UNITED STATES v. GENERAL DYNAMICS CORP.
Legal provision: Internal Revenue Code
Argument of Alan I. Horowitz
Chief Justice Rehnquist: We will hear arguments next in No. 85-1385, United States versus General Dynamics Corporation.
You may start whenever you are ready.
Mr. Horowitz: Thank you, Mr. Chief Justice, and may it please the Court.
The issue in this case concerns the timing of a tax deduction.
Respondent is obligated under its collective bargaining agreement to reimburse its employees for certain medical expenses.
The question is whether its deduction of those reimbursement payments should be taken in the year in which the reimbursement claims were filed, evaluated by respondents' claims processors, and paid, mostly 1973, as we contend, or whether, in the alternative, respondent can accelerate this deduction into the previous year by computing and deducting a statistical estimate based on its past experience that is said to represent claims that can be expected to be filed in 1972... excuse me, expected to be filed later and ultimately paid in later years for medical services obtained by employees in 1972.
That is what the Federal Circuit held.
The background of this case can be briefly stated as follows.
Prior to 1972, respondent provided medical insurance for its employees by means of contracts that it entered into with outside insurance companies.
Beginning on October 1st, 1972, respondent terminated this coverage and took upon itself the responsibility of reimbursing its employees directly for their medical expenses, though it retained the coverage and benefit scheme that had been in effect under the insurance company plans.
At the same time, respondent transferred to its own books the so-called IBNR reserve, incurred but not reported reserve, that had been established on the books of the insurance companies.
This account is a standard one kept by insurance companies, an actuarial estimate intended to reflect their potential reimbursement liability for medical expenses that have been incurred but have not yet been reported to or evaluated by the insurance company.
At the close of 1972, the estimate in this reserve account was $5.5 million.
In 1977, respondent filed an amended return for 1972 in which it sought to deduct this figure as a business expense incurred during the year 1972 on the theory that its employees' visits to the doctor created a fixed and definite obligation for respondent to make a payment, presumably in the aggregate amount of $5.5 million that they deducted, even though respondent did not in fact effect any reimbursement until an employee filed a claim and the claim was evaluated and approved in whole or in part by respondent's processors in 1973 or in the later year.
Respondent's effort to deduct the amount of its IBNR reserve is not permitted by the Code.
This is a classic example of a reserve that is generally not permitted to be deducted by ordinary taxpayers.
It is not an amount that represents a real liability.
It is an aggregate estimate based on statistical analyses of past experience of predicted future liabilities.
It is an amount that stays on the books of the respondent in perpetuity and is adjusted year by year to see how the estimate is working out, but it does not purport to be a determination of an amount of a particular liability based on the actual facts that give rise to that liability.
Now, it is okay for insurance companies to deduct such a reserve.
That is specifically provided for by the Code, and Aetna and Prudential, the companies that used to administer these plans, the companies that used to insure respondent, were entitled to take this deduction.
That is because of the special problems involved in reporting insurance company income, because they take a lot of premiums into income in Year One, and if they did not use this sort of reserve accounting system there would be a serious mismatching of income as to when the premiums are taken in earlier and the benefits are paid out later, but it is clear that this particular deduction is not available, is not available to a regular taxpayer, and respondent in fact does not even rely on the insurance company provisions.
As Justice Brandeis said in Brown v. Helvering, these insurance company provisions are special technical provisions enacted for insurance companies that do not apply to other taxpayers.
In general, reserves are not deductible by taxpayers unless they are specifically authorized by the Code.
Now, in the AAA case this Court specifically rejected the notion on which respondent relies here of using aggregate statistical estimates as a substitute for actual facts about individual expenses.
We quoted the pertinent part of that at Page 42 of our brief, and I would like to read it here.
"When a company projects an expense on a statistical basis. "
and I am now quoting,
"without regard to individual expense but consistently with overall experience, its accounting doubtless presents a rather accurate image of the total financial structure, but it fails to respect the criteria of annual tax accounting and may be rejected by the Commissioner. "
Unknown Speaker: Mr. Horowitz, that quotation might be read to refer to the second prong of the all events test, the amount, the accuracy of the estimate of the amount, and as I understand it you don't challenge that point of the decision below.
Is that correct?
You think it is wrong but you didn't raise it in your cert petition.
Mr. Horowitz: We don't challenge the accuracy of the amount, but we do challenge the method by which they have computed their deduction.
Unknown Speaker: So is it correct that the case before us would be the same from the point of view of the issue we have to decide?
If you had a small business with only one employee who got sick and you knew precisely what the amount of the medical expense was but he just hadn't filed the claim until after the first of the year.
Mr. Horowitz: No, it wouldn't be precisely the same, although we think we would win that case also, but I think that this case would be--
Unknown Speaker: What would be the difference?
What would be the relevant difference insofar as the--
Mr. Horowitz: --Well, it would depend how they... what deduction they took there, what amount they took as a deduction.
Unknown Speaker: --Well, they took the amount that was exactly the amount that the employee told them the bill was, and he showed them the--
Mr. Horowitz: That is the crucial difference, because here they don't know anything about what the actual bills were.
You see, in your case--
Unknown Speaker: --But doesn't that go to the amount question rather than the certainty of the obligation?
Mr. Horowitz: --Well, I don't think so.
Maybe I can explain it.
What I think the case is the same... what we don't rely on or what we don't quibble with here is the fact that they were... that their estimate was wildly inaccurate and was off by more than $1 million.
The Court could treat this the same as if they had in fact paid out $5.5 million, the amount that they estimated, rather than only paying out $4.4 million.
That discrepancy is not the subject of our petition.
Unknown Speaker: Aren't the two questions kind of intertwined with one another?
When you say that all the events haven't occurred, the final event would be the determination of the amount, perhaps.
Mr. Horowitz: Well, the final event here is the evaluation of the claim determining what they are going to pay.
Now, there might be--
Unknown Speaker: Yes, but your position applies equally to undisputed claims, if I understand it correctly.
Mr. Horowitz: --Well, they don't know whether the claims are disputed or not.
Unknown Speaker: But certainly there are some of the claims that are undisputed, and I am just trying to posit a hypothetical with the claims undisputed as to liability, undisputed as to amount.
