Taxpayer use of statistical sampling.

Even though the IRS has acted to clarify a taxpayer's ability to use statistical sampling in making a broad range of determinations needed to establish tax liabilities, some questions still remain in light of recently released Field Service Advice (FSA) 200209028.

Director's Memorandum

In March, the Large and Mid-Size Business Division issued its long-awaited "Director's Memorandum" to examiners on taxpayer use of statistical sampling techniques. While in the past the Service has allowed taxpayers to use statistical sampling in certain circumstances, the new memorandum marks the first instance in which it has acknowledged implicitly and approved taxpayer use of the methodology broadly, without regard to a specific underlying issue of income, deduction or credit. The memorandum's stated purpose--to establish guidelines in "evaluating samples and sampling estimates by taxpayers"--may be interpreted to encompass taxpayer positions taken on returns, amended returns and refund claims.

The memorandum directs examiners to determine first whether a taxpayer "has appropriately used a probability sample to support or be the primary evidence of tax amounts" Rather than describing the circumstances for which a probability sample would be appropriate, the memorandum states that the determination will be based on the taxpayer's facts and circumstances and the situation.

Some of the factors used in determining whether sampling is appropriate include the time required to analyze large volumes of data, the cost of analyzing data and other books and records that may exist independently or have greater probative value. Further, sampling "generally should be considered appropriate if there is a compelling reason" for its use "and taxpayers cannot reasonably obtain more accurate information." However, sampling "generally should not be considered appropriate if evidence is readily available from another source that can be demonstrated to be a more accurate answer, or if the use of sampling does not conform to Generally Accepted Accounting Principles."

If the taxpayer's sampling technique is found to be appropriate, the memorandum provides statistical guidelines for testing the validity of the sampling in the specific situation. These guidelines generally parallel those set forth in Internal Revenue Manual 42(18)1, which IRS computer audit specialists use for their own statistical sampling audits. Generally, a taxpayer is expected to have a written sampling plan prior to undertaking the sampling; the instructions have a list of items that the taxpayer should include in the plan. The memorandum continues by providing suggested sampling techniques.

In ascertaining appropriate uses of statistical sampling, the statistical sampling method cannot be used to relieve taxpayers of their responsibilities for maintaining records required by the Code "which have specific documentation requirements for the entire population." This appears to mean, for example, that a taxpayer's ability to ascertain a statistically provable amount of entertainment expenses would not be sufficient to support a deduction in the absence of the collection and maintenance of the required records of those expenses. The memorandum leaves untouched a number of specific instances in "formal regulations, rulings, and procedures" that have addressed specific applications of the method.

The memorandum is directed solely to Service employees. However, the text and informal public statements of IRS officials indicate that the Service intends to provide guidance to taxpayers (and their consultants) that wish to use statistical sampling, while at the same time not stating a formal and precedential "official pronouncement of the law or the Service's position." Technically, the memorandum may not be relied on in any given case, as it may not be cited as precedent. However, it provides the working rules governing IRS examinations; if the taxpayer complies with the stated requirements, Service agents are expected to accept the sampling as reflective of the underlying taxpayer records.

FSA 200209028

FSA 200209028 focuses principally on the substantiation requirements set forth in Sec. 274(d), rather than on the nature and function of statistical sampling. It held that taxpayers must produce adequate substantiation for each item of travel and entertainment expense, or the IRS would disallow that item. The statistical sampling approach does not satisfy the strict Sec. 274(d) substantiation requirements:

Irrespective of the actual validity of the proposed methodology, sampling is nonetheless a form of a close approximation, the application of which is expressly prohibited by the regulations ... the specificity with which the taxpayer must substantiate each and every item covered under section 274(d), requires an expenditure-by-expenditure determination, and not a general determination that the deduction was substantiated.

The FSA appears to be concerned principally with the possibility that sampling would be used to compensate for gaps in a taxpayer's meal and entertainment records. However, the more common scenario is that a taxpayer determines on its original return an amount for the meal and entertainment deduction based on an item-by-item determination (periodically through the year and totaled at year-end), and the Service then samples that total statistically. Whether correct or not, the effect of FSA 200209028 appears to be minimized by the issuance of the Director's Memorandum.

FROM DANIEL J. WILES, J. D., AND DAVID MICHAEL REPASS, J.D., LL.M., WASHINGTON, DC

Editor:
Annette B. Smith, CPA
Partner
Washington National Tax Service
PricewaterhouseCoopers
Washington, DC