Oral argument — Timothy O'Brien v. Caterpillar Inc.

Our second case for the morning is O'Brien v. Caterpillar. Mr. Keller. Thank you, Judge. May it please the court. Taken together, I think it's fair to say that the district court's holdings make the ADA disparate impact remedy that the Supreme Court created in Meacham and at Smith all but useless. The court started off by holding that if a policy borrows its eligibility criteria from another policy or practice, it ceases to be a policy. There is no authority, zero authority for that holding, and it has absurd implications. Suppose Katz said, we won't let anybody get promoted who is eligible to retire under our retirement policy. That would obviously be a policy about promotions, but according to the district court, it wouldn't be, because it borrows its eligibility criterion from a different policy. Next, the court held that because the sub-liquidation policy resulted from a single decision, and because it wasn't repeated once the fund was distributed, it wasn't a policy. I think everyone agrees that this court held to the contrary in Council 31. Caterpillar says Council 31 is no longer good law, but neither the Supreme Court nor this court has overruled it or hinted that it should be overruled. If an employment program has the earmarks of a policy, it doesn't matter that it resulted from a single decision, because most policies do, and it doesn't matter that the program isn't repeated once its objectives are accomplished. The district court's reasoning would overrule almost every decision that held that a one-time hiring exam, for example, was unlawful under a disparate impact theory, because it wasn't repeated. Now, it's true. Some actions, like tearing down a dilapidated housing project in the Joliet New West case, they really are one-off, factually dependent decisions, and as the opinion itself put it in that case, they aren't part of a policy at all. But the sub-liquidation program was part of a policy. It was part of a company-wide policy to get rid of sub-programs wherever they existed, and it wasn't a one-off event. Over a four-month period, it took applications from a large group. It applied a written rule to grant or deny a unique money benefit, and we know exactly what it was within that policy that caused the disparate impact against older employees. Caterpillar hasn't suggested what purpose of the ADA would be served by holding that this wasn't a policy simply because it wasn't repeated once the fund was distributed, and none would be. Does anything depend on the existence of a fund? Suppose Caterpillar just made an offer out of its own corporate treasury that anybody who takes early retirement, anybody who's eligible for retirement now and takes early retirement, gets $35,000 cash. Right. Is that a violation of the ADEA? Absolutely not. There's nothing wrong with an early retirement fund. Why is that different? Why does it matter what the source of the funds is? It doesn't matter at all. What matters is that when the fund is in the normal situation that you posit, Judge, when a fund's being distributed to people who want to retire, there is no discrimination in the program. You have an option. You can retire and get money, or you can not retire and not get money, but nobody is discriminated against on the basis of age for that program. What brought the discrimination in was that after the original decision was made to have an incentive retirement program used from the fund, another group was brought in which was unconditionally eligible for that, and that's the decision that created the disparate impact. That's where I have some trouble. There is no disparate treatment in the program as I identified it, but there's certainly a disparate impact between those who go and those who don't, and you're making a disparate impact claim. That's why I asked whether you're contending that all early retirement incentives create disparate impacts and violate federal law. No, I don't think that is our theory of disparate impact. The disparate impact is that there are two groups, one predominantly younger, one predominantly older. One group has to incur a cost no matter what choice it makes. Either it has to give up the distribution by saying no to retirement or it has to give up their job. The other group, the younger group... All early retirement incentives create exactly that system. No, because in an early retirement incentive, whether you retire has nothing... You're all eligible to retire. You're all of the same age group. There's no age distinction there. People have a choice. But if the cash benefit is paid only to those who take early retirement within a window... And in my hypothetical, there's no similar offer to anybody who's younger. Correct. There was no offer to the younger group. It was just given to them. They got money unconditionally. The older group had to make a choice, and either choice that they made, they incurred a cost, a big cost, either the cost of foregoing about $38,000 or the cost of giving up their jobs. When you impose a cost on the older group, no matter what they do, but you don't impose that cost on the younger group, no matter what they do, that's the disparate impact, and it's a real adverse impact. If you just have a simple retirement program, you don't have that feature, Judge. Who are the plaintiffs in this case? The plaintiffs are some 48 people of the group who were retirement eligible and chose to keep working. And chose to keep working. Correct. So a lot of the arguments and the particular stripe of argument you were just making makes it sound like the plaintiffs are all of those who are retirement eligible, whether they took the deal or not. But I don't see how you can make an argument on behalf of those who took the deal. They aren't plaintiffs. Absolutely. That would be like arguing in a testing case that the people in the minority group, and there would be a disparate impact shown on pass rates. It's true. Some people in the minority group will have passed the test. And so in that sense, they may not