(a) DLA contracting officers are required to obtain price and cost-type information, without requiring certification (see FAR 15.402(a)(2)(i) and (ii)), when TINA does not apply and there is no other basis for determining that the proposed prices are fair and reasonable (e.g. through market research and price analysis techniques). In these situations when TINA does not apply, the distributor/dealer must provide its cost-type data when required by the contracting officer. Additional guidance is contained in DFARS Procedures, Guidance, and Information (PGI) Subpart 215.4. Moreover, if the contracting officer is relying on previous prices paid by the Government, the contracting officer must establish that a thorough price or cost analysis was performed on the previous buys (DFARS PGI 215.403-3(4)).
(b) It is critical that all levels of management support the contracting officer in acquiring cost-type data necessary for determining fair and reasonable prices. If the exclusive distributor/dealer’s representative does not provide the information requested by the contracting officer, the situation should be elevated to higher levels of management within both the government and contractor organization. When TINA does not apply, any distributor/dealer who does not comply with the requirement to submit cost-type information for a contract or subcontract is ineligible for award unless the Head of Contracting Activity (HCA) determines that it is in the best interest of the Government to make the award to that offeror in accordance with FAR 15.403-3(a)(4). The contracting officer must ensure the following is clearly documented in support of the request to the HCA:
(1) The effort taken to obtain the data;
(2) The need for the item or service; and,
(3) Increased cost or significant harm to the Government if award is not made.
(c) In the event of an exigent situation when the HCA has approved award without obtaining requested cost data, the contracting activity must notify the Procurement Integrity and Pricing Division (J-73), and J-7 will then notify Defense Procurement, Acquisition Policy, and Strategic Sourcing (DPAPSS) (see mandatory PGI 15.402-90).
15.402-91 Pricing Policy - EMall
(a) FAR 15.403-3(c)(1) states that at a minimum, the contracting officer must use price analysis to determine whether the price is fair and reasonable whenever the contracting officer acquires a commercial item. The fact that a price is included in a catalog does not, in and of itself, make it fair and reasonable. If the contracting officer cannot determine whether an offered price is fair and reasonable, even after obtaining additional information from sources other than the offeror, then the contracting officer must require the contractor to submit information other than certified cost and pricing data to support further analysis in accordance with FAR 15.404-1 proposal analysis techniques. In most situations the contracting officer should be able to obtain information needed to perform price analysis on commercial catalogs. However, in limited situations where price analysis is not adequate, the contracting officer should request other than certified cost and pricing data and perform cost analysis.
(b) Determining the price reasonableness of a commercial catalog for the DOD EMALL poses a challenge due to the number of items in a catalog, e.g. 50,000 or more, with minimal or no historical demand or sales data, coupled with low (below the micropurchase threshold) prices. The contracting officer shall use the mandatory procedures in PGI 15.402-91 in making price reasonableness determinations for new commercial catalogs to be placed on EMALL, additional items added to existing catalogs, or renewal of catalogs.
(c) Situations for using statistical analysis. Statistical analysis can be invaluable to use in developing the Government objective for contract prices based on historical values. Historical costs or prices are often used as a basis for prospective contract pricing. When several historical data points are available, you can use statistical analysis to evaluate the historical data in making estimates for the future. Statistical analysis is particularly useful in the analysis of commercial catalogs with large numbers of items where an overall price reasonableness decision is made on the catalog based on the items statistically selected for review.
(d) Applicable principles in using statistical analysis. Most audit universes are widely dispersed. Usually, there is a wide variation between the smallest and largest individual dollar amounts, with most of the amounts being relatively small and only a few amounts being very large. Since a random sample from the entire universe would probably include only a few large (high dollar) items, the reliability of the results would be correspondingly low. This is possible because wide variations are likely between questionable amounts for individual large items and the average of questionable amounts from the universe. These issues can be addressed by a combination of stratification and random sampling.
(e) Stratification involves the evaluation of a large quantity of data without sacrificing quality. Stratified sampling techniques allow examination of 100 percent of the items with the greatest potential for cost reduction, while random sampling helps assure that there is no general pattern of overpricing smaller dollar items. Stratification of the universe into several dollar ranges or strata can be used to improve the review reliability and reduce the overall number of items evaluated. Normally, the universe is stratified into high-dollar stratum (for 100 percent evaluation) and several other strata from which samples are selected for evaluation. The review effort is concentrated on the high-dollar items where the risk is greatest. The samples are statistically selected (random sampling) from each of the other strata, which are used as the basis for making decisions on the price reasonableness of the corresponding universe.
(f) The exact number of strata for statistical sampling will depend on the dollar range of various catalogs (based on judgment of the evaluator). The underlying assumption of random sampling is that a sample is representative of the population from which it is drawn. If the sample is fairly priced, the entire stratum is assumed to be fairly priced. Conversely, if the sample is overpriced, the entire stratum is assumed to be proportionally overpriced.
(g) We do not want Contracting Officers and Cost/Price Analysts to perform 100 percent review of the entire commercial catalog, because of the limited resources available, and the relative risk of overpayment by not performing 100 percent review. However, we must have an acceptable methodology for determining price reasonableness of the entire catalog through a viable review process. Therefore, we will use statistical sampling in our reviews. Statistical sampling is preferred over judgmental sampling, because of its advantages, which include objectivity, overall defensibility, and measurability of the risk of substantial (material) sampling error. In a randomly selected sample each item has a known chance (or probability) of being selected. The results of a statistically selected sample can be objectively applied to the universe from which it was drawn to assist the evaluator in projecting the results of evaluation of the sample to the universe. You should not attempt to project results to the universe if the sample was not randomly selected.
(h) In sampling for variables, there is no single “best sample size.” Sample size is a compromise between the inversely related considerations of precision and evaluator time.
Evaluating too many sample items can result in achieving greater precision than necessary. That is, more resources will have been devoted to sample analysis than necessary. Therefore, the use of EZ-Quant software is mandated to ensure selection of the proper sample size. The EZ-Quant sample size estimation option allows the evaluator to determine the optimum sample size for variable sampling based on three factors: (1) presumed error rate or the results of a sample from a similar evaluation universe; (2) precision amount; and (3) confidence level.
(i) Sampling plan. The successful application of statistical sampling begins with the design of the sampling plan. General sampling plan elements are list below.
(1) Briefly state the objective of the sample or what the evaluator is looking for in the universe. For example, “The objective is to determine price reasonableness of offeror’s catalog.”
(2) Describe the universe and state its size.
i. Describe the sampling unit (i.e. the basic auditable item to be examined);
ii. Specify the scope of the review so that all sampling units pertinent to the sampling objective can be identified; and,
iii. State the size of the universe (i.e. the total number and amount of all sampling units).
(3) Select the universe size - the number and value (if applicable) of all sampling units. An example is the universe of 5,000 catalog items totaling $2,000,000.
(4) Describe the sampling frame. This is the physical or electronic representation of the universe to which the mechanics of sampling will be applied, for example, an Excel spreadsheet of the offeror’s parts catalog with associated prices.
(5) Select a suitable sampling approach. When performing variable sampling the reviewer may choose physical unit sampling or dollar unit sampling (DUS).
(6) Develop the sampling reliability parameters.
(7) When sampling for variables establish a sample size using sample sizing utilities in DCAA EZ-Quant.
(8) Describe the sample selection method.
(9) A copy of the sampling plan and a cross reference to the items reviewed for price
reasonableness shall be retained in the contract folder for each catalog reviewed.
