100 BLS SURVEY UNION AGREEMENTS DAVIS-BACON DETERMINATION STATE WAGE SURVEYS INCUMBENT CONTRACTOR OTHER SOURCES As you can see, surveys by the Bureau of Labor Statistics were the source of data for roughly 70 percent of the determinations in effect at the end of each year. Union agreements were the source of data for slightly less than one-fourth of the determinations, and the other 6 or 8 percent were based on data from various sources, including a very few-less than 1 percent-based on wage board schedules determined under the Coordinated Federal Wage System. In contrast, the major sources of data relied upon for wage determinations under the Davis-Bacon Act are (1) collective bargaining agreements and (2) surveys conducted by mail requesting wage data for different types of employees. Department officials advised us that recently they have been placing a heavy emphasis on personnel responsible for administering the Davis-Bacon Act conducting onsite surveys and that this is now a major source of wage data. BLS surveys are not used as bases for Davis-Bacon Act wage determinations. According to the Department, several factors account for the fact that wage determinations are issued for all contracts subject to the Davis-Bacon Act. First, data on prevailing wages in the construction industry were more readily collected and applied than data under the Service Contract Act. There are only four basic types of construction work, and construction crafts are well established, are not changing rapidly, and are clearly defined. For example a carpenter's duties do not vary significantly from one project to the next and have not changed much in recent years. Also, potential sources of data are easy to identify-collective bargaining agreements are common, labor unions have or know where to get wage data as do contractors' trade associations, and contract awards are publicized. Although minimum wage rates determined under the Davis-Bacon Act are applied, at times, to contracts for certain types of services, this is not usually the case because the work performed under construction contracts subject to the Davis-Bacon Act is not usually similar to the work called for by service contracts. Also, the types of employees involved are not usually similar. The technique of using mail surveys, frequently used under the Davis-Bacon Act, has also been used by the Department on occasion in obtaining data for making wage determinations under the Service Contract Act. BLS makes regular wage surveys whether the data are to be used for Service Contract Act wage determinations or not, so the Department uses them because the data are readily available. In other areas where regular BLS surveys are not performed, the Department believes that there might be some, although very limited, potential for expanded use of mail surveys under the Service Contract Act. This will depend, however, on the difficulty in identifying potential sources of data. According to Department officials, the wage data surveys conducted by BLS upon which most Service Contract Act wage determinations are based are similar in some respects to the surveys conducted in connection with determining wage board rates under the Coordinated Federal Wage System. BLS and wage board data collectors are required to coordinate their efforts so that if they plan to visit the same employer, a single visit jointly by the two data collectors can be arranged. BLS surveys on which Service Contract Act wage determinations are based are usually more broadly based; that is, they collect data from more employers in more industries than do wage board surveys. Also, the procedures used to determine the wage rates once the basic data are collected are substantially different under the Coordinated Federal Wage System than they are for the Service Contract Act. Under the Service Contract Act, the objective is to determine the local prevailing wage rates for the various types of workers for work similar to that called for under the proposed service contract. If a rate established by a collective bargaining agreement is dominant, that rate is designated as prevailing. If a BLS survey is available, a median or mean rate based on the BLS survey is designated as the prevailing rate. Under the Coordinated Federal Wage System, the objective is to determine the local prevailing wage rates for various skill levels, regardless of the type of worker or work to be performed. For example, if it is determined that carpenters and plumbers require the same level of skill and will therefore be paid the same wages, it may not be necessary to determine the prevailing local rates for each skill or to relate the prevailing rate to the type of work to be performed. The Department believes it is generally precluded from applying wage board rates to service contracts because provision for such application was included in the proposed Service Contract Act of 1964 but was not included in the similar bill enacted in 1965. The Department also believes that the emphasis on skill levels under the Coordinated Federal Wage System surveys makes the wage board rates generally inappropriate for application to service contracts. As I mentioned, in a few cases, the Department has used wage board survey data in establishing prevailing rates