Paul L. Caron

Monday, October 26, 2020

Lesson From The Tax Court: Losing Gambler Gets Twice Lucky In Tax Court

Taxpayers rarely walk away from casinos richer than when they entered.  The odds are not in their favor.  If a slot machine pays out $1,200 or more, however, the casino will still report that win on a W-2G, even if the taxpayer loses all of it before leaving the casino.  The theory is that $1,200 is income to the taxpayer and the taxpayer’s choice to use it for more gambling is no different than the taxpayer’s choice to use that $1,200 for other consumption.

The IRS recognizes that the reality is different from theory and so it permits taxpayers to net their gambling gains and gambling losses for each visit to---or session at---a casino.  In the unlikely event they leave the casino a net winner, those wagering gains are gross income which must be reported.  If they leave a net loser, they may be able to deduct those wagering losses against wagering gains up to the amount of wagering gains.  Tax Court precedents uphold this per-session method of accounting for gambling gains and losses.  In addition, plenty of precedent requires taxpayers to substantiate their wagering losses for each session.

In John M. Coleman v. Commissioner, T.C. Memo. 2020-146 (Oct. 22, 2020), Judge Lauber bucked both sets of precedents to allow the taxpayer a gambling loss deduction equal to over $350,000 of gambling wins reported on various W-2Gs.  There are good reasons for why Judge Lauber did this, but the bottom line is that this taxpayer got twice lucky.  It helped that he was represented pro bono by two high-powered tax attorneys from Morgan Lewis.  Let's look at what we can learn from them and from this case.  Details below the fold.

Background: Section 165(d)
Congress has a general policy of allowing taxpayers to deduct from their income the money it takes to produce that income, and taxes only the net.  There is no principled distinction between netting expenses against gross receipts to arrive at gross income (e.g. Cost of Goods Sold rules in §471) and netting expenses against gross income to arrive at net income (e.g. business expenses rules in §162).  All reflect a Congressional choice to tax income net of associated costs.  In contrast, Congress does not permit taxpayers to deduct personal expenses. §262.

Losses are a special kind of cost and Congress applies the distinction between business-related losses and personal losses in §165.  Subsection (c)(1) and (2) allows deductions for losses incurred in a trade or business or in transactions entered into for profit.  And (c)(3) permits casualty loss deductions.  Otherwise, however, §165 does not permit taxpayers to deduct personal losses, such as losses from the sale of principal residence.

Gambling presents a special problem for taxation.  When a single wager wins, the wager is treated as a cost of producing the win and, consistent with the general policy of deducting what it takes to make money, the amount of the win is reduced by the cost of the wager to determine the gain from the wager transaction.  Bonaparte v. Commissioner, T.C. Memo. 2017-193 at page 8, note 3.  It’s like a COGS.

Gambling losses are more difficult and are addressed by §165(b).  It provides that “losses from wagering transactions” may be deducted but only up to the amount of “gains from such transactions.”

Section 165(d) is good news for casual gamblers.  It permits deduction of what would otherwise be a personal loss. The deduction allowed by §165(d) is consistent with the principle that the cost of producing income ought to be deductible from the income produced.  That is good news indeed.

Section 165(d) is also bad news, however.  Losses in excess of winnings are simply disallowed, regardless of whether the taxpayer is gambling as a hobby or as a business.  Humphrey v. Commissioner, 162 F.2d 853 (1947) (wagering losses were subject solely to §165(d) and thus not deductible as costs under §162(a) or the general rules of §165).  Unlike, say, §172, there no allowance for a net gambling loss to be used against prior-year or later-year gambling winnings.

Background: Taxing The Churn
There is worse news for casual gamblers.  Section 165(d) by itself simply tell taxpayers what deductions are allowed.  Determining “what is allowed” does not complete the deduction analysis.  One must then see where the deduction is allowed: may it be taken above the line, from gross income, or must it be taken below the line, from adjusted gross income (AGI)?

If the gambling activity is a trade or business, then the loss deductions permitted by §165(d) go above the line as part of the Schedule C determination of gambling income.  See LaPlante v. Commissioner, T.C. Memo. 2009-226.  Chief Counsel Generic Advice 2008-13.  That is all well and good and consistent with the overall policy of taxing income net of expenses to produce it.

If the gambling activity is a hobby, however, the deductions go below the line on Schedule A.  That is not good news.  True, by being listed in §67(b), the §165(d) deduction is not subject to the 2% floor imposed by §67(a) or to the total disallowance temporarily imposed by §67(g).  Still, being forced to take the §165(d) deduction below the line has the very unhappy consequence of taxing the churn.