have been adversely affected. But nobody would say that the statistical analysis you do to show the adverse impact should exclude anybody in the minority group who passed the test. It's the pass rates of minorities versus the pass rates of non-minorities. And so when we're looking for adverse impact here, Judge, what we do is we take the people who were adversely affected by the choice that they had to make, and we see what the age composition of that group was, and we compare it with the age composition of the people who weren't adversely affected no matter what choice they made or who didn't have to make a choice at all. And when you do that comparison, you find that everybody in the older group suffered relative to everybody in the younger group, no matter what choice they made. People who decided to retire gave up their jobs. There's no reason to assume that they would have done that had we not structured the program the way they did because we normally assume that we structure incentive retirement programs to induce people to retire who otherwise wouldn't. So when you make that comparison, everybody in the retirement-eligible group is worse off than everybody in the non-retirement-eligible group, no matter what choice they made. And that's essentially what this court held in the Solon case. Let me go to the RFOA issue. The RFOA issue, it seems to me, it's simple. The court's error was very simple. It didn't analyze the reasonableness of the decision that was essential to creating the adverse impact that we complain of. All the court did was to say, look, it's reasonable to abolish the subfund, and it was reasonable to use its money for an incentive retirement program. But abolishing the subfund didn't discriminate against anybody on the basis of age. Sure, the union didn't like it, but they had the right to do it. Moreover, simply having an incentive retirement program, which was their original decision, that would not have discriminated against plaintiffs on the basis of their age. That's the statute, and that's this court's and every other court's decision. It's just a program. What did create the disparate impact was the later decision to bring in non-retirement eligible people and to bring them in on better terms than the retirement eligible people did. That's a decision that had to be justified to carry the burden that Meacham imposed on Caterpillar. Now, why did they make that later decision? The modification decision, as we call it. There's no dispute about why it made it. The Catt negotiator who developed this idea, Dave Stevens, testified why. He said, look, and it wasn't because Catt thought that non-retirement participants had to share in the distribution. In other places they didn't, and in the original proposal they didn't. And it wasn't because Caterpillar wanted to treat non-retirement eligible people equally with retirement eligible people because it didn't treat them equally. It was an ad hoc labor negotiation reason. Caterpillar was demanding tough concessions in these bargaining negotiations, and a strike was looming. And it realized that four of the six people on the negotiating committee were non-retirement eligible. And Stevens thought, maybe if we extend this liquidation to non-retirement eligibles, maybe it'll soften up the committee. And in two words, you can say that Catt's justification for a modification that imposed a cost on an older group that wasn't imposed on the younger group was, we think it would achieve labor peace. Now, can this reason justify affirming a summary judgment that the district court, we think, entered for a wrong reason? And it can't. Because a jury could find, and we think it would have to find, that the modified proposal hurt, not helped, Catt's goal of labor peace and that Catt knew it was hurting and stuck to it anyway. Pleasing a union, for starters. Pleasing a union strikes us as a shaky justification for programs that disadvantage an older group relative to a younger one. The law is clear that agreement by a union to a discriminatory program is not a defense, and if that's so, why should pleasing a union be a defense to a disparate impact claim? But in any case, there wasn't any evidence that the modified proposal pleased a union and a lot of evidence that it helped produce a strike, not prevent it. So here, Catt basically has two arguments to you. First, it implies that the union invited this modified proposal, and it relies on one offhand comment by one bargaining committee member early in the game when the proposal was first announced to use the whole fund for a retirement incentive program. But when the committee responded a week or so later to that modified proposal, it said no, it thumpingly rejected it, and both Stevens and the union's negotiator have testified that the union never indicated that bringing non-retirement eligible people into the liquidation plan would make it acceptable. The union objected to any distinctions in distributing this fund, and even after the strike began, the union tried to get Caterpillar to back off of that feature that produced this lawsuit. And the second thing that Caterpillar says is, well, the modified proposal must have helped get a contract because, look, the union ultimately accepted it. And that's a non-sequitur. The union ultimately agreed to every one of Catt's give-back demands, but that doesn't show that any of those demands was helpful in achieving labor peace. It just means that Caterpillar got what it needed by beating a strike. And if the modified proposal had helped to achieve labor peace, there would be evidence to show it. It would be easy for Caterpillar to carry its burden, if that evidence existed. We have statements by everybody who said anything in labor negotiations for months. Well, Catt argues, well, look, once we made this proposal and the union rejected it, did we really have to withdraw it or change it? And the answer is simple. It did have the right to bargain as hard as it liked. We don't say the contrary. But if the proposal that you're determined to cram down is one that