§ 4.51 Prevailing in the locality determinations.

(a) Information considered. The minimum monetary wages and fringe benefits set forth in determinations of the Secretary are based on all available pertinent information as to wage rates and fringe benefits being paid at the time the determination is made. Such information is most frequently derived from area surveys made by the Bureau of Labor Statistics, U.S. Department of Labor, or other Labor Department personnel. Information may also be obtained from Government contracting officers and from other available sources, including employees and their representatives and employers and their associations. The determinations may be based on the wage rates and fringe benefits contained in collective bargaining agreements where they have been determined to prevail in a locality for specified occupational class(es) of employees.

(b) Determination of prevailing rates. Where a single rate is paid to a majority (50 percent or more) of the workers in a class of service employees engaged in similar work in a particular locality, that rate is determined to prevail. The wage rates and fringe benefits in a collective bargaining agreement covering 2,001 janitors in a locality, for example, prevail if it is determined that no more than 4,000 workers are engaged in such janitorial work in that locality. In the case of information developed from surveys, statistical measurements of central tendency such as a median (a point in a distribution of wage rates where 50 percent of the surveyed workers receive that or a higher rate and an equal number receive a lesser rate) or the mean (average) are considered reliable indicators of the prevailing rate. Which of these statistical measurements will be applied in a given case will be determined after a careful analysis of the overall survey, separate classification data, patterns existing between survey periods, and the way the separate classification data interrelate. Use of the median is the general rule. However, the mean (average) rate may be used in situations where, after analysis, it is determined that the median is not a reliable indicator. Examples where the mean may be used include situations where: