Paul L. Caron

Monday, October 30, 2017

Lesson From The Tax Court: Substantiation And The Cohan Rule

The married taxpayers in Partyka v. Commissioner, T.C. Sum. Op. 2017-79 (Oct. 25, 2017), fell prey to scammers who scheme was to rent a furnished house and then steal the furnishing. The scammers stole a substantial amount of household furnishings from the taxpayers and the taxpayer pegged the value of the stolen items at just under $30,000 for which they claimed a §165 loss deduction.

There was an interesting timing issue in the case, but what struck me was the long discussion of substantiation, and how what may be sufficient documentation of loss for one purpose (police investigation, prosecution) may not be enough for tax purposes. Indeed, the Tax Court found that the taxpayers were potentially liable for the §6662 accuracy-related penalty. I question, however, whether the Tax Court properly applied the Cohan rule. For details, see below the fold.

Here’s how the Tax Court described the Partyka’s efforts at documenting their losses:

Petitioners began cataloging and listing the missing items during the end of 2011 and early 2012 in order to have a basis for a criminal or civil action against the tenant and her husband. ***

Petitioners were able to find and offered at trial photographs of the household furnishings taken shortly before the tenant and her husband moved in during September 2011. On the basis of the photographs and coupled with their recollection, petitioners made a schedule of the items that were the basis of their $29,979 theft loss deduction. In order to show fair market values of the items, petitioners found examples of the items for sale on websites such as eBay. They also researched the cost of fabrication of items such as drapery to arrive at values. They used the prices from their research and applied some discounts to arrive at their estimates of fair market values.

So the taxpayers created a document that itemized the losses. That is, truly, the first step in proper documentation of loss. However, the loss deduction is limited to the lesser of basis or change in fair market value when items are damaged. Here, the taxpayers did recover some of the items but those were damaged. For all items, the Tax Court generously assumed that the taxpayers had paid more than the items were worth and so did not demand that the taxpayers prove their basis in each and every item listed in the schedule. That is, the Tax Court assumed that the change in fair market value would be what the stolen furniture and household items (all used) would be worth on the market.

The problem with the taxpayers’ substantiation lay in their inability to document their claimed fair market value for the items listed. For some of the items the Tax Court therefore simply disallowed any deduction. For example, the taxpayers listed the loss of a cocktail table and valued it at $800. But the taxpayers were unable to give the Court “sufficient detail for the Court to objectively estimate a value” for the lost cocktail table.

For other items, while the Court disagreed with the taxpayer’s estimate of value it did not disallow the deduction entirely. Instead it substituted its own estimation of value. Here are two examples:

Petitioners listed living room and two other sets of draperies and reached a composite value of $10,000 ($5,000, $2,300, and $2,700). They arrived at that value by calculating the cost to make new draperies. There is no evidence in the record establishing the age and condition of the draperies, and the replacement cost would not represent the total fair market value. Under the circumstances we find the total fair market value of the draperies to be $2,000. Petitioners also listed two beds with mattresses, box springs, linens, and covers for a composite value of $6,000. Here again, without evidence establishing the age and condition of these used bedding items, we find the total fair market value to be $1,200.

There are two lessons here. First---and most important---notice what substantiation means. It is not that the taxpayers failed to document what was stolen. They failed to document the fair market value of what was stolen. Their documentation of the loss may have been enough for police to conduct an investigation, for prosecutors to prosecute the scammers and perhaps even for a claim to their insurance carrier (although I am guessing they had no insurance here). But in tax law, the concept of “substantiation” means more than simply keep track. In this case, it means also proving up the fair market value.

The second lesson is perhaps less important, but I find it more interesting. Notice that for some items (e.g. the cocktail table) the Tax Court refused to allow any amount but for other items (drapes, beds) the Tax Court substituted its opinion on the worth of the stolen or damaged items. That is, the Tax Court instead allowed at least something. The Tax Court can do this under what is known as the Cohan rule. Cohan v. Commissioner, 39 F.2d 540, 543-544 (2d Cir. 1930)

What is interesting to me is how the Tax Court applied the Cohan rule. The Tax Court described the rule this way:

When a taxpayer establishes that he or she paid or incurred a deductible expense but fails to establish the exact amount of the deduction, the Court normally may estimate the amount allowable as a deduction. There must be sufficient evidence in the record, however, to permit the Court to conclude the amount of a deductible item.

I don’t know where the Court gets the language that I bolded.  I do not think that’s a fair reading of Cohan. In Cohan, the taxpayer was an impresario, a producer of Broadway shows. Heck, it was George M. Cohan!! You know, the dude who made Broadway Broadway.

Mr. Cohan was too busy schmoozing and dealing to keep track of expenses he incurred in the carrying on of his business. That got him in trouble in Tax Court which held that he could deduct zero expenses since he could not substantiate the amount of his expenses. Mr. Cohan appealed to the Second Circuit Court of Appeals, sitting in New York. Here is the relevant part of Judge Learned Hand’s opinion:

In the production of his plays Cohan was obliged to be free-handed in entertaining actors, employees, and, as he naively adds dramatic critics. He had also to travel much, at times with his attorney. These expenses amounted to substantial sums, but he kept no account and probably could not have done so. At the trial before the Board he estimated that he had spent eleven thousand dollars in this fashion during the first six months of 1921, twenty-two thousand dollars, between July first, 1921 and June thirtieth, 1922, and as much for his following fiscal year, fifty-five thousand dollars in all. The Board refused to allow him any part of this, on the ground that it was impossible to tell how much he had in fact spent, in the absence of any items or details. ... But to allow nothing at all appears to us inconsistent with saying that something was spent. True, we do not know how many trips Cohan made, nor how large his entertainments were; yet there was obviously some basis for computation, if necessary by drawing upon the Board's personal estimates of the minimum of such expenses. The amount may be trivial and unsatisfactory, but there was basis for some allowance, and it was wrong to refuse any, even though it were the travelling expenses of a single trip. It is not fatal that the result will inevitably be speculative; many important decisions must be such. We think that the Board was in error as to this and must reconsider the evidence.

Bottom line for the Cohan Court: the Board should make as close an approximation as it can, bearing heavily if it chooses upon the taxpayer whose inexactitude is of his own making.

I don’t think the Tax Court correctly followed Cohan in Partyka. For example, the Court accepted as true that the taxpayers really did lose a cocktail table to the scammers. But the Court refused to allow them any deduction because, it said, the taxpayers did not give “sufficient detail for the Court to objectively estimate a value.” I am not sure Cohan says the Court must be able to “objectively” estimate a value. In fact I am pretty sure the whole point of Cohan is that if the Court is convinced of the fact of the expense (or, here, loss), then the Court should peg a value to it instead of just disallowing it entirely. As in Cohan, here “there was obviously some basis for computation” in the form of either a picture or even just the words "cocktail table."  In such cases I read Cohan as instructing the Court to allow something, even if “the amount may be trivial” and will “inevitably be speculative.” The Court disallowed a total of about $7,700 in claimed losses not because it disbelieved the taxpayers actually lost the items (e.g. the cocktail table) but because the taxpayer had not given sufficient detail for the Court to objectively estimate a value. But the Court could still peg a value of 10% of what the taxpayers claimed.  Sure, that's a trivial amount and sure that is speculative.  But that is the point, I think, of Cohan.  

Applying Cohan correctly is especially important here I think because the amount allowed to the taxpayers will determined whether they are subject to the accuracy related penalty of §6662(a).  So even giving taxpayers 10% of what they claimed would bump up their allowed losses by $770 and may allow them to escape the penalty.

Bryan Camp, New Cases, Tax | Permalink


@Kathleen and Publius: I think y’all raise an important point. The Cohan rule is the Tax Court’s own adaptation of what Judge Hand opined in the Second Circuit opinion. The Fifth Circuit might well approve the Tax Court’s approach. We’re in Golsen territory here. My point (as West Coast noted) was that this approach is a deviation from what the Cohan court actually held. And, normatively, I think Hand’s approach is the better one. If a judge believes the fact of an expense, then sure a taxpayer should be allowed something above zero.

