Mohamed v. Exxon Corp.

796 S.W.2d 751, 1990 Tex. App. LEXIS 1368, 58 Empl. Prac. Dec. (CCH) 41,238, 53 Fair Empl. Prac. Cas. (BNA) 261, 1990 WL 75700

• Court: Court of Appeals of Texas
• Decided: June 7, 1990
• Docket: C14-89-00886-CV
• Status: Published

Opinions

OPINION

J. CURTISS BROWN, Chief Justice.

In this civil rights case the trial court ruled that res judicata barred the plaintiffs from litigating state law claims, after a federal court had dismissed an earlier suit grounded on federal law. We must determine the prior judgment’s preclusive effect under federal law. The complexity of the inquiry makes us wish there were a state-to-federal counterpart to Tex.R.App.P. 114, which provides for certification of state láw questions from federal appellate courts to the Texas supreme court; such a reverse Erie rule would let us ask the Fifth Circuit about its caselaw. Because there is no such rule, we turn to the facts.

I

In their first lawsuit plaintiffs alleged employment discrimination. They sued Exxon Corp. and Exxon Minerals Co. (Exxon) in state court under 42 U.S.C. § 1981. Exxon removed the case to federal court, and it ended up on the docket of United States District Judge Ross Sterling.

Soon thereafter the plaintiffs exhausted their state administrative remedies and procured a right-to-sue letter from the Texas Commission on Human Rights. They went once again to state court, but in the second suit the plaintiffs alleged violation of state civil rights law. See Tex.Rev.Civ.Stat.Ann. art. 5221k. The factual allegations remained essentially the same. In the second suit the plaintiffs also added a new defendant, their supervisor Roger Kust. Although the complaint in the § 1981 case had referred to Kust by name as an alleged wrongdoer, Kust was never officially a party to that proceeding. (This variation comes into play during the analysis whether preclusive effects of the § 1981 case extend to Kust in the art. 5221k case.)

Exxon again sought removal to federal court, and the art. 5221k case was assigned to United States District Judge Carl 0. Bue. At this point there existed two separate suits in federal court, each having arrived via removal from state court, each arising out of the same transaction, one stating a § 1981 claim and the other stating an art. 5221k claim. Plaintiffs filed a motion to remand in each case. Defendants went to Judge Sterling and filed a motion to consolidate the two actions into a single proceeding before him. Judge Sterling denied the motion to remand, and he refused to rule on the consolidation request. Judge Bue ultimately granted the motion to remand the art. 5221k case, but not before a complication had arisen in the § 1981 case.

It seems Judge Sterling had ordered an outright dismissal of the § 1981 case for want of prosecution. In all fairness, plaintiffs hardly lacked diligence; they were, after all, prosecuting two lawsuits at once. Instead, it was the defendants who had inadvertently failed to follow Judge Sterling’s docket control procedures. Those procedures required the defendants to deliver a docket control order advising plaintiffs of a docket call date. When the plaintiffs did not appear, Judge Sterling dismissed the action. Plainly, due process considerations would not have allowed such an outcome to stand, and the Fifth Circuit would assuredly have found an abuse of discretion if there had not been a change. See Callip v. Harris Cty. Child Welfare Dept., 757 F.2d 1513, 1519 (5th Cir.1985) (recognizing dismissal with prejudice as an extreme sanction, available only in the most egregious cases).

Interestingly, however, plaintiffs decided not to seek reinstatement of the § 1981 case, an avenue they could have taken to remedy the due process problem. They opted instead to accept the dismissal — with one proviso. In their consent to dismissal they asked Judge Sterling to recognize their lack of fault respecting the failure to appear, requesting him to label the dismissal as one without prejudice. If Judge Sterling had granted that motion to modify, this case would not be here today. Alternatively, if he had expressly designated the dismissal as one with prejudice, this case would at least be far simpler. Unfortunately, he did neither. He simply denied the motion.

At the core of the dispute before us is the question whether Judge Sterling’s dismissal of the § 1981 case was with prejudice. If it was, then the state district court in the art. 5221k case correctly regarded that prior judgment as a predicate for invocation of res judicata. On the other hand, if Judge Sterling’s dismissal was without prejudice, then there was no judgment on the merits upon which to base a finding of res judicata, and we would have to reverse.