It does that in two ways.

First, the allowed deduction fights with the standard deduction. That is, §63 requires taxpayers to choose between itemizing their below-the-line deductions and taking the standard deduction.  They cannot do both.  Once they choose, they are stuck with that choice. §63(e)(1).

Gamblers who are in the trade or business can thus take their §165(d) deductions PLUS the standard deduction.  But gamblers who are merely hobbyists must choose either/or.  Thus, their gambling WINS must be counted as gross income, but their gambling LOSSES are separated and taken later (if at all) below the line.

This fight with the standard deduction makes it difficult to truly offset gaming winnings.  For example, assume a taxpayer gambled in 2019 as a hobby.  The standard deduction for a single taxpayer in 2019 was $12,200.  If the taxpayer had total wagering gains of $10,000 and wagering losses of $18,000, the taxpayer would report $10,000 of wagering gains as gross income.  While §165(d) would then permit a deduction for $10,000 of the wagering losses, §62 would send that deduction below the line where it would fight with, and lose to, the standard deduction, assuming other itemized deductions were no greater than $2,200.  For this taxpayer, it would be as if there were no deduction allowed for wagering losses.  That’s taxing the churn, i.e. the volume of play.

In contrast, a similar taxpayer who could claim they were in the “business” of gambling would be able to deduct $10,000 above the line, thus wiping out the $10,000 wagering gains while also claiming the $12,200 standard deduction.  For a real-life example of this consequence see Viso v. Commissioner, T.C. Memo. 2017-154.

Second, taxpayers who actually are able to deduct wagering losses below the line still end up reporting phantom AGI.  That can have various effects on other parts of a return.  For example, assume our casual gambler in the above example gambles for just one day in 2019.  The taxpayer withdraws $8,000 from the bank and loses it all by the end of the day.  The taxpayer is economically poorer by $8,000.  However, assume during the day of play the taxpayer had total wagering gains of $10,000 and total wagering losses of $18,000.  If the taxpayer had to report the $10,000 as gross income it would increase the taxpayer’s AGI by $10,000 of phantom income.  It was just churn.  If that does not bother you, just add zeros.  Taxpayer now withdraws $80,000 and loses it all except the churn is higher.  Say the taxpayer has $100,000 of interstitial wagering gains but $180,000 of interstitial wagering losses.  Now the taxpayer reports the $100,000 as gross income but cannot deduct the $100,000 of losses until after calculating AGI.

The IRS has limited the damage here with the idea of “per-session” in Chief Counsel Advisory 2008-11.  The idea permits taxpayers to net all wagers made during an identified session of gambling (not exceeding a 24 hour period).  Applying the per-session method to my example, the taxpayer would not be taxed on the churn because the taxpayer would not have any “gains from wagering transactions” at the end of the session.

But what if we extend my hypo so that the taxpayer loses the $8,000 over the course of the year with the interstitial wins of $10,000 and interstitial losses of $18,000 spread over multiple sessions.  Some sessions end with a win and others end with a loss.  Now is the taxpayer being taxed on the churn or do we have “instances of undeniable accessions to wealth, clearly realized, and over which the [taxpayer has] complete dominion.”  Commissioner v. Glenshaw Glass, 348 U.S. 426, 431 (1955)?  Well, if the taxpayer cashes out some sessions with a win, those dollars won can be used or invested or saved.  That sure looks like gross income and not just churn.  So that is why the IRS adopts the session method.  When a taxpayer nets out a win, that’s gross income to be reported.

Taxpayers who are net losers in a year would like to not have to report any wagering gains as gross income.  They would like to employ a yearly netting to avoid being taxed on the churn.

The Tax Court has adopted the per-session method and rejected the yearly netting method.  Shollenberger v. Commissioner, T.C. Memo. 2009-306.  There, the taxpayers had one really lucky visit to a casino, hitting a $2,000 slot machine jackpot.  The taxpayers had taken $500 to the casino and, after hitting the jackpot, proceeded to continue playing until they had $1,600 left.  They then quit and went home.

On audit, the Revenue Agent dinged the taxpayers for $2,000 in unreported income, using only the $2,000 winnings reported by the casino on Form W-2G.  In effect, the IRS ignored the Shollenbergers’ other wagering during their one visit.

Using the session method, Judge Thornton held that the taxpayers did not have $2,000 of unreported income, concluding that “on March 29, 2005, petitioners entered the casino with $500 and took home $1,600 of winnings” and thus “the amount of gambling income which petitioners should have reported on their 2005 return was $ 1,100.” Op. at 3.

