Do you offer any advanced statistical tools and methodologies for recommending labor rate bands such that financial risk is minimized?
The answer below is the best answer we could give based on the inability to expand on statistical analysis in this format.
Open full Question Details
1.a. FAR 42.302(a) concerns responsibilities for the administrative contracting officer (ACO). DCMA has responsibility for forward pricing rate agreements (FPRA). In addition, DCMA has responsibility to review contractor labor rates, Department of Labor has responsibility for prevailing wage rates.
b. A statistical tool can be used to determine various items under descriptive statistics. The tool would have two tabs; the left tab would reflect definitions and the right tab would be the tool. Descriptive statistics are methods used to describe data in a collection set. Description of the data involves understanding the dispersion of the data. This dispersion is described as “central tendency,” the distribution of data around the center point of data point. A normal center of the data points is the mean, described as the mathematical average of the data points. Utilizing labor rates from a determined band 1, the statistical analysis would reflect the average labor rate, the median or the half of the data points. The standard deviation would also show the relationship between a set of data points to the mean. It is considered more accurate because an outlier can exaggerate the mean.
2a. It is a matter of business judgment in terms of what is reasonable in the variability of labor rates. For example, if using the OPM table that lists the federal government pay tables (without locality pay) from GS 01 through GS 15. The table lists the low end of the grade (step 01) as well as the midpoint (step 05), and the high end of the grade (step 10). You would create a column that reflects the difference between the annual pay at step 10 and the annual pay at step 1. Then the difference is divided by the annual pay at step 1 to determine the percent difference in each of the grades. The variability of labor wages within each grade is roughly 30%.
b. Another way to look at it is by dividing the federal government pay scale into three (3) pay bands; the first pay band consists of GS-01 step 01 through GS-05 step 10, the second pay band consists of GS-06 step 01 through GS-10 step 10, and the third pay band consists of GS-11 step 01 through GS-15 step 10. Again, determine the difference between the low end of the pay band (GS-01 step 01, GS-06 step 01, and GS-11 step 01) and the high end of the pay band (GS-05 step 10, GS 10 step 10, GS-15 step 10), and then divide the difference by the lower end of the pay band to determine variability. Because the pay bands are large, the variability ranges from 95% to 158%.
c. From the two examples, utilizing the same data, what is considered fair and reasonable in terms of variability? It depends. There is not a specific number that is correct in terms of variability. It depends upon business judgment, with some degree of logic behind the judgment and this logic required the ability to be justified under scrutiny.
3. For further information and clarity we advise that you take on more training in statistical analysis; a good place to start would be CON 270. As far as tools, once you have the know how you can create a statistical decision tool to assist in your development of a range right from your spreadsheet program. Of course others can be purchased, however we (DAU) do not release tools for statistical analysis nor do we have any that we are able to recommend. Please see your statistical analysts or IT personnel for assistance.