II

A

Characterization of the dismissal in the § 1981 case is a matter of federal civil procedure. The governing rule is Fed.R. Civ.P. 41, which supplies gap-filling instructions to deal with a dismissal order that fails to specify whether the dismissal was without prejudice. Rule 41 recognizes two types of dismissals, voluntary and involuntary. Generally speaking, under rule 41(a) the plaintiff may voluntarily dismiss his action as of right, and “[ujnless otherwise stated in the notice of dismissal or stipulation, the dismissal is without prejudice.” Rule 41(b) addresses the question of involuntary dismissal, including a dismissal for want of prosecution: “[ujnless the court in its order for dismissal otherwise specifies, a dismissal under this subdivision and any dismissal not provided for in this rule ... operates as an adjudication on the merits.” The salient issue is therefore whether the dismissal was voluntary, within the meaning of federal law. Cf. Rizk v. Mayad, 603 S.W.2d 773, 775 (Tex.1980) (dismissal for want of prosecution is not considered on the merits under Texas law); Gracey v. West, 422 S.W.2d 913, 917 (Tex.1968) (same); Texas Atty. Gen. v. Daurbigny, 702 S.W.2d 298, 300 (Tex.App.-Houston [1st Dist.j 1985, no writ) (same).

The parties have cited a considerable number of relevant decisions, virtually all of them coming from the Fifth Circuit. We are quite content to confine our analysis within the boundaries drawn by the litigants in this case, because Fifth Circuit jurisprudence is sufficiently well-developed and coherent to inform our decision today. Moreover, there is every reason to regard that jurisprudence as a reliable guide for interpretation of a decree from one of the District Courts within the Fifth Circuit. Nevertheless we note that Fifth Circuit precedent as such does not bind us. What binds us is federal law — here, a federal judgment whose scope depends on federal rule 41 — because the Supremacy Clause says so. See U.S. Const, art. VI. As has long been recognized, “The laws of the United States are laws in the several States, and just as much binding on the citizens and courts thereof as the State laws are.... The two together form one system of jurisprudence, which constitutes the law of the State; and the courts of the two jurisdictions are not foreign to each other_” Claflin v. Houseman, 93 U.S. 130, 136-37, 23 L.Ed 833 (1876); see Tafflin v. Levitt, — U.S. -, 110 S.Ct. 792, 800-01, 107 L.Ed.2d 887 (1990) (Scalia, J., concurring) (on the authority of federal law in state courts). Our reading of federal law could diverge from that of the Fifth Circuit in a given case; today it does not.

At first glance one might conclude that rule 41(b) obviously applies, since that section classifies dismissals for want of prosecution as involuntary. We believe such an approach unduly facile. In this age of expanded discovery and more or less unsupervised pretrial processes, it is not always clear when it is the court taking the initiative and when it is a litigant. For example, in Plumberman, Inc. v. Urban Sys. Dev., 605 F.2d 161 (5th Cir.1979) an intervenor asked to restructure its pleadings so as to take on plaintiff status, after all other parties had dropped their claims and sought dismissal with prejudice. The trial court granted the intervenor’s motion and gave it 10 days to file an amended complaint, lest the intervention be dismissed. The intervenor never filed an amended complaint, and dismissal followed. Two years later litigation resumed but the second trial court found the former intervenor’s suit barred by res judicata, in light of the earlier dismissal. The Fifth Circuit reversed. Reasoning that the stipulation to a dismissal if an amended complaint were not filed was voluntary, the court of appeals applied rule 41(a) and held res judicata unavailable.

A contrary result occurred in Dillard v. Security Pac. Brokers, Inc., 835 F.2d 607 (5th Cir.1988). There the trial court had dismissed a litigant’s counterclaim as a sanction for failure to appear at a deposition. On appeal the court distinguished Plumberman and squarely held the dismissal to be on the merits under rule 41(b).

Appellants rely on Plumberman as authority for their argument that by consenting to the dismissal, they successfully persuaded Judge Sterling to “transform[ ] the federal court’s initial judgment of dismissal from an involuntary dismissal to a voluntary dismissal.” This conclusion they derive from his dismissal order, which expressly recognized their refusal to seek reinstatement. That assertion seems to us a non sequitur. True, Judge Sterling recited the positions of the parties; but he never grounded his refusal to modify the order on the offer of plaintiffs’ consent. Had he said, “The motion is denied because plaintiffs’ consent makes this dismissal voluntary and therefore without prejudice,” we would agree with plaintiffs now. He did not say that, and the reason is evident. If Judge Sterling had felt that way, he needed only to grant the motion for clarification. It seems far more plausible that he took the view espoused in defendants’ opposition to the motion, namely that plaintiffs should have to prosecute the § 1981 case in federal court or suffer its dismissal on the merits, since plaintiffs had filed separate suits in the first place. This position better explains denial of the motion.