The only thing that was not done is the further processing.
It seems to me your argument would be the same in that case insofar as you are talking about the first prong of the all events test.
Mr. Horowitz: Well, let me give an example where we would agree.
Unknown Speaker: Well, let me just... would your argument be the same or not in that case?
And if not, what is the difference?
Mr. Horowitz: The claims are not disputed?
Unknown Speaker: That's correct.
Mr. Horowitz: I'm sorry.
Could you ask the question--
Unknown Speaker: The amount is undisputed, the fact of the service is undisputed, everything is undisputed except he just didn't process it, didn't perform the final event of filing a written application.
Mr. Horowitz: --Well, if he has not filed the claim, our position would be the same.
Unknown Speaker: Would be the same.
Mr. Horowitz: I mean, we believe the filing of the claim is a--
Unknown Speaker: Let me put it another way.
Isn't the effect of your position to place the taxpayer on a cash receipts and disbursements basis?
Mr. Horowitz: --No, we don't think so.
In this case, because of the way they run their plan, it may have that effect, because they say that as soon as they evaluate the claim and approve it they immediately send out a check, so it may turn out to be the same thing, but they don't have to run their plan that way.
They could set up the plan... say the plan provided that General Dynamics has to make reimbursement within 90 days of when the claim is filed, and in order to maximize their interest they in fact never pay the claims until the 90 days has elapsed, even though they may evaluate them within ten days after they come in.
Now, if they had a bunch of claims that had been evaluated by the end of the year, by December 31st, and they knew what their liability, what they were going on pay on that, even though they weren't going to pay it for another three months, they could accrue those on their 1972 tax return because they would have determined the liability and all the events would have occurred that determined the amount and the fact that the liability--
Unknown Speaker: Mr. Horowitz, the AAA case, it does have some nice language in it, but really the Court wasn't addressing the kind of situation we have here.
Wasn't it the case there that it was undisputed that the services had not yet been rendered in the year in question?
They hadn't been asked for in the year in question.
Wasn't it an attempt to estimate what services would be demanded the next year, whereas here you still have an estimation, to be sure, but what is being estimated is not what will be demanded in the future but what liability has already been incurred this year, if you accept the fact that liability exists before the actual demand is made?
Mr. Horowitz: --Well, I think it is pretty similar, because in AAA they could... it was clear that services were going to be rendered some in one year and some in the next year, and they were trying to make some sort of estimate of when the services will be rendered, and they did it, and here it is clear as a matter of... or it is at least to be expected as a matter of probability that there were some people who went to the doctor during 1972 although in fact they don't know that.
As far as they know it is possible that no employee went to the doctor.
Unknown Speaker: What about the claims that have already been presented?
Mr. Horowitz: Well, first of all, they don't... there is no way of breaking down in this case which claims have been presented and which haven't, so I am not sure the Court needs to get to that, but as far as claims that haven't been presented, but if you assume that all the claims had been presented before the end of the year and just hadn't--
Unknown Speaker: Right.
Suppose we had nothing but claims that had been presented, the only step left is that the company hasn't evaluated them and decided that it will pay, but they have been presented, so the last act that the claimant has to take has been taken.
Mr. Horowitz: --Well, then you get to the question of how significant of an event this is, that they have to evaluate it, and our position is that in this case, under the plan that they have, that is still a significant event, because there is a lot of... until they evaluate the the claims and make a judgment over what should be paid there is a lot of uncertainty over what is going to be paid.
Unknown Speaker: Well, Mr. Horowitz, that wasn't the finding of the Claims Court, of course.
They said it was purely ministerial.
Was that clearly erroneous?
Mr. Horowitz: I don't, as we state in our reply brief, I don't read that as a finding of fact.
I read that as a conclusion of law as to how the all events test should be applied.
It is clear from... it is clear from the record--
Unknown Speaker: Well, it seems to... how can you say that their determination of what has to be done for processing is not a determination of fact--
Mr. Horowitz: --Well, there is no dispute over what is done--
Unknown Speaker: --as a prerequisite to then applying the all events test?
Mr. Horowitz: --I don't think there is any dispute about what is done in the processing.
They look at the claims and they determine what amount is going to be paid.
Now, the respondent's claim is that that is ministerial in the sense that it is just like plugging it into a computer.
Unknown Speaker: That was the finding of the Claims Court.
Mr. Horowitz: Well, I don't read the Claims Court finding as being ministerial in the sense that respondent tries to use it here, in the only sense that it would be relevant, which is that if they gave these claims form to any one of their claims processors or punched it into a computer that it would always spit out the same answer.
That is clearly not the case, and the testimony shows that.
There is a lot of room for judgment in what claims are going to... on what amounts are going to be paid on a particular claim, and in respect to no particular individual claim can it be said until that claim has been evaluated and approved what the ultimate liability is going to be admitted by respondent.
The testimony shows that there were plenty of lawsuits about claim denials.
Testimony shows that there were issues such as whether there was a preexisting condition, whether something was cosmetic surgery, whether something was reasonable or customary on which there were going to be disputes.
Now, it is true, of course, that a lot of the claims are going to be boilerplate ones on which there will be no dispute, but as far as the all events test is concerned, it is our position that there is a sufficient uncertainty here until the claim is evaluated and a judgment is made as to what amount should be paid, that the last event is still necessary.
I should reiterate, though, that it is really not necessary for the Court to get that far here, because it is clear that the previous step, the filing of a claim, is necessary before there can be... before there can be any obligation on respondent's part to make a payment.
If a claim is never filed they never make a payment.
If a claim is filed beyond the one-year period or 90-day period that is specified in the plan, they said they sometimes make a payment but they are not obliged to do so.
So until a claim is filed, we don't see how there could possibly be an obligation to make a payment that would justify an accrual.
And since they have not broken down their estimates in any way we think that that point alone is dispositive of the case, and it is not really necessary for the Court to decide whether the second, the final event is necessary.
Unknown Speaker: Well, it depends a little on whether you treat the filing of the claim within a certain period as a sort of statute of limitations within which it can be filed.
Or look at it in terms of some kind of waiver, I guess.