Judge Thornton, however, rejected any broader concept of netting.  The taxpayers had also proved up an additional $2,264 in other losses from wagering transactions during the tax year.  They argued that they should be able to net these losses against the jackpot earnings.  Judge Thornton rejected that argument:

Insofar as petitioners mean to suggest that section 165(d) permits their gross income from slot machine play to be calculated by netting all their 2005 slot machine gains and losses, we disagree.  ... To permit a casual gambler to net all wagering gains or losses throughout the year would intrude upon, if not defeat or render superfluous, the careful statutory arrangement that allows deduction of casual gambling losses, if at all, only as itemized deductions, subject to the limitations of section 165(d). Id.

What the taxpayers were unable to do is prove that their other gambling sessions during the year always netted out net losses.  Thus, unless they could tie those various losses to identifiable sessions, the Court would not permit them to use those undifferentiated losses to offset the $1,100 gain they had to report on the income side.

Background: Player Accounts
Casinos track players.  Players have long been able to sign up for Player Accounts, which the Casinos use to monitor play and give “comps” to those who play enough.  For a detailed description of how Casinos tracked players in the pre-electronic card days, see Barbiero v. Commissioner, T.C. Memo. 1992-381.

Computer technology has enhanced Casinos' ability to track play and, since about 2005, Casinos have given players the option of plastic cards with a magnetic strip across the back that have the look and feel of a traditional credit card.  You can read an interesting history of player tracking systems in Pepin, Player Tracking: You’ve Come a Long Way, Baby. Global Gaming Business Magazine, May 25, 2011.

A player loads money into their Player Account at a cage or self-service kiosk. The player uses the card by inserting it into the slot machine and entering a security PIN that allows the machine to access the account and load account balance information into the machine.  When the player is done at that machine, the player presses a button to end play and the machine transfers the play data back to the Players Account.  The player can then remove the card and go to another slot machine, or perhaps a table game.  When the player is ready to cash out, the player must return to the cage or a self-service terminal to convert the account balance into cash or refunds to a credit card.

Players do not always use their player cards.  Sometimes they think using the cards adversely affects their luck.  Sometimes they think if they don’t use the cards, the Casino cannot track them.  Sometimes they lose their cards, or forget to bring them.

Background: Substantiation and Cohan Doctrine
Taxpayer must prove up their entitlement to §165(d) deductions no differently than proving up any other deduction.  The Tax Court has historically been punctilious in requiring gamblers to prove their net losses using the session method.  Shollenberger is one example.

Another example is LaPlante v. Commissioner, T.C. Memo. 2009-226, where the taxpayer claimed to have made 20-25 visits to one casino (Foxwoods) during 2004.  One time she won an $8,000 jackpot but lost $4,000 of it during the same session.  She testified that was her only winning session.  So she reported the $4,000 net gain as income and then deducted from that the undifferentiated losses during the year, claiming they totaled more than $4,000.  The taxpayer’s problem was that the IRS received various W-2G’s from Foxwoods showing payments of over $30,000.  Still, the taxpayer was able to provide a letter from Foxwoods that showed she had lost a total of $35,480 at the slots during 2004.

The Court did not accept proof of the undifferentiated losses as adequate to offset the reported W-2G winnings.  Focusing on the per-session method, the Court rejected the letter as adequate substantiation, even though it accepted that “the letter is helpful in confirming the overall picture that petitioner lost money for 2004, a point not in dispute.”  The fact that the taxpayer netted losses over the entire year, however, was not enough to substantiate the net win or net loss for each session. The Tax Court noted that Ms. LaPlante could have asked for the casino records of her player account and submitted those to show session results.  It drew a negative inference from her failure to do that.

The standard advice to gamblers follows the per-session method.  For example, this website suggests not only that a taxpayer track each session but also track different types of gambling (keeping separate track of slot machine play and craps tables, for example).  That seems excessive to me, but the basic idea is important:  track your activity.

In practice, many gamblers fail to keep proper contemporaneous records of their casual trips.  Generally, they know they are going to lose, so why bother?  Then comes the thrilling bit hit on the slot machine.  And now they realize their failure to keep records means they will basically be taxed on the churn.  But ...

