DCAA Compensation Reviews - Current Methodologies Used

Those of you following compensation issues as they relate to costs charged to Government contracts are no doubt familiar with the two ASBCA (Armed Services Board of Contract Appeals) cases from 2012; JF Taylor (ASBCA Nos. 56105, 56322) and Metron (ASBCA Nos. 56624, 56751, 56752). The ASBCA found for the contractor (mostly) in both decisions while criticizing the methodologies employed by DCAA in deriving its estimates of "reasonableness". Refer to the following postings for a brief summary of those decisions.

     Fatally Flawed Compensation Audit

     Fatally Flawed Compensation Audit - Part 2

     Executive Compensation - The Government Loses Again

Over the next few days, we will be reporting on our own post-JF Taylor/Metron observations and experiences. We have two general observations. DCAA has not substantively changed the way it audits compensation costs, even though the methodologies it continues to use were called into question in these cases.  Secondly, contracting officers seem to be less likely to follow or rely upon DCAA recommendations and in some cases, have become their own compensation "experts".

We will address the first observation today. The question we keep asking ourselves is why does DCAA continue to employ the same methodologies today, that were called "fatally flawed"  in the J.F. Taylor case. That methodology averaged two or three statistical studies (compensation surveys) and added a 10 percent range of reasonableness.

The short answer here is that DCAA does not believe it lost the J.F. Taylor case on its merits. In 2014 correspondence addressing the applicability of the J.F. Taylor case, DCAA wrote:

... as specifically noted in the Board decision, the JF Taylor expert's statistical arguments were accepted by the ASBCA because the evidence was "unrebutted" by the Government. Had we rebutted the evidence, the outcome may have been different. Therefore, we continue to utilize our methodology...

So there you have it. DCAA continues to add a ten percent range of reasonableness to the average (i.e. 50th percentile) of several surveys as benchmarks to assess the reasonableness of compensation.