But it seems to me that at least the Claims Court said the obligation was fixed.
Mr. Horowitz: Well, again, I think that is a legal conclusion.
I understand that's what the Claims Court said, but I don't think that's a finding of fact, that the obligation is fixed.
I mean, the respondent puts that in a statement of facts and cites to a page in the record that that is the testimony of their witness that the obligation is fixed, but that is a judgment for this Court to make, not for respondent's witness.
Unknown Speaker: Mr. Horowitz, perhaps it is a little unfair to you, but it seems to me your argument about filing the claim has a certain reminition to the slot machine not being pulled yet--
Mr. Horowitz: It is not fair.
Unknown Speaker: --which I found persuasive in that case, as you may recall.
Mr. Horowitz: Well, this case is much easier for us than the slot machine case, we think.
It had better be.
One big difference between this and the slot machine case is the point I started to make a little while ago, and that is that in the slot machine case there was... the amount that the casinos were trying to accrue was based on real facts, and they were real amounts, real individual liabilities.
They would walk around the casino and look at the face of each slot machine and mark down what the amounts were on the machine, and they could say that we have a liability to pay that amount.
Now, we had a dispute in that case over whether there was still a contingency outstanding, but there is no question that they had a liability in the amount on the machine that was either fixed or almost fixed but still contingent.
Unknown Speaker: xxx they had a liability.
xxx If they had a liability if they had a liability.
The only question about it was its contingency.
I don't... that doesn't make too much sense to me.
Mr. Horowitz: They had a liability in a particular amount based on a particular fact.
Now, in this case, in this case, and I think this is a very important point, and I am not sure we made it as well in our brief as we might have, so I would like to focus the Court's attention on it, but the respondent just has, they just take a historical estimate.
They are not looking... they claim to be deducting a liability and expense, but the number that they attach to that has nothing to do with what the liability really is.
It is just based on historical evidence.
Unknown Speaker: But that gets back to the fact that I thought you didn't challenge the amount.
Mr. Horowitz: No, that's what I am trying to explain.
We don't challenge the accuracy of the estimate.
The Court can assume that the estimate has landed right on the button instead of $1 million off, but it is just impermissible under the Code just to take an estimate like this.
What they have done is, they have sort of skipped the all events test, and the all events test... I guess what I would like to convey to the Court is what a radical claim they are really making in this case.
It is one that is different from any that has ever been made in any of these accrual accounting cases before, and that is why I think we allowed ourselves in our brief to fall somewhat into the trap of just plugging this case into the all events test, but you can't really do that because the all events test presupposes that the... the ultimate question is whether the amount that respondent seeks to deduct is... in fact represents a liability and whether it represents a fixed liability, and the all events test kind of presupposes that that amount is based on these events that give rise to the liability, and then the all events test looks at those events to see whether they create a sufficiently fixed liability.
Unknown Speaker: I think what you are saying is that in these other cases what is estimated is the amount of the liability, whereas what is being estimated here is the very existence of the liability.
Mr. Horowitz: That is a very good way of putting it, Justice Scalia.
Unknown Speaker: They don't really know that any particular claim is out there.
Mr. Horowitz: They don't know that any claim is out there.
Unknown Speaker: They just estimate that a certain number of claims are out there.
Mr. Horowitz: They are just estimating historically, and they can do almost anything if they are allowed to get away with that kind of estimates.
Unknown Speaker: Yes, but these 56,000 people, there's just no realistic basis for assuming there is some liability?
Mr. Horowitz: Well, as a matter of probability, surely it is likely there is going to be some liability, sure.
Unknown Speaker: They are pretty strong probabilities, aren't they?
Mr. Horowitz: Sure.
Unknown Speaker: Like about 99 million to one?
Mr. Horowitz: But that is not... but it's misleading to... you can't just look at the all events test and say there is a liability and then pick another number out of somewhere else and attach it to that liability.
They have got to go together.
The amount that you are trying to deduct has to be connected to the liability for which you are entitled to a deduction.
This all sounds very abstract.
I would really like to try to make it a little more concrete by giving an example, but let me try to give an example of a case where I think respondent's plan would satisfy the all events test if they had computed their deduction in the proper way.
Let's assume that it was really ministerial, that the plan was just one line, and it said for every doctor bill that is submitted, every doctor bill that is submitted to us, we will pay $10 reimbursement.
Now, at the end of the year, and let's assume further that it still takes them a long time to process these claims because they like to keep all these various statistics about what is going on, but every doctor bill is going to be paid off for $10.
Well, at the end of a year they may have a stack of 100 bills that have been submitted to them, 100 claims sitting in a little desktop.
Now, if they went through and counted up those claims we would agree that they could accrue a $1,000 deduction because they would know that they have 100 claims and they are going to be paid at $10 apiece, even if they don't pay them until later, so we would agree that they could take a $1,000 deduction at that point, but we would not agree that they could go back, look at their history of what their claims were like in 1970, come up with some predicted number for what claims they were going to have sitting in their box at the end of year end 1972 and take some actuarial estimate, which might be $1,000, might be $1,100, might be $900.
They can't do that.
That is not... that is not a liability that could be accrued.
Unknown Speaker: Could I ask, suppose that this estimate... suppose you lose this case, but also suppose that their estimate was off, that they had overaccrued.
Would they have to take something back into income the next year?
Mr. Horowitz: Not necessarily.
They would not.
If the next year they were--
Unknown Speaker: Well, they took a deduction, they accrued a deduction that exceeded what their... what their actual liability was.
Mr. Horowitz: --But, you see, they are taking the deduction for the reserve.
They are not taking the deduction for the payments.
Unknown Speaker: Right.
Mr. Horowitz: So the reserve is, if the Court would hold that it was reasonably accurate, the reserve was right, and so their 1972 deduction would still be good.
Now, the reserve is kind of self-correcting over time, so the next year they would have only $4.5 million in payments and that would already leave $1 million left in the reserve.
If they found that they were consistently overestimating they would probably have to reduce the reserve in later years, and that would have the effect, especially if they had to in later years--
Unknown Speaker: They might... the next year the service might say you are accruing too much.