.... the Tax Court has a doctrine that may help such gamblers.  It’s a general rule that if a taxpayer can convince the Court of facts that give rise to an item of deduction, but are unable to establish the exact amount of the deduction, the Court will guesstimate a reasonable amount if the taxpayer gives the Court a reasonable basis on which to make the guess.  It’s called the Cohan rule for reasons I explained in Lesson From The Tax Court Substantiation and the Cohan Rule, TaxProf Blog (Oct. 30, 2017).

The year at issue is 2014.  Mr. Coleman did not file a return. The IRS sent him an NOD based on a Substitute For Return it had prepared based on third-party information returns it had received from various casinos.  Those information returns reported payouts of $350,000 in gambling winnings.  Mr. Coleman petitioned the Tax Court in 2017 to contest the asserted tax liability.

In 2014 Mr. Coleman was married and self-employed as an insurance consultant.  His wife had wage income and he earned about $77,000 from self-employment as an insurance consultant.  In addition, Mr. Coleman received $150,000 as settlement for a personal injury claim.

In 2014 Mr. Coleman gambled in four different casinos for a total of at least 193 days.  He played slots almost exclusively.  At each casino Mr. Coleman had a player rewards account, but like many gamblers, he used his player card inconsistently.  Based solely on his player accounts, two of the casinos showed Mr. Coleman as having net gambling winnings for 2014: almost $53,000 at Dover Downs and $1,500 at the Baltimore Horseshoe.  The record contains no information from the other two casinos.

During 2014 Mr. Coleman made 210 draws upon either his credit card or his bank account at ATM machines located within the four casinos.  The withdrawals totaled just over $240,000 and the IRS agreed to stipulate that all of the money withdrawn (less about $2,500 in transaction fees) was spent on gambling.

Lesson 1: The Cohan Rule Applies to Gamblers Too
Judge Lauber framed the case this way: “whether petitioner has substantiated gambling losses in excess of $350,241, the amount of his gambling winnings reported to the IRS.”  Op. at 1.

Before the Court, Mr. Coleman rolled the dice.  He was unable to substantiate his net gains or losses from any session of gambling.  Instead, he and his lawyers argued for a yearly netting and to even get that they needed to convince the Court that it could, under the Cohan rule, reasonably estimate his losses for the entire year.

As to the Cohan rule, they rolled a seven.  Judge Lauber concluded that “petitioner has provided sufficient evidence to show that his gambling losses were at least $350,241.”  Op. at 17.  It appears that Judge Lauber relied on the following three sources of evidence.

First, and critically, was the testimony of an expert witness, Mr. Mark C. Nicely.  Mr. Nicely basically said a gambler who spent the amount of time Mr. Coleman spent playing slots would have overall net gambling losses of at least $151,690 during 2014.”  This was, said the expert, a “99% level of certainty.”   Critically, Mr. Nicely testified — and the Court accepted —  that “it would be virtually impossible for [Mr. Coleman] to have annual net gambling winnings.”  Mr. Nicely testified that the odds of Mr. Coleman “having enjoyed even $1 of net gambling profit, for the entirety of 2014, were at least 140 million to 1."  Op. at 11-12 (emphasis supplied).

Second, Mr. Coleman’s attorneys introduced evidence that: (a) Mr. Coleman was a compulsive gambler; (b) Mr. Coleman’s bank accounts started and ended the year with the same small account balances, (c) his credit card balances were higher at the end of the year than at the start, (d) he and his wife took no vacations and made no major purchases in 2014, and (e) in fact the family experienced financial hardship during the year, including failure to pay their property taxes.

Third was the evidence that Mr. Coleman had received a $150,000 personal injury settlement check during the year.  Judge Lauber put two and two together.  He noted that the money was gone by the end of the year.  That fit well with Mr. Nicely’s testimony that Mr. Coleman’s net losses for the year were likely just over $150,000.

In short, Mr. Coleman was able to give the Court facts to show both that he incurred losses and a sufficient basis for the Court to exercise its discretion under Cohan to estimate the amount of those losses to be in excess of the reported winnings.

Lesson 2: Yearly Netting
Mr. Coleman's yearly netting argument also came up aces.  Mr. Coleman was unable to show net gains or losses for any identified session.  While he was able to identify 193 days of gambling, he was apparently unable to substantiate the result for each of those days.  He apparently had no contemporaneous records.  While he was able to show how much cash he withdrew on each of those days, he did not show whether he ended each session with a win or a loss.

None of that mattered to Judge Lauber.  Go back a re-read the bolded parts of Mr. Nicely’s testimony.  Notice that it is proof of a yearly netting and not a per-session netting.  Judge Lauber seems to have accepted — sub silentio — the argument rejected in LaPlante: that if a taxpayer can substantiate a net yearly loss, then that entitles the taxpayer to a §165(d) deduction equal to the winnings reported as gross income.