In any event, we need not get mired in a slough of speculation, because there is an independent reason for regarding the order as on the merits. We simply cannot accept appellants’ contention that their post facto consent alters the involuntary character of an existing order. If rule 41 permits a you-can’t-fire-me,-I-quit argument, the pages of the Federal Reporter do not reflect that fact. It would be a strange rule indeed that allowed avoidance of sanctions by belated acquiescence. Perhaps Judge Sterling abused his discretion in refusing to modify the order, or in refusing to spell out its intended effect. But the remedy for such error was to appeal his ruling. See Nagle v. Lee, 807 F.2d 435, 443 (5th Cir.1987) (“[Plaintiff] forfeited review of that decision by not appealing the dismissal in a timely manner.”) We therefore apply rule 41(b) and find the order to be with prejudice. Accordingly, we must consider its preclusive effect under federal law.

B

Texas courts follow federal principles of res judicata for determining the effect of a prior federal judgment. Jeanes v. Henderson, 688 S.W.2d 100, 103 (Tex.1985). The Fifth Circuit requires four conditions for the invocation of res judicata:

(1) the prior judgment must have been rendered by a court of competent jurisdiction,

(2) there must have been a final judgment on the merits,

(3) the parties must be identical in both suits, and

(4) the same cause of action must be involved in both suits.

See, e.g., Dillard v. Security Pac. Brokers, Inc., 835 F.2d 607, 608 (5th Cir.1988); Nagle v. Lee, 807 F.2d 435, 439 (5th Cir.1987); Nilsen v. City of Point Moss, 701 F.2d 556, 559 (5th Cir.1983) (en banc). Before applying this test, we pause to observe that federal res judicata is an equitable doctrine, the employment of which is within the trial court’s broad discretion. United States v. Thomas, 709 F.2d 968 (5th Cir.1983); Nations v. Sun Oil Co., 705 F.2d 742 (1983); see also In re Braniff Airways, 783 F.2d 1283, 1289 (5th Cir.1986) (when “reasonable doubt exists as to what was decided in the first action, the doctrine of res judicata should not be applied.”); Kauffman v. Moss, 420 F.2d 1270, 1274 (3d Cir.1970) (same).

Moving to the merits of the preclusion argument, we find no dispute over Judge Sterling’s jurisdiction. And although there exists a dispute over whether he intended to dismiss the case with prejudice, we have already answered that question in the affirmative. The first two elements are therefore satisfied. As to identity of parties and identity of causes of action, the analysis becomes more complicated. Where the Fifth Circuit’s formulation says res judicata requires identity of parties, that does not mean identity in the usual sense. It means rather a relationship so close that a court may justifiably deny the plaintiff a second bite at the apple. In a case remarkably similar to this one, the Fifth Circuit grounded claim preclusion on a finding of sufficient identity between a corporate defendant — Exxon, as it happened — and two of its employees. Lubrizol Corp. v. Exxon Corp., 871 F.2d 1279, 1289 (5th Cir.1989). The court reasoned that Exxon’s vicarious liability derived precisely from the alleged acts of its agents, and that res judicata ought to apply. This view is the majority position among federal courts. See id. at 1288 (collecting cases); see also Pollard v. Cockrell, 578 F.2d 1002 (5th Cir.1978) (suggesting other relationships which might supply sufficient reason for finding identity: beneficiary and trustee, parent corporation and subsidiary, etc.). Furthermore, this result follows logically from Jeanes v. Henderson, where the Texas supreme court noted and relied on the federal abandonment of strict privity: “As a leading authority states, ‘[ojnce it is concluded that [res judicata] is asserted against a person who properly may be bound by the judgment, the inquiry shifts to ask whether there is some special reason for denying its benefits to a non-party.’ ” 688 S.W.2d at 105 (quoting 18 C. Wright, A. Miller & E. Cooper, Federal Practioe & Procedure § 4488, at 410 (1981)). What our supreme court correctly perceived in 1985 as a forceful development is all the stronger in the wake of Lubrizol. We therefore find identity of parties.

Nor is Nagle v. Lee, 807 F.2d 435 (5th Cir.1987), to the contrary. There the court denied a res judicata claim by two police officers after an initial suit against a named sheriff and two unknown actors, “John Doe” and “John Smith.” Because the two officers were never served in the first lawsuit, the court held they never became parties to it. Accordingly, the remaining question was whether the defendants could enjoy the effects of the prior dismissal as persons whose acts would have made the sheriff vicariously liable. To that question the court said no. Id. at 440 n. 4.