is harder on older than younger employees, and you stick to it, and you succeed in getting the union to accept it with everything else, you take the risk that when the smoke clears, people who were prejudiced by it are going to sue you. And when they do, if your argument is that you should get a pass because the proposal helped to achieve labor peace, you need to be able to prove that, because Meacham makes it your burden to prove the RFOA. So this court should reverse, and it should remand, either with instructions to grant summary judgment for plaintiffs, because Catt declined to do any kind of justification of its real reason for this, or to hold a trial where Catt will have to prove that it really was, to a jury's satisfaction, that it really was, its modified proposal really was reasonably designed to achieve labor peace. That's what we ask the court to do. Thank you. Mm-hmm. Mr. Torres. May it please the court, good morning. Plaintiffs try to make much of this idea that a single event cannot somehow constitute evidence of a disparate impact. They're arguing that Council 31 somehow needs to be overruled. In fact, the distinctions here that are at issue really are ones of the facts of the particular cases and what particular evidence was in the particular case. So in City of Joliet, we have a particular decision to tear down one building. In Ilhart, we had a decision to lay off one part-time attorney. In Bennett, it was one screening panel that was all white. The common feature in all of those cases is that there was no evidence that that conduct was repeated at any other point in time. The situation here involving Caterpillar's decision to liquidate the Joliet subfund falls within the same factual purview. In fact, the union admits as much in its statement of facts where it indicates, or it admits, that this was the only instance where Caterpillar sought to liquidate a subfund and use the distribution of those funds to incentivize retirement. So the question here is whether it lines up with the City of Joliet set of facts or the Ilhart set of facts or the Council 31 set of facts. The Council 31 set of facts are markedly different because in that case, the evidence was that the state routinely conducted reductions in force pursuant to formulas and guidelines that existed. And so in that case, there was evidence that the particular incident that the court was being faced with was being repeated over and over again. There is no such evidence in this case. In fact, we have a situation where once the subfund was liquidated, the agreement by which that liquidation occurred had no more force in effect, there was no more funds to distribute, and there's no evidence of Caterpillar ever engaging in similar conduct with respect to any other group of individuals. So the... If you take the classic disparate impact case, the very beginning, back to Griggs against Duke Power, suppose Duke Power had had a high school graduation requirement and it used it for only four months. Would that mean that it was not, in fact, the kind of thing for which disparate impact analysis was appropriate? If it was used for four months over a period of time, I could see an argument that it was affecting people, it was a policy that they intended to rely upon. I guess the question would be, under the facts that you pose, Judge Easterbrook, whether it was intended to achieve some discrete event or whether it was a policy... That was just a hiring rule. They wouldn't hire anybody who lacked a high school degree. It was in force for four months. And the rule is challenged on the basis it effectively screened out black applicants for employment. I think the evidence of a four-month policy that was intended to screen out applicants would give rise to a claim under... No, no, no, no, no. I'm asking about disparate impact. If it was intended to screen out blacks, you've got a completely different case. I'm asking about disparate impact, about a requirement which they thought was suitable for business reasons, but it turned out to have this big effect. Would you say, well, it was in force for only four months, therefore no disparate impact analysis? I don't think the duration necessarily determines whether it's a disparate impact case or not. The one-off decision in City of Joliet to tear down or decide to tear down an apartment building presumably could have occurred over a period of time, and yet this court said it was a one-off event, it wasn't pursuant to some policy... Yes, alas, the litigation about that case has dragged on for more than a decade, but it was still a one-off decision. Yes. But, you know, this may be very hard to figure out what a one-off decision amounts to. I have a different kind of question for you. In the Kentucky retirement case, the Supreme Court essentially said that how the ADEA treats pensions requires an analysis of the three separate pension clauses in the ADEA. Are you relying on any of those three clauses? No, Your Honor. Relying upon the principle... But that's what the Supreme Court said we should turn to and think about. When trying to figure out how pension decisions interact with the ADEA, the justices say, think about those three clauses, and you're not. At least I understand that. Well, the principle from Kentucky retirement that we thought was particularly useful here was this idea that the correlation between age and years of service doesn't necessarily mean that you can't take into account one consideration without inexorably considering the individual's age. The problem with this argument is that plaintiff's case, as I understand it, depends on the fact that all of the older workers, but none of the younger workers, was required to give up something. They had to give up current income in order to get the $37,000. I don't think that's the kind of thing that's sheltered by any of the three pension clauses in the ADEA. I think that the distinction here, though, was not that there was a decidedly younger population and a decidedly older population. Almost 90% of the people who were eligible to