Kathy SMITH, Plaintiff, Appellant, v. F.W. MORSE & CO., INC., Defendant, Appellee.
In this appeal, the plaintiff invites us to overrule the district court's adverse decision under Title VII of the Civil Rights Act of 1964, 42 U.S.C. §§ 2000e-2000e-17 (1988) (Title VII), and to reinstate her common law causes of action for breach of contract and wrongful discharge. We decline the invitation in all its aspects.
We chronicle the events that preceded the filing of suit and then recount what transpired thereafter.
A. Chronology of Events.
Damar Plastics & Metal Fabricators, Inc. (Damar) operated a job shop in Somersworth, New Hampshire, where it crafted custom components for high-technology applications. Plaintiff-appellant Kathy Smith joined Damar in 1976 and advanced steadily through the ranks until she reached the position of production manager almost a decade later. In that capacity, Smith scheduled production runs and coordinated delivery dates. In late 1987, after an imbroglio with Darrol Robinson (Damar's owner and general manager), she requested and obtained reassignment to a different post having no responsibility for production scheduling.
On December 23, 1988, defendant-appellee F.W. Morse & Co., Inc. (Morse), a firm owned by Chris Bond, acquired Damar's business and assets. Damar then had fewer than forty employees, including seven managers reporting directly to Robinson: Michael Hickman (production control); Robert Lane (shipping); Ronald Paradis (production/machining); Marc Shevenell (production/sheet metal); Gary Bickford (engineering); Michael Seeger (sales); and Smith. Though not titled, Smith testified that she was considered to be a de facto manager who, largely because of Hickman's inadequacies, performed many of the duties of the production control manager.
Bond promptly concluded that Damar had too many chiefs and too few Indians. Within days of the closing, he fired Hickman. Then, in concert with Maryann Guimond, the new general manager (who had authority to hire, fire, and discipline personnel), he interviewed a number of employees, including Smith. In the aftermath of this review, the company cashiered Lane. To fill the void created by the two executive-level departures, Morse promoted Smith to the newly created position of materials manager, consolidating responsibilities for scheduling, production control, inventory control, purchasing, shipping, and receiving that had previously been spread among three managers.
All told, Morse's initial reorganization efforts substituted Guimond for Robinson and pared second-echelon management from seven to five. In addition to Guimond, the reconfigured management team comprised Paradis (machining); Shevenell (sheet metal); Bickford (engineering); Seeger (sales); and Smith (materials). In recognition of Smith's increased responsibilities, Morse twice hiked her pay (once in January and again in March), thus increasing her weekly stipend by roughly twenty-five percent.
At about the time of the takeover, Smith informed Bond that she had become pregnant and would need a maternity leave. Morse, a tiny company, had no formal maternity leave policy. Bond nonetheless honored Smith's request and assured her that her position was “secure.” In preparation for her leave, Smith held several meetings with Guimond, Shevenell, and Paradis. The company temporarily distributed her managerial duties among other supervisors and arranged for a newly-hired secretary, Kelly Gilday, to perform her clerical functions. Along the way, Guimond informed Smith that either Paradis or Shevenell likely would be discharged, and told her that she would be promoted again upon her return from maternity leave. Guimond also indicated that, in all probability, Bickford would be demoted, and Smith would be asked to assume a portion of his duties. While these changes presumably would warrant increased remuneration, Guimond did not mention an amount.
On April 7, 1989, Smith began her maternity leave, planning to return to work in approximately six weeks. She gave birth two weeks later. Meanwhile, Guimond, expecting the “sky to fall,” held regular “reality check” meetings with Shevenell and Paradis. To her surprise, the plant functioned very well.1 Guimond reported the good news to Bond.
Smith visited the plant on May 1 and informed Guimond that she wished to return to work one week earlier than originally anticipated. Guimond inquired about whether Smith desired more children, and Smith replied affirmatively. The following day, Guimond queried Karen Vendasi, Smith's sister and co-worker, about Smith's plans to have a larger family. Vendasi relayed this conversation to Smith and told her of nascent rumors to the effect that she might not return to work. Smith contacted Guimond and demanded an explanation. Guimond denied any knowledge of the rumors, dismissed them as idle buzznacking, and again assured Smith that her job was secure. Guimond repeated these assurances during a chance meeting on May 4.