Likewise, where caselaw purports to require identity of causes of action, the scope of res judicata actually extends to all claims which “might have been presented” in the original suit. In re Reed, 861 F.2d 1881, 1382 (5th Cir.1988); In re Air Crash, 861 F.2d 814, 816 (5th Cir.1988). For example, in Langston v. Insurance Co. of N. Am., 827 F.2d 1044 (5th Cir.1987) the dispute centered around alleged employment discrimination, just as this case does. The plaintiff first filed a common law tort action for wrongful discharge. During this time he had an age discrimination complaint lodged with the EEOC. Later he received a right-to-sue letter from the EEOC but neglected to raise the age discrimination matter in the pending ease. Instead he brought a second suit, alleging only age discrimination. When the first court dismissed his tort action on the merits, the second court found his discrimination claim barred by res judicata:

Plaintiff contends that he could not have brought his age discrimination suit in Langston I because he had not completed the requisite filings with the EEOC at the time the suit was filed....

... After removal of Langston I and receipt of the right-to-sue letter, plaintiff had a fair opportunity to raise the age discrimination claim in the already pending cause of action. Instead, plaintiff filed the case sub judice resulting in two separate suits arising out of substantially the same set of facts, yet awaiting resolution by two different judges. Clearly, it is the prevention of this kind of claim splitting and “dilatoriness” which the doctrine of res judicata contemplates.

827 F.2d at 1048; see also Miller v. United States Postal Serv., 825 F.2d 62 (5th Cir.1987) (same result where the causes of action alleged sex discrimination in one case and discrimination on the basis of handicap in the other); Fleming v. Travenol Labs., 707 F.2d 829 (5th Cir.1983) (same result where the causes of action arose under Title VII in one case and 42 U.S.C. § 1983 in the other).

Lubrizol, Langston, and Fleming all make clear that failure to amend one's complaint can be held against a claimant in a court of law. For this reason we reject plaintiffs’ suggestion that res judicata cannot include claims arising after the filing of their original petition. Nevertheless we must inquire what would have happened if they had sought to prosecute the art. 5221k claim in the § 1981 case. One could argue Judge Sterling would have disallowed such amendment, refusing to exercise his discretion to hear a pendent state claim; if that is so, then the art. 5221k claim fails to satisfy the requirement that it “might have been presented.” In order to decide whether that requirement is met, we must examine the doctrine of pendent jurisdiction. Under that doctrine a court of the United States may exercise its judicial power over a claim which would not, on its own jurisdictional feet, support the invocation of federal authority pursuant to Article III. Pendent jurisdiction as a concept has fairly fuzzy boundaries, so the inquiry whether a prior judgment can cast its shadow over what would have been a pendent claim is a difficult one. The Texas supreme court laid out the pertinent standard in Jeanes v. Henderson.

There the court unanimously adopted as the rationale for its decision comment e to § 25 of the Restatement (Second) of Judgments (1982). 688 S.W.2d at 104 n. 4. Comment e provides for an exception to the rule of preclusion in one circumstance:

If, however, the court in the first action would clearly not have had jurisdiction to entertain the omitted theory or ground (or, having jurisdiction, would clearly have declined to exercise it as a matter of discretion), then a second action in a competent court presenting the omitted theory or ground should not be held precluded.

What is notable about this formulation is its placement of the burden to establish whether or not the federal court would have exercised pendent jurisdiction. It starts from the premise that the federal judge would have adjudicated the state law claims. Only if the federal judge would “clearly” have declined to exercise pendent jurisdiction is preclusion inappropriate. Accordingly, in cases of total silence, the state court is left to presume res judicata applies. This rule might not be intuitively obvious, but it is supported by the overwhelming weight of authority. See Anderson v. Phoenix Inv. Counsel of Boston, 387 Mass. 444, 440 N.E.2d 1164, 1168-69 (1982); Merry v. Coast Comm. Coll. Dist., 97 Cal.App.3d 214, 158 Cal.Rptr. 603, 613 (1979); Maldonado v. Flynn, 417 A.2d 378, 383-84 (Del.Ch.1980); see also McLearn v. Cowen & Co., 48 N.Y.2d 696, 422 N.Y.S.2d 60, 61-62, 397 N.E.2d 750, 751-52 (1979) (discussing the rule), rev’d on other grounds, 60 N.Y.2d 686, 468 N.Y. S.2d 461, 455 N.E.2d 1256 (1983).