participate in the pension plan became eligible simply by years of service alone. So the distinction that the plaintiffs are trying to make, that somehow this is an older versus younger group, is not bore out by the actual demographics of the group. Then why aren't you relying on section 623L1AI? But you're not. I mean, that's the main pension clause in ERISA. I'm sorry, in ADEA. Too many employment statutes. I guess the reason we haven't focused as much on the pension piece of this is that it was simply borrowing the eligibility formula in order to identify the population that was going to have to retire pursuant to this retirement incentive. So the distinction that was being drawn here, years of service, is what defines who had to elect to retire and who was not required to retire, because they were ineligible to do so. So the idea that there was a choice to be made here, however distasteful that the plaintiffs may have viewed it, was a distinction that was being drawn based upon their years of service, not age. So when the disparate impact... All right, so let's take it that way. Let's suppose that Caterpillar said, if you are too young to retire, we will add $40,000 to your account. If you are old enough to retire, we will add $20,000 to your account. Stop. That's all that's going to happen. Is that shielded from challenge under the ADEA? If the distinction that was being drawn was based on age alone, like the Solon case, Your Honor, then it would not. I specified what the facts were. If you're too young to retire, if you're old enough to retire. In your view, is that then protected from challenge under the ADEA? No, Your Honor, it's not. But that's not the distinction that the facts here show is being drawn. That's why this is called a hypothetical. Going back to the... Even if it's a one-off decision, that's actionable. Well, again, the question of whether or not the distinction that's being drawn, like the Solon case, was a situation where they had an ongoing practice of drawing distinctions between age. So in that situation, at least there was a finding that what you're really relying upon is something that's age-based. And they specifically distinguish in that case the treatment of individuals based upon things like years of service. So if you couple that with the idea that it was only occurring in one instance, I think you still have the problem of them having decided to enact a policy. Again, under the hypothetical, we don't know how long it was going to... There's nothing in the statute or the case law that requires that a policy be in existence for a particular duration before it becomes actionable. Well, the Inclusive Communities case and the City of Joliet also... I mean, they just stand for the proposition that it's harder to prove the claim if it's a one-off decision, not that there's some requirement that the policy that's being challenged be in existence for a specific period of time. I'm sorry, yes, you're right. You're not pushing that hard on this. It's not a question of duration. It's a question of, in the case of these RIFs that are at issue in the Snead case and the Green case, they found that there was no evidence that they had actually ever applied these criteria to anyone else. This was a unique set of employees that was at issue in Green. So again, you can take a long time to tear down an apartment building. It may take four months to distribute money, but the duration itself is not the key to whether or not it's a one-time event or a discrete event. It may be difficult to determine which one falls on which side. But here, the facts show that the Caterpillar specifically did this, eliminated the fund, the agreement had no more force in effect, and the plaintiffs admit that there was no other situation where Caterpillar had ever allowed to do this, we would suggest fall more squarely within the framework of City of Joliet and the Snead and the Green case, as opposed to Council 31, where, as I said before, there was evidence that, in fact, that policy was being repeated time and time again. With the time that I have remaining, I just wanted to talk a little bit about the RFOA arguments. A lot of what plaintiffs... If you could use real words rather than obscure acronyms, it would help. I apologize. The reasonable factor... We're not specialists. I apologize, Your Honor. With respect to the reasonable factor other than age argument, the evidence, in fact, shows, contrary to Mr. Gallen's argument, that the proposal by Caterpillar to include the nonretirement-eligible individuals certainly was reasonable to the extent that Caterpillar... The evidence shows Caterpillar was reacting to a statement by the other side that the proposal to only give the money to the retirement-eligible individuals shortchanged the nonretirement-eligible individuals. The question here was whether it would be reasonable in response to that statement, however obscure or brief it might have been for the company to say, in order to get a deal, I am going to propose that those individuals also share in the distribution of this fund. The subsequent facts here actually show the union repeatedly... I'm not sure where this is taking us. Come back to my hypothetical in which Caterpillar offers $40,000 each to the people who aren't eligible to retire and $20,000 each to those who are. And it just says, we're going to put them in your account. And Caterpillar says, the reason we did this is not that we dislike older workers, but the union insisted on it. Is that a reasonable factor other than age? Under your hypothetical, I'm not just so sure it is, Your Honor, but I don't know that those are the facts that this record shows here. The question of whether it was reasonable for the company to want to try and accomplish... But the question I'm asking is whether if you're thinking about whether something is a factor other than age. Does it matter that the company was trying to achieve labor peace or to get the union to do something that it didn't really want to do so Caterpillar had to structure this the way the union