Mr. Horowitz: --Well, that would get back to the reasonable estimate--
Unknown Speaker: Yes.
Mr. Horowitz: --business, but I mean, the nature of reserving accounting is that if they keep accruing too much at some point they are not to be adding to the... in the first year you just take the amount of the reserve.
After that the deduction goes for an addition to the reserve to get it back up to where it belongs.
Unknown Speaker: You put your finger on it when you said a reserve is self-correcting.
Certainly a bad debt reserve is like this.
You can overreserve some time and it will correct itself in the long run.
Mr. Horowitz: Right, at some point it corrects itself, and if you have to reduce the reserves later you will take it into income in a later year, but that is--
Unknown Speaker: Mr. Horowitz, does the record tell us whether when... the first year they operated they took over the business from the insurance carriers, did they pay for the reserve in the face amount of the reserve?
Did they put out dollars equivalent to the--
Mr. Horowitz: --I think they did.
Unknown Speaker: --They did, so that--
Mr. Horowitz: Yes.
Unknown Speaker: --To go back to your simple example there where... suppose they didn't pay out $10 on each doctor's bill, there was some element of qualification in order to get paid.
They know that... and they go and count all the bills, and they have a long history that 90 percent of all of the bills are approved, and they have done that for years and years.
Or make it even 95 percent.
So they add up all of these submitted doctors' bills at the end of the year and then they try to accrue 95 percent.
You wouldn't let them do that, would you?
Mr. Horowitz: I think they could not do that, no, but that is the closest--
Unknown Speaker: Even 99 percent.
Even if they know from past experience that all except one bill will be approved in its totality, you would still say that--
Mr. Horowitz: --Well, they know as a matter of statistics?
Unknown Speaker: --in this whole stack of bills they cannot say categorically that any single one of those is definitely a claim, is definitely a valid claim, right?
Mr. Horowitz: If they can't say that any single one of them is a valid claim, then our position would be that the all events test does not permit them to do that, but I think that is the closest question in this case, and that is really the third ground on why their position is wrong, and I don't think the Court needs to get to it.
The first ground is that they are not... they just can't take an estimation like this that is not related to the actual facts involved, and the second ground is that the claim has not been filed, and if you get past those two hurdles, then I think on the third ground, although that is our position, I think that is the most difficult question that is presented to the Court.
I would like to talk a little bit also in the manner of an example about the Milwaukee case which was discussed in our brief, the case that was up to this Court once and back to the Seventh Circuit, because the factual similarities there, I think, kind of illuminate what is going on here.
The Milwaukee case involved an attempt by the Milwaukee Transit Authority to take a deduction for estimates of what it was going to have to pay out on accident claims for accidents that it had in a particular... in a taxable year.
They went through and found that they had about 400 some accident reports from their drivers.
They went through these accident reports one by one and made an estimate of what they thought their liability was on each accident report.
In fact, the record says that the attorneys were authorized to settle the cases for that amount.
And then they totaled up all their estimated liabilities for these accidents and set up a reserve account to pay for them, and tried to deduct that in that year.
The Seventh Circuit originally allowed this deduction.
It stated that it was clear that each individual liability here was not accruable because it was contingent, it was contested, a well established proposition, but it held that all of them lumped together were statistically accurate enough that they could be deducted.
Now, this Court vacated that decision in light of AAA, sent it back to the Seventh Circuit, which agreed that AAA made that decision wrong.
Now, I think what is important to note here, respondent dismisses this case by saying that, well, it is well established that these were contested liabilities, and so therefore you can't accrue them, but if respondent's position here is correct, if they can use this estimation, than Milwaukee will be able to take these kinds of deductions because they will just step it back one step the same way respondent has done here.
They won't even bother to look at each individual claim and leave themselves open to the possibility that it is contested.
They will just take some actuarial estimate of what their accident liability is based on past experience for a given year, and they will take a deduction for that, and they will be able to do it the same way that General Dynamics does it here if in fact the Court approves this method.
I would like to reserve the remainder of my time.
Argument of Lynne E. Mc Nown
Chief Justice Rehnquist: Thank you, Mr. Horowitz.
We will hear now from you, Ms. McNown.
Mr. Nown: Mr. Chief Justice, and may it please the Court, the issue in this case is when it is proper to take a deduction for the liability that the taxpayers have for medical benefits provided to their employees.
The government presupposes the answer by saying that this is a future contingent liability which we are trying to deduct at the end of 1972.
First, I would like to keep in mind the difference between the cash basis and the accrual basis of accounting.
The cash basis of accounting for a deduction requires actual payment of the claim.
That is clear.
The accrual basis of accounting doesn't require payment, it requires that there be an established liability, and that is typically prior to payment of the item.
Now, there are two parts to the all events test.
Both parts have to be met in order to take the deduction.
The first part of the test is that all events have occurred which establish the fact of liability.
The second part of the test is that the amount has to be determinable with reasonable accuracy.
They are two separate tests.
They both have to be met in order to take the deduction.
Unknown Speaker: I take it then that you accept the application of the all events test.
Mr. Nown: We do accept the application of that test.
That test was enunciated first by this Court in United States versus Anderson in 1926.
It has been with us for a very long time.
It was put in the Treasury Regulations, and it has been referred to by this Court.
Unknown Speaker: That doesn't always assure our position on it, though.
Mr. Nown: True.
But that was the birthplace--
Unknown Speaker: I just wanted to be sure of yours.
Mr. Nown: --It is clearly the test, and it was reaffirmed by this Court just last term in the Hughes properties decision.
Now, the only issue that is before this Court is the liability issue, whether all events have occurred which establish the fact of liability.
The amount issue I respectfully submit is not before this Court.
The Solicitor General is attempting after the fact to put part of that issue before this Court, saying, we object to how they computed their number.
Unknown Speaker: Well, are those two necessarily totally separate, Ms. McNown?
Mr. Nown: There is some interrelation between the tests in certain cases, but they are two distinct requirements.
You can meet one requirement of the test and fail the second.
Unknown Speaker: Supposing you are the Milwaukee Transit Company and you knew that as of December 20th, 1986, you had... your bus had been in an accident, and that you were liable but you didn't yet know the amount.