Applying this standard, we look to the procedural history of the § 1981 case in Judge Sterling’s court. To be sure, this is not an instance of total silence. Plaintiffs point out Judge Sterling’s refusal to rule on a motion to consolidate. They regard his refusal as tantamount to declining pendent jurisdiction over the state law claims. In addition, they say Judge Bue declined pendent jurisdiction by remanding those claims to state court. Thus it is argued that both federal courts clearly turned down the opportunity to entertain the art. 5221k case, and that we should not now find it barred by res judicata. We disagree.

When Judge Bue found removal improvident and ordered a remand, he did so not because he wanted to, but because he had no choice. Far from declining to exercise pendent jurisdiction, he had absolutely no jurisdiction to decide the matter at all: there was no controversy which would support federal jurisdiction in the first place, and therefore there was no core claim from which a state law claim could “pend.” And when Judge Sterling would not consolidate the cases into his court, he was hardly drawing a clear line so as to reserve some issues for the state judiciary. He was simply doing nothing. Furthermore, plaintiffs were not the ones who sought consolidation; it was the defendants who asked Judge Sterling to consolidate, over the vigorous objections of the plaintiffs. We would have a different case if the plaintiffs had sought to amend their complaint in the § 1981 case and add the art. 5221k theory. That did not occur. We cannot say it is clear, then, that Judge Sterling would have refused to hear the state law claims. For this reason we must consider the art. 5221k case as one which “might have been presented” in the § 1981 proceeding, and res judicata applies.

Ill

At this juncture we should distinguish between two types of uncertainty which haunt cases like this one — where a state court must determine the effect of a prior judgment from federal court, and where a state law theory was omitted from the federal case. The first type of uncertainty is the sort we have encountered just above, in deciding whether a federal court “would clearly have declined to exercise” pendent jurisdiction. That is plainly a matter of judgment for the United States district court, a matter which affords the federal judge a zone of discretion, so our job becomes somewhat more difficult. Nevertheless it is our duty to find preclusion unless the federal judge would clearly have declined the pendent claim, as is explained by comment e of the Second Restatement’s § 25 and Jeanes v. Henderson. The second type of uncertainty relates to the scope of the prior judgment. This uncertainty grows out of the ever-present problem of interpreting words within their context. Whereas the first variety comes from the peculiar nature of pendent jurisdiction (and can come up only when a federal court rendered the initial judgment), the second derives from the ordinary difficulty of construing a written instrument. This latter kind of uncertainty is far less problematic. It poses a problem only in the rare case where a reasonable person has genuine cause for doubt over what the first court decided. It is this sort of uncertainty that leads to language such as “if reasonable doubt exists as to what was decided in the first action, the doctrine of res judicata should not be applied.” In re Braniff Airways, Inc., 783 F.2d 1283, 1289 (5th Cir.1986). The “reasonable doubt” standard has nothing whatsoever to do with the problem of pendent jurisdiction.

As we have explained, the types of uncertainty differ, both in source and in effect. The only uncertainty today is the type involving pendent jurisdiction, and the rule of Jeanes v. Henderson requires us to resolve that issue against the plaintiffs. With this background we can consider plaintiffs’ penultimate argument that application of res judicata would be inequitable and unfair. See, e.g., United States v. Thomas, 709 F.2d 968, 972 (5th Cir.1983). And to that argument we cannot improve upon the words of the United States Supreme Court:

we do not see the grave injustice which would be done by the application of accepted principles of res judicata. “Simple justice” is achieved when a complex body of law developed over a period of years is even handedly applied. The doctrine of res judicata serves vital public interests beyond any individual judge’s ad hoc determination of the equities in a particular case.

Federated Dept. Stores, Inc. v. Moitie, 452 U.S. 394, 401, 101 S.Ct. 2424, 2429, 69 L.Ed.2d 103 (1981). The result today comes as a direct outcome of calculated procedural machinations at the trial level, machinations which contravened no ethical canon but which render hollow any complaint that equity ought to forbid the reaping of what has been sown.

The final point of error challenges the trial court’s action in ruling on the motion for summary judgment without the benefit of oral argument. We find no error in the court’s action; but even if we did, we perceive no harm. Summary judgment practice is decidedly a textual affair rather than a matter of oral advocacy. See City of Houston v. Clear Creek Basin Auth., 589 S.W.2d 671, 677 (Tex.1979) (“To permit ‘issues' to be presented would encourage parties to request that a court reporter record summary judgment hearings, a practice neither necessary nor appropriate to the purposes of such a hearing.”). We overrule all points of error.

Affirmed.