A few days later, Guimond concluded that the materials manager's position was superfluous and decided to eliminate it. She told Smith of her decision on May 11. During this telephone conversation, Guimond asked Smith if she preferred people to be told that she had decided to stay at home with her infant child rather than that she had been discharged. Smith rejected the suggestion. Nevertheless, a Morse employee repeated this canard to several customers.2
Following Smith's severance, Guimond gave most of her duties to Paradis in his new capacity as operations manager. Shevenell assumed the role of manufacturing manager (in charge of both machining and sheet metal work). Guimond also promoted two lower-ranking employees, Peter Lapanne and Brian Hoffman, to assistant manager positions (though evidence adduced at trial demonstrated that Lapanne had been an assistant manager as far back as 1984, and that neither man assumed any new responsibilities or received any salary increase in connection with his new title). Gilday continued to perform the clerical functions associated with Smith's former position. When the second round of the reorganization wound down, the plant had three second-echelon managers-Paradis (operations); Shevenell (manufacturing); and Seeger (sales)-in lieu of the original seven.
B. Procedural History.
Smith sued Morse in a New Hampshire state court alleging, inter alia, wrongful discharge based on gender discrimination, intentional infliction of emotional distress, and breach of contract. Morse removed the case to federal district court on the ground that Smith's claim “arose under” Title VII, thus prompting federal question jurisdiction. See 28 U.S.C. §§ 1331, 1343(a)(3), 1441, 1446; see also 28 U.S.C. § 1367 (conferring ancillary jurisdiction over appended nonfederal claims). Smith thereafter filed an amended complaint that made her Title VII claim explicit.
Early in the proceedings, Morse moved for partial summary judgment. The district court (Stahl, J.) granted the motion on the common law wrongful discharge and emotional distress claims. See Smith v. F.W. Morse & Co., No. 90-361-S, slip op. at 12 (D.N.H. Sept. 26, 1991) (unpublished) (Smith I ).
Several years later, the parties simultaneously tried the Title VII claim to the bench (McAuliffe, J.) and the breach of contract claim to a jury.3 At the close of the plaintiff's case, the district court entered judgment as a matter of law in the defendant's favor on the breach of contract claim and disbanded the jury. The Title VII case proceeded before the district judge. Morse asserted that it scrapped the materials manager's position and laid off the appellant as part of an overarching strategy to streamline a top-heavy managerial structure, and that even if Smith had not been on maternity leave she would have been flattened by the downsizing steamroller. The district court agreed and entered judgment accordingly. See Smith v. F.W. Morse & Co., 901 F.Supp. 40, 45 (D.N.H.1995) (Smith II ). This appeal ensued.
II. THE TITLE VII CLAIM
The crown jewel of the appellant's asseverational array is her contention that the district court erred in finding that Morse did not discriminate against her on the basis of her sex. Our appraisal of this contention is in three parts.
A. Standard of Review.
Following a bench trial, the court of appeals reviews the trier's factual determinations for clear error, see Cumpiano v. Banco Santander P.R., 902 F.2d 148, 152 (1st Cir.1990); Fed.R.Civ.P. 52(a), but affords plenary review to the trier's formulation of applicable legal rules, see Johnson v. Watts Regulator Co., 63 F.3d 1129, 1132 (1st Cir.1995). The jurisprudence of clear error constrains us from deciding factual issues anew. See, e.g., Jackson v. Harvard Univ., 900 F.2d 464, 466 (1st Cir.), cert. denied, 498 U.S. 848, 111 S.Ct. 137, 112 L.Ed.2d 104 (1990); Keyes v. Secretary of the Navy, 853 F.2d 1016, 1019 (1st Cir.1988). Indeed, we may not disturb the district court's record-rooted findings of fact unless on the whole of the evidence we reach the irresistible conclusion that a mistake has been made. See Cumpiano, 902 F.2d at 152; RCI Northeast Servs. Div. v. Boston Edison Co., 822 F.2d 199, 203 (1st Cir.1987).