That's the lesson: if you can show a net yearly loss then, logically, your failure to keep per-session records should not matter: the yearly loss would be the same yearly loss.  This yearly netting gambit won't help taxpayers who filed a return electing the standard deduction.  The reason it helped Mr. Coleman was that he had not filed a return at all and was contesting a Substitute For Return prepared by the IRS.  So he was not foreclosed from electing to itemize.

Thus, while Mr. Coleman still must report wagering gains of $350k, this decision permits him an itemized deduction of the same amount.  Note he still has the problem of an inflated AGI.  He is still being taxed on the churn.

The lesson's logic supports a more ambitious conclusion, however.  One might argue that Mr. Coleman should not have had to report the $350k as income in the first place.  He was a compulsive gambler.  Yearly netting for a compulsive gambler makes sense in the same way as the IRS per-session method makes sense.  Here's why.

The per-session method recognizes the reality that winning a single $1,200 slug in a trip the casino should not count as income because, as a practical matter, the taxpayer will plow it right back into more gambling.  The reason that the IRS gives for using the session method is basically that there is little danger of a taxpayer using interstitial winnings for anything other than gambling during a session.  The taxpayer has not exercised sufficient control over the accretion to wealth until after they leave the casino.  So the $1,200 interstitial win represents a decrease in the cost of the gambling adventure and not income.  Once the taxpayer leaves the casino with extra money jingling in their pocket, however, that is an accretion to wealth because they have control to do with that wealth whatever they please.

What is true for most taxpayers is likely not true for a compulsive gambler.  The money will not be used for anything else other than more gambling.  In effect, then entire year is one big gambling session that goes on as long as there is money to spend.  All wins are interstitial wins.  Mr. Coleman's compulsive gambling problem likely played into Judge Lauber's decision to accept the yearly netting, contrary to prior Court cases like LaPlante and IRS guidance.  But if so, then the same logic would entitle Mr. Coleman to simply not report any wagering gains since they were all wiped out.

But don't press your luck.

Coda 1:  Those interested in the topic of gaming taxation might enjoy my recent article: Taxation of Electronic Gaming,  77 Wash. & Lee L. Rev. 661 (2020).

Coda 2:  Mr. Coleman also got lucky in finding two tax attorneys at Morgan Lewis to handle his case:  Mr. John B. Magee and Mr. Eric J. Albers-Fiedler.  They did a great job in framing this as a compulsive gambling case and to hire Mr. Nicely.  I think that made all the difference.

Coda 3:  The more standard approach is the professional gambler gambit.  That is, taxpayers try to move their §165(d) deduction above the line by claiming their gaming activity is a business and not a hobby.  Websites such as this one target such taxpayers by advertising how they can help hobbyists disguise themselves as businesses.

It still comes down to luck: luck in avoiding audit, luck in drawing a sympathetic Revenue Agent if audited, and luck in drawing a sympathetic judge or jury.  That is because any activity can be a “trade or business” if the taxpayer pursues it with sufficient regularity and with the objective of making a profit.  Commissioner v. Groetzinger, 480 U.S. 23, 35 (1987)(unemployed marketer who spent entire year at dog tracks, winning $70,000 and losing $72,000 was engaged in the business of gambling).  Treas. Reg. 1.183-2(b) sets out a non-directive multi-factor test to determine whether a taxpayer is so genuinely engaged in an activity that it crosses the line from hobby to business.

Taxpayers sometimes get lucky with this framing as well.  See Myers v. Commissioner, T.C. Summary Op. 2007-194 (compulsive gambler held to be in trade or business of slot machine gaming); Le v. Commissioner, T.C. Summ. Op. 2010-94, (2010)(husband’s use of Feng Shui to determine his lucky days was a business decision (!)  Hence, slot machine winnings of over $800,000 could be offset above the line by losses of over $1 million).

Bryan Camp is lucky to be the George H. Mahon Professor of Law at Texas Tech University School of Law.  He invites readers to return every Monday for a new Lesson From The Tax Court.

Bryan Camp, New Cases, Scholarship, Tax, Tax Practice And Procedure, Tax Scholarship | Permalink


In the UK, tax on gambling winnings was abolished just after the turn of the century (2001). That also includes lottery wins. There is also no tax to be paid on playing the Forex market. Maybe the US should follow suit and make life easier for all concerned?

Posted by: Simon P | Oct 26, 2020 8:36:00 AM