Now, would it be possible to say that all events had accrued so that you could accrue that even though you had paid nothing during 1986?
Mr. Nown: If these are uncontested claims.
In Milwaukee Suburban Transport these were contested claims.
Unknown Speaker: No, I am just using that as a hypothetical.
Let's call it the Minneapolis Transit Company.
Mr. Nown: Okay.
If you have an accident and you know you are liable and you are not contesting that liability, if you... the fact of liability has been established.
The events which establish liability happened when your bus driver ran over somebody.
That is Part 1 of the test.
Part 2 of the test is whether or not you can determine the amount of that liability with reasonable accuracy.
If you can you are entitled to deduct it.
Unknown Speaker: So in my--
Mr. Nown: Even though you haven't paid it.
Unknown Speaker: --In my hypothesis you would fail Part 2.
Mr. Nown: If you couldn't come up with a reasonable estimate of what that liability is you would not be able to deduct it because you have to satisfy both parts of the test.
Failing one, whichever it is, results in no deduction.
Now, the government in saying that they object to the method by which this is--
Unknown Speaker: Before you go on, what do you mean by an uncontested claim?
Suppose I acknowledge that I am liable in the Chief Justice's example.
I acknowledge that the company is liable for the consequences of the bus accident, but I very much contest whether particular injuries were attributable to the accident and so forth, so I really don't know what the amount... I don't... I do contest the amount of my liability.
Mr. Nown: --I think contesting the amount of the liability is sufficient contest to dispute the claim.
Unknown Speaker: To dispute the claim.
Mr. Nown: You can dispute either the basic underlying liability or the amount that you are going to have to pay to the person.
The dispute becomes on either one of those issues on a--
Unknown Speaker: What if you don't dispute it but you just don't know it?
I mean, in the example the Chief Justice gave the company just didn't know what the amount was.
I can hardly contest it until somebody makes me an offer, tells me a number that I can contest.
Mr. Nown: --The prospects of disputes in tort claims and personal injury claims are so substantial that the courts have looked very askance at those claims because they foresee contests, but we aren't talking--
Unknown Speaker: So you would treat that like a contested one, right?
Mr. Nown: --I think you would have to on a case by case basis look at those situations to see if the taxpayer never contests them and always pays out, I don't see the fact that it is a tort claim with no contest either on liability or amount would mean you could not deduct it, but that is not ordinarily what happens in tort claims.
They are usually contested in some fashion.
Unknown Speaker: Would your answer be the same even though the Minneapolis Transit Company almost invariably pays out something on all claims, but it may be a lot different than what the plaintiff is demanding.
Mr. Nown: If there are contests you cannot deduct the liability at the opening.
Unknown Speaker: Well, I suppose there are a good many medical claims that result in some kind of contest, aren't there?
Mr. Nown: The record in this case is that there are not substantial medical claims that are contested.
This is a situation where the taxpayers have put in an employee benefit plan.
They are interested in good employee relations.
They have agreed in advance.
We are willing to pay the amounts that are set forth in this plan to any employee who receives these benefits while employed by the company.
These plans are very specific.
They are very detailed.
They state, for instance, under basic medical, we will pay you $60 a day for your room and board charge at the hospital.
We will pay you $200 for an appendectomy.
The major medical benefits which are provided are equally specific.
Unknown Speaker: Are there any deductibles?
Mr. Nown: There is typically on the major medical a $100 deductible.
The government says the deductible applies across the board.
The deductible does not apply to the basic medical plan, but these plans are very specific.
It is clear to the administrators and to the employees in the booklets that are provided that the plans do not pay 100 percent of gross medical costs incurred, and therefore, while the government suggested in the trial below that the fact that the plans don't pay 100 percent of gross medical amounts, that somehow this means they are contests.
That is just not true.
It is not true as a fact.
It is not true analytically.
If somebody goes on business travel and his employer says, we pay coach air fare, and the employee goes out and says, well, I am going to go first class, if the company policy is, we pay coach, when he submits his bill and the company pays him the coach fare instead of the first class fare, that doesn't provoke litigation.
It doesn't provoke contest.
It is very clear what that plan is going to provide.
The trial court correctly analyzed this situation and determined, rightfully, we say, that not paying what is not a covered benefit is not a contest, and there is no evidence in this record that that engendered contest.
Unknown Speaker: Well, do you... is it your position that you could just use historical evidence as to what the deduction should be?
Mr. Nown: No, we do not just use historical evidence on the deduction.
Unknown Speaker: Well, do you know what medical services have been performed during the year?
Mr. Nown: We know at the end of the year that a... we know specific information for all of those employees for whom claims have been filed, and there are employees for whom a claim has been filed with the company.
It has got the name, it has got the background information and the doctor's bill.
Unknown Speaker: What about... but aren't there some people you deduct for who have just received medical services but no claim has been filed?
Mr. Nown: There are also people for whom no claim is filed.
Unknown Speaker: Do you know that they have received medical services?
Mr. Nown: Some of those people we do know have received medical services.
Unknown Speaker: But not all of them?
Mr. Nown: But not all of them.
For instance, hospitals--
Unknown Speaker: But you just estimate that there are going to be a lot of them?
Mr. Nown: --We have 56,000 employees.
We pay millions of dollars of medical benefits every year.
I submit there is no three-month period where these people are not submitting bills.
Unknown Speaker: You are guessing the claims exist.
You are not just guessing about the amount of the claims.
You are guessing on the basis of historical experience that claims exist.
You cannot categorically come in and lay your life on the line and say there are claims and here are the people that have them.
You don't know who the people are.
You don't really know for sure that the claims are there.
It is just--
Mr. Nown: --We do know many of the claims are there.
Unknown Speaker: --Yes, but you don't know--
Mr. Nown: --if people are out of work we probably know that, but knowledge is not required.
We could know it.
The facts exist.
These people, the gentleman who has been to the doctor in the first week of December for an office visit, surgical procedure has been there.
We know that we could collect that information.
If we walked the floor, like in the casino, if we walked the floor and had supervisors go out and ask every employee, have you been to the doctor, we would get answers.
We would get answers from specific people about specific doctor visits, about specific procedures which were approved.