This deferential standard extends not only to factual findings simpliciter but also to inferences drawn from the underlying facts. See Cumpiano, 902 F.2d at 152. Similarly, findings regarding an actor's motivation fall within the shelter of Rule 52(a), and, therefore, if the trial court's reading of the record on such an issue is plausible, appellate review is at an end. See Foster v. Dalton, 71 F.3d 52, 56-57 (1st Cir.1995); Anthony v. Sundlun, 952 F.2d 603, 606 (1st Cir.1991).
B. The Jurisprudence of Title VII.
Title VII provides, inter alia, that it is an unlawful employment practice for an employer to discharge an individual because of her sex. See 42 U.S.C. § 2000e-2(a)(1). After the Supreme Court held that this phraseology did not proscribe discrimination on the basis of pregnancy, see General Elec. Co. v. Gilbert, 429 U.S. 125, 145-46, 97 S.Ct. 401, 412-13, 50 L.Ed.2d 343 (1976), Congress augmented Title VII by enacting the Pregnancy Discrimination Act of 1978 (PDA), Pub.L. 95-555, § 1, 92 Stat. 2076, 2076 (1978) (codified at 42 U.S.C. § 2000e(k)). The PDA made clear that:
The terms “because of sex” or “on the basis of sex” include, but are not limited to, because of or on the basis of pregnancy, childbirth, or related medical conditions; and women affected by pregnancy, childbirth, or related medical conditions shall be treated the same for all employment-related purposes, including receipt of benefits under fringe benefit programs, as other persons not so affected but similar in their ability or inability to work.
42 U.S.C. § 2000e(k). Thus, at the time Smith and Morse parted company, Title VII's ban on gender discrimination encompassed pregnancy-based discrimination.
Like other Title VII plaintiffs, an employee claiming discrimination on the basis of pregnancy may proceed under either a disparate treatment or a disparate impact theory. See generally Furnco Constr. Corp. v. Waters, 438 U.S. 567, 575, 579-80, 98 S.Ct. 2943, 2948-49, 2950-51, 57 L.Ed.2d 957 (1978) (explaining the dichotomy). Here, the appellant alleged disparate treatment. Consequently, she had the burden of proving that the defendant purposefully terminated her employment because of her pregnancy.
In cases predating the Civil Rights Act of 1991, see supra note 3, the framework for proving intentional discrimination varies depending on the availability of direct evidence. See Fields v. Clark Univ., 966 F.2d 49, 51-52 (1st Cir.1992), cert. denied, 506 U.S. 1052, 113 S.Ct. 976, 122 L.Ed.2d 130 (1993); Cumpiano, 902 F.2d at 153. Absent the evidentiary equivalent of a “smoking gun,” the plaintiff must attempt to prove her case by resort to a burden-shifting framework. See Texas Dep't of Community Affairs v. Burdine, 450 U.S. 248, 254-56, 101 S.Ct. 1089, 1094-95, 67 L.Ed.2d 207 (1981); McDonnell Douglas Corp. v. Green, 411 U.S. 792, 802, 93 S.Ct. 1817, 1824, 36 L.Ed.2d 668 (1973). Under this framework, a plaintiff can establish a prima facie case of pregnancy discrimination by showing that (1) she is pregnant (or has indicated an intention to become pregnant), (2) her job performance has been satisfactory, but (3) the employer nonetheless dismissed her from her position (or took some other adverse employment action against her) while (4) continuing to have her duties performed by a comparably qualified person. See, e.g., Cumpiano, 902 F.2d at 153; Lipsett v. University of P.R., 864 F.2d 881, 899 (1st Cir.1988). Establishing the prima facie case raises a rebuttable presumption that discrimination sparked the adverse employment action, see Cumpiano, 902 F.2d at 153, and imposes upon the employer a burden to put forward a legitimate, nondiscriminatory motive for the action. See Burdine, 450 U.S. at 254-55, 101 S.Ct. at 1094-95; Lipsett, 864 F.2d at 899. If the defendant clears this modest hurdle, the presumption of discrimination vaporizes, see Mesnick v. General Elec. Co., 950 F.2d 816, 823 (1st Cir.1991), cert. denied, 504 U.S. 985, 112 S.Ct. 2965, 119 L.Ed.2d 586 (1992),4 and the plaintiff (who retains the ultimate burden of persuasion on the issue of discriminatory motive throughout) must then prove that the employer's proffered justification is a pretext for discrimination, see St. Mary's Honor Ctr. v. Hicks, 509 U.S. 502, ----, 113 S.Ct. 2742, 2749, 125 L.Ed.2d 407 (1993); Mesnick, 950 F.2d at 823-24.
