from the Administrator].... The Board of Service Contract Appeals has previously recognized on numerous occasions that the Administrator is granted broad discretion in interpreting the SCA. Service Employees International Union, AFL- CIO, CLC, BSCA Case No. 92-01, Aug. 28, 1992. The express language of Section 4(c) [at issue in this case] certainly does not dictate an interpretation different from the Administrator's. The Administrator's reading of this provision is reasonable and not a departure from accepted canons of construction. Therefore, the Administrator's interpretation is accorded great weight. The Board should not substitute its own policy preferences for those of the official in whom primary responsibility for enforcing the statute is vested. See A. Vento Construction, WAB Case No. 87-51, Oct. 17, 1990 and Titan IV Mobile Service Tower, WAB Case No. 89-14, May 10, 1991.

ITT Federal Services Corp (II), ARB Case No. 95-042A (July 25, 1996), 1996 WL 415926 at *3, 4. Thus, our inquiry on review is focused simply on whether the Administrator's decision reflects a reasonable interpretation of the statute and regulations, not whether we believe it to be the best policy choice.

   Although the basic concept behind the Service Contract Act i.e., that employees on Federal service contracts should not be paid less than the locally-prevailing wage and fringe benefit rates is straightforward, the implementation of the statute is complex and raises many difficult questions. Fundamental concepts of "locality" and "prevailing" are critical threshold issues in wage determination matters, but they are followed by a host of equally challenging problems such as competing methodologies for collecting and analyzing wage data. In many of these situations requiring interpretation of the statute or its regulations, there is no single "right" or "obvious" answer to these questions. Instead, the Administrator must choose from a variety of options while trying to reconcile several interests: the statutory mandate that local labor standards be protected; the need to establish predictable and enforceable policies; the goal of promoting stability in the Federal procurement system; and the obligation to be an effective steward of the resources provided by Congress for implementing the statute, using them as efficiently as possible. It is not an easy job.

   The Armed Forces and Intervenors raise important concerns in challenging the Administrator's CMSA-based wage determinations. The Administrator's shift to wage determinations based on CMSA data results in significant pay increases for workers in the outlying counties of the Seattle metropolitan area. Furthermore, as the Armed Forces point out, the geographical area encompassed within the Seattle CMSA is very large and the "old" urban centers that formed the basis for the predecessor wage determinations continue to be urban focal points that could be used to support a series of wage determinations based on the smaller PMSAs, rather than the sprawling CMSA if such subregional wage data were available. Indeed, we acknowledge that the very same analysis of OMB's standards for determining metropolitan areas that we adopted earlier in finding the Seattle CMSA to be an acceptable "locality" under the SCA (supra at 17-24) could be used to support a finding that the smaller PMSA units also would be acceptable "localities" under the Act. The Armed Forces' policy argument that these smaller geographical units are preferable "localities" for SCA purposes is not without substance.

   The Board's limited mandate, however, is to determine whether the Administrator's determination is consistent with the law and reasonable, and not to weigh whether the Administrator's approach is "better" than the one advocated by the Petitioners. Almost any change in wage data methodologies will advantage some groups and disadvantage others, but this is precisely the kind of judgment call that is uniquely within the Administrator's power. In this instance, merging all the regional wage data into a single survey tends to increase SCA wage rates in the Tacoma and Bremerton-Shelton areas (where wage rates historically have been lower than in King County); as a corollary, it would appear that this same process ultimately will result in lower SCA wage rates in the Seattle urban core, compared with the rates in the predecessor Seattle wage determinations. Today, the Army, Navy and Air Force assert that the new wage rates in Tacoma and Bremerton-Shelton are too high, from the perspective of the contracting agencies and contractors performing work at facilities in these outlying areas; tomorrow, service employees working in downtown Seattle may assert that the CMSA wage data has produced SCA wage rates that are too low, from the workers' perspective, when compared with wage rates paid to workers performing similar work in the Seattle urban core, which presumably is their preferred "locality." But once the Wage and Hour Administrator has selected an appropriate "locality" for SCA purposes, the statute does not require the wage determination process to be balkanized to suit the unique perspectives of specific constituencies.

   In this decision, we have found that the Administrator's shift to using CMSA-based data represents a reasonable interpretation of "locality" under the Act because government statistical data show that the Puget Sound region is economically integrated. In addition, we have found that the wage and economic data supplied by the Armed Forces does not make a compelling case that the Administrator's CMSA-based wage determinations for the Tacoma and Bremerton-Shelton areas must be reconsidered. We also have found that the Administrator's 15% capping methodology is a reasonable transitional device, based on the record before us. To the extent the Armed Forces argue to this Board that an alternative approach would be preferable to the methodology chosen by the Administrator, they have raised a question of policy that does not properly belong before us.

   Accordingly, the petitions for review submitted by the Armed Forces are DENIED, and the Administrator's April 20 and July 20, 1998 rulings are AFFIRMED.

SO ORDERED.

       PAUL GREENBERG
       Chair

       CYNTHIA L. ATTWOOD
       Member

APPENDIX - MAP OF WASHINGTON STATE COUNTIES