Unknown Speaker: --Well, how did you--
Mr. Nown: And that is very different than not having any knowledge at all.
Unknown Speaker: --Well, I know, but the fact remains that when you arrived at this figure you deducted you estimated how much medical services.
You just didn't know all of the medical services that had been provided.
Mr. Nown: We didn't in fact know all the medical services provided, but that knowledge isn't required.
I think this Court's decision in Hughes--
Unknown Speaker: So you could just estimate... well, if that is true, then you don't need to know that any.
All you have to do is to pick out a historical figure.
Mr. Nown: --Well, there are two issues.
The first issue is whether the fact of liability has occurred.
Since the employees have been to the doctor, they have been to the hospital and received services, the fact of liability has occurred.
The rule doesn't require that we know each instance in which it has occurred.
For instance, in a unilateral contract we say I will pay Joe Blow to go to Philadelphia, and Joe goes to Philadelphia on the last day of the year.
When he goes to Philadelphia the contract, the unilateral contract has matured, and we have an obligation.
Unknown Speaker: So you are saying--
Mr. Nown: We may not know.
He hasn't called us up.
Unknown Speaker: --You are saying if there is a challenge to our deduction and you had a proceeding by that time you would know exactly how much medical services had been provided and you could then say that really established the fact of liability.
Mr. Nown: Oh, we know much sooner than that.
The vast majority of these claims come in in the first three months of the subsequent year.
The record shows that the payments of the claims that occurred in 1972, $4.5 million occurred in 1973, and $80,000 occurred in 1974.
These claims come in very quickly.
They are paid promptly.
Unknown Speaker: But as to some of them you don't know even by the end of the year.
Mr. Nown: That is correct.
But in Hughes Property where you had the slot machine case and the progressive jackpots, there was no winner at year end.
You had amounts which were on specific--
Unknown Speaker: But it was certain that there would be a winner.
Mr. Nown: --I think it is much more certain in this case that we... we have winners.
In this case people have been to the doctor.
They have pulled the handle.
They haven't stepped up to the window and said, pay me.
Unknown Speaker: You don't know who they are.
Mr. Nown: But they have won.
Unknown Speaker: You don't know who they are.
Of course, you didn't in the jackpot either.
The equivalent of the slot machine case here is to have knowledge that somebody has been to the doctor and has got an appendectomy which costs so much money, but not to know the name of the individual.
That is the equivalent of the slot machine case.
You don't know who the payment is going to go to, but you know that a payment is due.
Here you don't know the individual.
You don't even know for sure that there has been an appendectomy with respect to all of those for whom no claim has yet been filed.
Mr. Nown: Well, I submit we are easier than the Hughes case because we could go out and find those employees.
They are in the... in the factories, in the offices.
In the slot machine case nobody had won and you didn't know when that machine would be played and when the handle would come down.
Unknown Speaker: Do you think the slot machine case would have come out the same way if the company had in fact not gone around and looked at the machines and gotten the amount of liability off of each machine?
Mr. Nown: Going around and looking at the machines went to the second issue in that case.
It went to the amount issue.
It went to can they determine their amount with reasonable accuracy, and by reading the machines they could in that case.
But that issue is not before the Court in this case.
Unknown Speaker: 0 xxx.
I don't think at least in this case those issues are quite as sharply severable as you indicate.
I think one inevitably rubs off on the other.
Mr. Nown: Well, I think that in this case the information that we do know, the facts that do exist are sufficient that this Court can find that there is fact of liability.
The issue, we have employees who have actually been to the doctor.
They have received the services.
The plans exist.
When the claims come in it is a ministerial process to figure out the benefits.
It is not a judgmental, a difficult procedure which any claims administrator has to get into.
It is very routine.
It is very ordinary, and it is like paying the travel expense where the employee gets on the plane and has a coach air fare ticket, and comes back.
He has to submit an expense voucher.
Unknown Speaker: But could you deduct for travel expenses if you didn't know that any specific person had traveled but only knew that somewhere in your vast empire there were employees traveling, and that last year there were 15 of them that traveled, and this year there would probably be 15?
Mr. Nown: I think that goes to the issue of reasonable amount.
If you can determine with reasonable accuracy what the amount of the travel expense reimbursement is, you can deduct that amount.
If you can't determine the reasonable amount, then your deduction would fail because you haven't satisfied the second part of the test, but the fact of liability, has an employee gone out and got on that airplane, and are you going to be liable to reimburse for the airplane ticket, has occurred, and the procedures by which the processing and the amounts are going to occur have... are in place.
Unknown Speaker: What if following your advice General Dynamics does deduct for 15 travel payments in 1986 because they had 15 in 1985, and then the Commissioner comes in and says, look, we have looked over this thing and nobody traveled for General Dynamics in 1986?
Now, under your theory that wouldn't... would that invalidate the deduction or not?
Mr. Nown: I think the Commissioner would say that that accrual at the end of that year was not a reasonably determined amount and would challenge the deduction on exactly that basis, and it would fail on that basis.
Unknown Speaker: Well, Ms. McNown, beyond that he would say there was no fact of liability.
You miss the first prong if nobody got on the airplane.
Mr. Nown: No, it... all right, that is true.
Unknown Speaker: For your case, you have to assume that in fact a lot of people traveled.
Mr. Nown: Yes.
Unknown Speaker: You don't have to know about it, but at least they have to travel.
You don't have any liability to nonexistent travelers.
Mr. Nown: Yes.
But we don't have nonexistent travelers in our case.
Unknown Speaker: But if you did have nonexistent travelers you would lose.
Mr. Nown: --That is correct.
Unknown Speaker: Doesn't that suggest that we are not just talking about the estimate, but we are talking about the first prong?
Mr. Nown: --No, because here when the year ends nobody submits their expense voucher because there aren't any expense vouchers.
Here there are expense vouchers because there are employees.
We have claims that have been filed and will be completed for processing.
We have claims--
Unknown Speaker: 0 xxx.
Mr. Nown: --The claims that haven't been filed will come in.
They do come in, and we paid $4.5 million of them.
Unknown Speaker: If your answer to the Chief Justice's question about the airplane tickets is the way it is, why did the insurance companies need a special statute?