On the relatively rare occasions when a smoking gun is discernible-that is, when a plaintiff produces direct evidence that the protected characteristic was a motivating factor in the employment action-the McDonnell Douglas framework is inapposite. See Fields, 966 F.2d at 52. In those cases, direct evidence of discriminatory motive-say, an admission by the employer that it explicitly took actual or anticipated pregnancy into account in reaching an employment decision-serves to shift the burden of persuasion from employee to employer. The latter must then affirmatively prove that it would have made the same decision even if it had not taken the protected characteristic into account. See Price Waterhouse v. Hopkins, 490 U.S. 228, 258, 109 S.Ct. 1775, 1794-95, 104 L.Ed.2d 268 (1989) (plurality op.); id. at 265-67, 109 S.Ct. at 1798-1800 (O'Connor, J., concurring).
The seeming neatness of this dichotomy is illusory in certain respects, for evidence rarely comes in tidy, geometrically precise packages. In many cases, the line between McDonnell Douglas, on one hand, and Price Waterhouse, on the other hand, is blurred. In those situations, classification depends on both the quantity and quality of the proof that a court deems sufficient to constitute direct evidence of discriminatory animus.
Discretion is sometimes the better part of valor, and courts often wisely decide to sidestep difficult theoretical questions if answers to them are not essential to the proper resolution of a given case. We have here a good example of such a prudential approach. The trial court largely bypassed any differential direct evidence/circumstantial evidence tamisage, preferring to go directly to a finding that, on the totality of the evidence presented, Morse had proven that gender discrimination did not trigger the firing. See Smith II, 901 F.Supp. at 44-45. This approach negates any need for us to pursue the question of an analytic framework to a definite conclusion. While we agree with our concurring colleague that the decisional process is important, there comes a point at which slavish insistence upon process for its own sake serves only to exalt the trappings of justice over its substance. Here, the district court's finding on causation, if sustainable, resolves the Title VII claim whether the appellant's prima facie case arises under the McDonnell Douglas or Price Waterhouse paradigm. And as we illustrate below, see infra Part III(C), that finding passes muster.
C. The Merits.
Consistent with the district court's approach, Morse must be assumed to have had the burden of proving that it would have taken the same action-the elimination of the materials manager's position-whether or not the appellant became pregnant, took a maternity leave, or planned to bear more children. The court found that Morse carried the devoir of persuasion on this pivotal issue. It concluded that Morse's decision was “motivated by business judgment and represented an effort to economize by placing the most qualified personnel in the fewest number of managerial positions possible, and was not based on plaintiff's gender, pregnancy, or her expressed desire to have more children.” Smith II, 901 F. Supp. at 44. The court also concluded “that even if Guimond is assumed to have considered impermissible gender-based factors, the same decision to eliminate plaintiff's position would still have been made at the same time” for reasons of business necessity. Id. The crux of our inquiry is whether these findings are clearly erroneous.