under the Service Contract Act. The Department believes, however, that it has used this data to the maximum extent feasible. The Department prefers to use BLS survey data in making wage determinations under the Service Contract Act because BLS surveys cover more employers and industries and because the use of professional data collectors provides more assurance of reliable data. In this regard, data collection for wage board surveys is performed by wage board employees, half of whom are recommended by involved labor organizations, while surveys for the Service Contract Act are conducted by professional data collectors. NUMBER OF EMPLOYEES COVERED BY INDIVIDUAL CONTRACTS One of the proposed amendments to the act would require the Department of Labor to issue a wage determination for every proposed service contract expected to involve 25 or more employees unless the Secretary determined on the record, after opportunity for a hearing, that the issuance of a determination would seriously impair the conduct of Government business or would not be in the public interest. To assist the subcommittee in its consideration of this proposal, we were asked to provide information on how many notices of intention to enter into service contracts the Department received but for which it did not issue wage determinations in fiscal years 1970 and 1971 indicating that 5, 10, 15, 20, or 25 or more employees were expected to be involved. It was subsequently agreed with the subcommittee staff that data for only fiscal year 1971 would be provided because of the amount of resources required to compile the information and the short time in which the work had to be performed. Also, fiscal years 1970 and 1971 seemed to be similar based on the number of notices of intention the Department received and the number covered by wage determinations. The compilation of this data was performed by GAO by recording data from the notices of intention on file at the Department of Labor. Only 56 percent of the 21,081 notices of intention the Department received in fiscal year 1971 were on file and contained the information needed for the analysis. According to Department officials, many of the notices received were not on file because they were either lost in moving or were removed from the file and not returned. Also about 1,200 mail-haul contracts were subject to the Service Contract Act in 1971 and although these were counted as notices received, notices of intention were not actually submitted because of a special procedural arrangement between the Postal Service and the Department of Labor. Also, many of the notices on file did not show the number of service employees expected to be utilized under the contract. In making our analysis, we assumed-as Department officials believed to be the case that the usable notices on file were representative of the total 21,081 notices received in fiscal year 1971, and projected our findings for the 56 percent to the total. Accordingly, the results of our analysis are qualified to that extent. As shown in the next chart (attachment 5), 7,409 contracts were covered by wage determinations in fiscal year 1971. If all contracts involving 25 or more service employees had been covered, 1,230 contracts not included in the 7,409 would have been covered. Similar information is shown for contracts involving 20, 15, 10, and 5 or more employees and for all contracts. Of the 7,409 notices of intention which were covered by wage determinations during fiscal year 1971, only 756, or about 10 percent, were expected to involve 25 or more employees. According to Department officials, this situation occurred because when an existing wage determination or existing data are applicable, the notice of intention is covered by a determination, regardless of the number of employees expected to be utilized. Accordingly, many covered notices of intention are for contracts expected to involve very few employees. Covering an additional 1,230 notices of intention would represent about a 17-percent increase in the number covered. Department officials advised us, however, that obtaining the additional coverage would have required more effort and resources than the 17-percent increase would indicate. This is because the Department decided not to make and issue wage determinations for many of those notices for which data would have been difficult to obtain and those which they believed would not be applicable to other proposed contracts. The Department believes this practice is necessary to obtain maximum coverage with the limited amount of resources available for this purpose. We were also asked to comment on the adequacy of the staff and funds being allocated to administer the Service Contract Act as it is now written. The only specific indication we have in this regard is an estimate by the Deputy Assistant Administrator of the Wage and Hour Division that if the professional staff were doubled (about five more employees) and their funds for BLS surveys were quadrupled (additional $1.9 million) about 40 to 50 percent of the service contracts could be covered by wage determinations compared to the approximately 35-percent coverage in 1970 and 1971. He said that to try to increase the percentage of contracts covered beyond 50 percent 82-021-72- -3 |