Rev. Proc. 2011-42 (Statistical Sampling)

What it holds

Sets out an IRS-blessed way to use statistical sampling in tax computations. It was written for other areas of tax, but the same sampling logic is persuasive for large cost-seg jobs.

Where this comes from

Some tax computations involve thousands of small, similar items. Counting and pricing every one can cost more than the answer is worth. By 2011 the IRS had already accepted sampling in scattered places, like meal and entertainment substantiation.

Revenue Procedure 2011-42 pulled the rules together. Its stated purpose is to give taxpayers guidance on the use and evaluation of statistical samples and sampling estimates. It tells you when a sample is appropriate and exactly how the IRS expects it to be built, documented, and computed.

What it established

The procedure says a taxpayer may use a statistical sample where the facts make it appropriate, and it conditions that on following three appendices. Appendix A sets the sampling plan standards, meaning how the population, strata, and sample design must be defined. Appendix B sets the documentation standards, meaning the records you must keep so the sample can be checked. Appendix C provides the technical formulas.

It also builds in a penalty for sloppy samples. The allowed estimate depends on the sample's relative precision. If precision is strong, within 10 percent, the point estimate may be used. If precision is weaker, the taxpayer must use a less favorable number tied to the 95 percent one-sided confidence limit. A loose sample costs you deduction dollars by design.

How it shows up in a study

The IRS cost segregation guide lists sampling or modeling as one of the recognized study approaches, and this procedure is the standard a sampling plan should meet. On a portfolio of similar properties, or one property with thousands of repeating units like hotel rooms or apartment fixtures, an engineer can fully analyze a properly drawn sample and extend the results.

The key on exam is the paper trail. The sampling plan, the population definition, the selection method, and the math all need to be in the file. A documented plan that follows the appendices turns sampling from a shortcut into a supportable method.

What it does not mean

This procedure was not written specifically for cost segregation, and it does not say every study may sample. It is persuasive support for a careful sampling plan, and the facts still have to justify using one. A single mid-size building rarely needs sampling at all.

It is also not a license for judgment sampling, where someone hand-picks convenient examples. The procedure's whole point is statistical rigor: random selection, defined strata, computed precision. An eyeballed extrapolation without a plan gets no help from this authority.

Category

Methodology & procedure

Applies to

All property types

Status

Vetted

This page explains a tax authority in plain words. It is not tax advice for your situation. The way this authority applies to your property is reviewed by a licensed tax professional. Citation is provided so you or your advisor can read the primary source.