But the Tax Court’s “modification” of the actual rule laid down in Cohan might be a useful way to avoid making otherwise difficult credibility determinations. Take the claimed loss of cocktail tables as an example. Now it might be that the Tax Court simply did not believe that these taxpayers actually lost a cocktail table. But it would be difficult to say “taxpayers have not convinced us they lost a cocktail table” which is basically making a credibility determination against the taxpayers. The only evidence of loss is the taxpayer’s word. So not believing that is to basically call the taxpayers liars. That’s a difficult determination to make. It’s easier to say, in essence, “we believe the taxpayers but gosh darn it we just don’t have enough information to assign a value to that lost table.” Now the taxpayers are truth-tellers but just lack information.

Posted by: Bryan Camp | Nov 2, 2017 5:49:26 AM

@Jack: I wondered that myself. I think Treas.Reg. 1.165-8 (with a x-ref to 1.165-7) provides that, for theft losses, the amount of the loss is limited to the lesser of basis or change in fmv and “the fair market value of the property immediately after the theft shall be considered zero.” I think the Court was assuming that the TP’s basis was higher than the change in fmv. That seems a reasonable assumption for household furnishings. It’s also consistent with the spirit of Cohan, which is to sand down some of the square corners taxpayers must turn in meeting their various burdens of production. To require taxpayers to prove up both the fmv of the items and to trace the cost basis for each and every household item lost would be a pretty darned sharp corner to turn.

Posted by: Bryan Camp | Nov 2, 2017 5:36:34 AM

Why is value the measure? If this was rental property, wouldn't the proper deduction be basis, under IRC 165(c)(1) or (2)? Would still need evidence, but not of value -- instead, of cost.

Posted by: Jack Bogdanski | Nov 2, 2017 1:21:15 AM

A survey of the cases suggests to me that as long as a taxpayer produces some evidence that can support an estimate a court must make an estimate, but failure to provide any such evidence leaves the court the discretion to either make an estimate or deny the deduction. That decision is probably grounded at least in some part on the court's degree of confidence that it has the competency to make a reasonable estimate under the facts and circumstances.

Posted by: Mike Petrik | Nov 1, 2017 4:53:55 AM

You don’t know where the 2d Circuit got the tule that “There must be sufficient evidence in the record, however, to permit the Court to conclude the amount of a deductible item?” Well, a court may estimate the amount of the deduction where the taxpayer has failed to show the exact amount of the deduction, so long as the evidence shows that the taxpayer is entitled to the deduction, and there is sufficient evidence in the record from which the Court may estimate the exact amount. See e.g., Trigon Ins. Co. v. United States, 234 F. Supp. 2d 581, 588-90 (E.D. Va. 2002). Courts may rely on Cohan, with appropriate caution, to estimate the amount of a claimed deduction in cases where the taxpayer is unable to produce evidence substantiating the exact amount of a claimed deduction. See e.g., Dunn v. Comm'r, 301 F.3d 339, 358 (5th Cir.2002); Ellis Banking Corp. v. Comm'r, 688 F.2d 1376, 1383 (11th Cir.1982).
But where a taxpayer fails to provide evidence that would permit an informed estimate of the amount of a deduction, the Cohan rule generally is not applied. Trigon Insurance Co. v. United States, 234 F. Supp. 2d at 588-90; Vanicek v. Comm'r, 85 T.C. 731, 743 (1985); Washington Mutual, Inc. v. United States, 996 F. Supp. 2d 1095, 1103 (W.D. Wash. 2014).

Posted by: Publius Novus | Oct 31, 2017 2:00:38 PM

I think Camp's point is that the court DID find sufficient evidence that the expense had been incurred, which is why it allowed the deduction. The goofy part is that the court said there had to be enough evidence to determine the AMOUNT of the deduction under Cohan, which is clearly not true.

Posted by: West Coast Tax Prof | Oct 31, 2017 7:59:08 AM

Dear Mr. Camp, case law has held the Cohan rule inapplicable where a taxpayer fails to produce sufficient evidence that he or she actually incurred any deductible expenses because the evidence is not sufficient to permit the court to make a reasonable estimation of the total deductible amount of expenses or losses the taxpayer incurred, or, not relevant here, where the taxpayer's records of deductible expenses could have been produced but for his or her failure to cooperate. I see a long history of case law on the matter, but the most recent case I find on the matter is Oatman v. Commissioner, T.C. Memo 2017-17, available on the Tax Court's website.

Posted by: Kathleen Adcock | Oct 31, 2017 4:43:13 AM