Mr. Nown: The insurance companies are a whole separate industry which has many different accounting provisions and many different income provisions.
Congress has chosen to set up a special section of the Code to deal with the whole taxation of insurance companies.
The fact that the Congress has done that does not mean that taxpayers who are not insurance companies applying the test that is applicable to noninsurance company taxpayers, namely, the all events test, cannot deduct an item because it is also an item which an insurance company might deduct.
There is no--
Unknown Speaker: But you do think it is superfluous.
You really wouldn't need it... you wouldn't need any special provision for insurance companies, would you, if we agree with your answer to the Chief Justice's question on the airline?
Mr. Nown: --On health and benefit claims I think the all events test would satisfy the liability... the reserving requirement for insurance companies.
I think that they had only the all events test for a medical benefit plan like the GD medical plan and that is what we are locking at in this case, they would not need special statutory provisions in order to authorize a deduction.
Unknown Speaker: Isn't it true that the special statute enables them to take accrual accounting on a lot of reserves that would not meet the all events test, such as your accident case in Milwaukee?
Mr. Nown: It does.
Unknown Speaker: I assume those are--
Mr. Nown: It does.
Casualty cases, contested cases, that... the regulation for insurance companies is all pervasive for that industry, and the fact that there are some items of overlap does not prove to me, I submit it is not logical to say that that means ordinary business taxpayers who satisfy the ordinary business rules can't take their ordinary deductions.
Unknown Speaker: --Wasn't there a time when insurance companies were hardly taxed at all?
Mr. Nown: I believe there was, and there has been a constant fight and reanalysis over insurance company taxation throughout the years before Congress.
I would like to point out that the government, while it says that the dollar amount of the reserve on the second issue is not before the Court, it somehow suggests that the aggregate basis of determining the reserve is.
In the petition for certiorari the government stated at Page 13, Footnote 2, we are not presenting for review here the question of respondent's ability to satisfy the second part of the all events test.
It didn't we are not submitting part of it.
It said we are not submitting the issue of their ability to satisfy the second part of the all events test.
And that is the reasonable amount portion of the test.
So the only issue that is before the Court is whether the fact of liability exists.
Now, in our case the government, the authorities of this Court support that claims processing does not preclude deductibility.
Starting with the Anderson case in 1916, there a tax was imposed on the sale of munitions, and the taxpayer said, I think that the time to deduct the tax on munitions is in 1917.
At year end 1916 the munitions had been sold, the tax statute existed, but the tax had not been assessed.
I expect the return hadn't been filed.
The year had to close before you could file your tax return.
Yet the government, contrary to its position here, said no, the munitions tax for 1916 should be accrued in 1916 even though it is not assessed, even though it is not due, and this Court agreed with that and determined that while technically tax had not been assessed and wasn't due, in a practical sense, in a realistic sense all of the events establishing that liability had occurred.
That failure to need a tax assessment shows that paper processing such as involved in this case is not the kind of paper processing which requires that fixation and liability await the end of paper processing.
The government said in its argument that it is a very difficult process, and they cited preexisting conditions and they cited prevailing rates and said in the court below we show this is a difficult process.
It is an iffy process.
The record in the court below is totally to the contrary.
Evidence was presented before the trial court on exactly what happened in claims processing.
Those very kinds of situations, prevailing rates, preexisting conditions were testified to by claims processors, and these witnesses testified that while terms which may be unfamiliar to you or I or other laymen are not unfamiliar to the claims processors and they handle these conditions on a routine basis, applying the plans in mass volume numbers, and it doesn't cause processing problems at all.
Now, the government at Page 7 and 8 of their reply brief suggest that the same example of a per diem charge for a medical benefit for a hospital stay will result in different answers if you present it to different processors.
That is just not correct.
The government looks to an example in the respondent's brief at Pages 22 and 23.
The example posed was a simple, basic medical plan that paid $60 a day for room and board charges in a hospital.
The employee is there for three days.
He is entitled to $180.
The government says, oh, another processor could come up with $210, and their citation for support of that is testimony by a witness who is talking about a plan that has not only a basic medical feature, it has also a major medical feature.
And the addition of the major medical feature means that additional dollars are paid to the employee, but if you have got two different plans you will have two different results on the same facts.
If you have the same plan, a basic medical plan, two or more processors will indeed reach the same determination as to what the benefits are that are due to that employee.
If you add another tier of benefits, you will, of course, get a different answer, but you have got a different medical plan that is at issue.
Unknown Speaker: May I ask this question?
If an employee is entitled to Medicare benefits, may he also be a participant in the plan?
Mr. Nown: Under the record in this case, Your Honor, I am not... the record is not clear on that.
There has been some developing law in terms of what coordination occurs, I think, with government plans.
But in the private sector, in the private sector there are specific rules as to who pays what benefits first, and there is coordination of benefits so there is not excess payments to employees above and beyond their out of pocket costs.
Unknown Speaker: Well, sometimes a Medicare claim is not paid or even determined for months.
Mr. Nown: There is no evidence in this record that claims processing was in any way delayed.
The claims processing was handled routinely.
There was no suggestion that claims were withheld in order to delay the payment of the liability and, as the government suggests, maybe get interest on the money.
There is no suggestion of that in the record at all.
The liabilities that are before this Court in this plan are real liabilities to real people.
The government, by saying that they want the process to wait until claims processing is completed, until the taxpayers are just ready to write the check and put it in the mail, does considerable violence to the all events test.
It substantially seeks retrenchment of this Court's decision in United States versus Anderson, and it really ignores everyday life in the world today.
There is paper processing for many, many things, and if routine paper processing of the type that was at issue in this case is required in order to take an accrual deduction, it will go substantially towards making accrual basis taxpayers function for expenses on the cash basis.
In conclusion, the benefit payments that were made in this case on a practical, reasonable basis show that the events that established the fact of liability had occurred when the employee had been to the doctor.
At that stage the taxpayers could not remove the liability.
They could not terminate the plan.
They could not reduce the benefits.
Unknown Speaker: But you also said xxx employees will have been to the doctor or were to the doctor by the end of the year.
Mr. Nown: But that goes to the second issue of the test.