was happy? If in the end what it structured was disparate treatment by age, it sounds like there's a problem. If it turns out that there is disparate treatment by age, there may be a problem. But here there's no evidence that anything that Caterpillar sought to achieve during all of the bargaining that ensued was because of age. Forget about what it sought to achieve. For that matter, forget about disparate impact for the moment. If you make an offer that is more valuable as an actuarial matter to younger workers, those not eligible to retire, than it is to older workers, because the older workers get the $37,000 only if they give up something in exchange. And the younger workers get the $37,000 without giving up anything. Why isn't that disparate treatment and we then just stop right there? Well, if you're asking whether this could be analyzed under a disparate treatment framework, yes, Your Honor, but that doesn't answer the question as to whether or not they've stated a claim for disparate impact. It may be, and I'm about to ask Mr. Gallant whether he's making a disparate treatment argument. Well, a lot of the horribles and a lot of the claims that he say will flow from the district court's dismissal do sound in disparate treatment. They all talk about intent to gerrymander or intent to favor one group versus the other. Disparate treatment does not turn on intent. Right? If you take my hypothetical, $40,000 to one group, $20,000 to the other, and the only reason they did it was to make the union happy, not because they wanted to make the older workers worse off, but just to make the union happy. Still, the disparate treatment question is, did you intend to do what you're doing? And is it disparate treatment? Yes, Your Honor. And in my hypothetical, yeah, the Caterpillar intended to do what it was doing. The question is, yes, whether or not, though, in that hypothetical, whether the reason that it wanted to placate the union was somehow... No, no, no, no. I don't want to have a long discussion of this, but if you think about some of the famous cases like the Johnson Controls case, where women but not men are subject to some employment requirement, and the employer says, we're not doing this because we dislike women. We're doing it because they could be pregnant and lead is in the air and it's a teratogen. And the Supreme Court said, that's still disparate treatment. The reason you may be doing it may be noble, but it's disparate treatment and it's not within any defense in the statute. That's the kind of question I'm asking. Is this maybe disparate treatment? And then we need to figure out whether it's within a defense in the statute. You're entitled, as long as we keep offering you opportunities, you can talk. I was mindful of the time, Your Honor. Your opportunity won't last forever. I'm sure it won't, Your Honor. Whether or not there is an ability for Mr. Gallen to prove a disparate treatment claim would turn on a whole myriad of facts as to what was prompting the company to act, whether or not the articulated nondiscriminatory reason for why it acted was pretextual. And so is it possible to construct a theory under which the payment of money to one group versus the other somehow becomes evidence of disparate treatment? I'm sure that there's a fact pattern out there that it could, but it doesn't answer the question as to whether our burden to articulate a reasonable factor under age under a statute which is supposed to have a higher burden for proof has been met. And we would submit, Your Honor, respectfully that it was. Okay. Thank you very much. Thank you. Anything further, Mr. Gallen? Just a couple of things. I really have to step up because you have a question. Yes. Well, my question for you is, are you making a disparate treatment claim? The answer is we did not. That's the honest answer. We thought long and hard about this, but the Hayes and Paper case and the Kentucky retirement case, which are disparate treatment cases, discouraged us from doing that, and we didn't. Okay. The two points I want to make are this. What we seem to be getting on this prima facie case argument is you have to repeat a program in order for it to be a practice. And my question remains, Title VII doesn't say it is unlawful to discriminate if you repeat the discrimination. There's no repetition requirement anywhere in Title VII, and it would serve no purpose of the statute to hold that a four-month program where the statistics are clear and undisputed and where the disparate impact is clear and undisputed. It would serve no purpose to hold that repetition is necessary. With respect to the RFOA, Mr. Torres said that the evidence shows that Catt was reacting to a comment by one of the bargaining committee members, and that's why it modified the program. That's not what the evidence showed. We called Mr. Stevens as the company's 30B6 witness, and we asked him point-blank, it's quoted word for word in the brief, why did you do this? When it dilutes the whole purpose to add non-retirement-eligible people, why, nonetheless, did you add them to do this? And he didn't say, we did it because the union said it really wanted it. He didn't say that at all. He said, no, we saw that four of these guys on the bargaining committee were non-retirement-eligible, and we thought that that might entice them to go along. They had their chance to say that, but they didn't, because it wasn't true. And the last thing is this. There was a hint of the correlation argument that Mr. Torres made, which is in his brief, and he says, look, the factor that produced the disparate impact wasn't age itself, but something correlated with age, and retirement eligibility can't be that factor. But that's contrary to Meacham. What Meacham says is the obvious thing. In every disparate impact case, the factor that produces the disparate impact is not age itself or race itself or sex itself. It's something correlated, and why should there be an exception for retirement eligibility? Thank you very much. Thanks to both counsel. The case is taken under advisement.