The second issue requires an estimation of determination of what the amount is.
But at the end of the year, for the employee who went to the doctor in the first week of December--
Unknown Speaker: Well, it goes to liability, doesn't it?
Mr. Nown: --No, the employee is there, he has been to the doctor--
Unknown Speaker: Well, you don't know that.
How do you know that he was there?
Mr. Nown: --In the case where he hasn't filed a claim, if we don't have the knowledge we don't specifically know that, but if he has been there we cannot by January 1 say we aren't going to pay your claim because you didn't get it on file by January 1, and we are going to stop paying claims.
Unknown Speaker: Well, if the Commissioner came in and audited you and you had claimed this deduction and he looked over your books, and it turned out that all of these medical services that you had estimated had been performed never were performed, it would be like the person... like the company whose employees didn't fly.
Mr. Nown: In this case, when the Commissioner would come in and look he would not find that services were not performed.
Unknown Speaker: But if he did, if he did, he would disallow your deduction.
Mr. Nown: If he found there were no services performed, he would disallow the deduction, and since fact of liability would not have existed--
Unknown Speaker: Exactly.
Mr. Nown: --we would not be entitled to--
Unknown Speaker: But you are guessing as to how much liability there is.
Mr. Nown: --We are estimating the amount of the liability.
That is correct.
Unknown Speaker: Not just the amount, but you are estimating the existence of the liability.
Not just the amount.
Mr. Nown: I disagree, Your Honor.
I think that given all the facts that we know and the repetitive nature of these claims coming in and the size of our--
Unknown Speaker: It is a very, very sure estimate, but as to the claims that haven't been filed you are guessing that there is a liability.
It is a very, very good guess, probably, you know, 99.9999 percent accurate, but it is still an estimation about the existence of the liability as opposed to an estimation of the amount of a known liability.
Mr. Nown: --It is no more an estimation than occurred in Hughes, I submit.
In Hughes the slot machine had not been played and won.
The Nevada gaming regulations said you couldn't reverse the amount on the machine.
But there wasn't a winner.
In this case we have all these employees.
They are winners.
If these businesses went out of business, if they went bankrupt, there would be thousands of employees.
Everybody who had been to the doctor would come in and submit their claim and say to the Bankruptcy Court, pay this claim, you can't relieve the taxpayers of this liability.
Unknown Speaker: Your real point as I understand it is, you don't have to know.
The question is whether there was or was not a liability to your end.
Mr. Nown: That's correct.
For the first part of the test--
Unknown Speaker: It doesn't matter whether you didn't know it.
Mr. Nown: --For the first part of the test--
Unknown Speaker: As long as you found it out by the time you filled out your tax return.
Mr. Nown: --That's correct, and as long as... and the knowledge is not necessary for Part 1.
It may factor into our ability to determine a reasonable amount under Point 2, the amount issue of the all events test.
Thank you, Your Honor.
Rebuttal of Alan I. Horowitz
Chief Justice Rehnquist: Thank you, Ms. McNown.
Mr. Horowitz, you have four minutes remaining should you choose to use them.
Mr. Horowitz: I will use a few more, Mr. Chief Justice.
The basic issue here is whether there is a fixed obligation to pay the amount that the respondent is seeking to deduct.
That includes a couple of things, and I would like to make a couple of points.
First of all, whether or not they know, they could know something about the claims by going to see what is sitting in their in box or taking a survey of their employees is not relevant here because the number that they computed has nothing to do with any of that knowledge.
That number is purely a historical estimate and would be the same even if they had no knowledge.
Now, our position is that you cannot do that, and that you don't even get into the all events test.
That is an impermissible means of doing it.
If they did go around to all their employees and surveyed them as to what claims they had outstanding or what claims they were going to be filing next year or something, then this would be a different case.
Then we would not be raising that first objection, and then we would get into the second part of our contention, which is whether the all events test in fact is satisfied here.
And our position there is that it is not.
There is no fixed obligation to pay at the time that the employee goes to the doctor.
The employee may have an obligation to pay the doctor.
The employee may have a right to seek reimbursement from the employer, but the employer has no obligation to pay anything to the employee until the employee files a claim.
Now, there is... in a sense there is a legal liability in the way that lawyers are used to talking about it, which is that ultimately down the road there may be a lawsuit brought at which the employee will recover something and the liability that results in this recovery dates back to when he went to the doctor.
But that is not what we are talking about here.
We are talking about a fixed obligation on the part of the employer to make the payment, and that just does not exist at that time.
It does not exist until the next year.
On this business of whether we have waived our entire case by not raising the second issue in the petition, there is a big difference between the accuracy point, which is just whether the bottom line is correct or not, and the question of whether the method by which this amount is computed is legitimate, and I think as the Court has said that goes to the fact of the liability.
That goes to the essence of the all events test, and we certainly do not agree that this is a permissible method for using... for using to reach this deduction.
The all events test as first stated by Justice Brandeis, and I don't think the regulation was intended to change it, says that all the events must occur which fix the amount and determine the liability of the taxpayer to pay it.
The two are entwined together, and here the amount that they are seeking to deduct just comes out of the air.
It has nothing to do with the events on which they claim there is fixed liability.
And it has to.
In order to apply the all events test you have to have... be computing the amount based on the real events that are there.
If there are any questions--
Unknown Speaker: Yes, I do.
Mr. Horowitz: --No, go ahead.
Chief Justice Rehnquist: Thank you, Mr. Horowitz.
The case is submitted.
Unknown Speaker: The Honorable Court is now adjourned until tomorrow at 10:00.
Argument of Justice Marshall
Mr. Marshall: In No. 85-1385, the US versus General Dynamics comes to us from the United States Court of Appeals for the Federal Circuit.
The Court of Appeals there held that an accrual-basis taxpayer, that provided medical benefit to its employees could deduct an estimate of its obligation to pay the medical care obtained during the last quarter of the year.
Claims to which had not yet been filed has over close of the tax year.
In an opinion filed with the Clerk today, we reverse the judgment of the Court of Appeals.
Justice O'Connor has filed a dissenting opinion in which Justice Blackmun and Justice Stevens have joined.