There is little doubt that an employer, consistent with its business judgment, may eliminate positions during the course of a downsizing without violating Title VII even though those positions are held by members of protected groups (pregnant women included). See, e.g., LeBlanc v. Great Am. Ins. Co., 6 F.3d 836, 844-45 (1st Cir.1993), cert. denied, 511 U.S. 1018, 114 S.Ct. 1398, 128 L.Ed.2d 72 (1994); Goldman v. First Nat'l Bank, 985 F.2d 1113, 1118-19 (1st Cir.1993); Montana v. First Fed. Sav. & Loan Ass'n, 869 F.2d 100, 105, 107 (2d Cir.1989); Dister v. Continental Group, Inc., 859 F.2d 1108, 1115 (2d Cir.1988); Pearlstein v. Staten Island Univ. Hosp., 886 F.Supp. 260, 268-69 (E.D.N.Y.1995). This is merely a reflection of a central theme that permeates the relevant jurisprudence: insofar as Title VII is concerned, an employer can hire or fire one employee instead of another for any reason, fair or unfair, provided that the employer's choice is not driven by race, gender, pregnancy, or some other protected characteristic. See Foster, 71 F.3d at 56; Keyes, 853 F.2d at 1026; see also Freeman v. Package Mach. Co., 865 F.2d 1331, 1341 (1st Cir.1988) (elucidating similar proposition in ADEA case). The flip side of the coin, however, is that an employer who selectively cleans house cannot hide behind convenient euphemisms such as “downsizing” or “streamlining.” Whether or not trimming the fat from a company's organizational chart is a prudent practice in a particular business environment, the employer's decision to eliminate specific positions must not be tainted by a discriminatory animus. See Goldman, 985 F.2d at 1118 n. 4; Maresco v. Evans Chemetics, 964 F.2d 106, 111 (2d Cir.1992); Mesnick, 950 F.2d at 825; Pearlstein, 886 F.Supp. at 268-69.
Against this backdrop, we believe that the evidence adequately supports the trial court's findings. When Morse took over, Damar had an inordinately high ratio of managers to workers and the managers' responsibilities overlapped.5 Both Bond and Guimond testified that from the very start they believed that Damar's sprawling organizational structure defied rhyme or reason. Accordingly, they set out to compress some of the sprawl. The district court credited their intention, noting that the witnesses' actions matched their stated objective. More to the point, Guimond testified that she terminated the appellant “because I had a position that I no longer felt needed to be filled.” Bond testified in the same vein, indicating that he, too, had become convinced that Smith's position was expendable. The court accepted this evidence, concluding that the materials manager's position would have been eliminated within the same time frame whether or not Smith had taken a maternity leave.
In our view, this determination, while not inevitable, is supportable. In the first place, the record strongly suggests that, in fact, the position was expendable. In the second place, any other choice would have entailed a loss of engineering expertise that Damar could ill afford.6 In the third place, the court's view is bolstered by the reception that the appellant originally received from the new ownership. Bond and Guimond apprised her of the planned downsizing and assigned her significant new responsibilities when other managers were dismissed. They also promoted her and increased her compensation. These actions, undertaken with full knowledge that the appellant was pregnant and would be taking a six-week maternity leave, are inconsistent with a bias against pregnant employees. In the fourth place, the district judge, sitting as the trier of fact, had the right to credit Bond's testimony that the “maternity leave never played a role in itself” because the same decision “would have been made in a very close time frame,” and Guimond's testimony to like effect. In a bench trial, such credibility judgments are the judge's prerogative. See Anthony, 952 F.2d at 606.
To be sure, the record could support a less innocuous conclusion. The chronal proximity of Guimond's questions anent Smith's plans to have more children and her dismissal, Guimond's ill-advised suggestion that customers and employees be told that Smith decided to stay at home to care for her daughter, and Smith's termination while on maternity leave are troubling-so much so that we, if free to write a palimpsest, might have characterized the impetus behind the appellant's ouster differently. But whether the trial court could have drawn an inference of discriminatory intent is not the test. See Foster, 71 F.3d at 55; Keyes, 853 F.2d at 1027. As long as a contrary inference is also supportable-and that is the situation here-then it is for the trial court, not the court of appeals, to call the tune. After all, “when there are two permissible views of the evidence, the factfinder's choice between them cannot be clearly erroneous.” Johnson, 63 F.3d at 1138 (citing Anderson v. City of Bessemer City, 470 U.S. 564, 574, 105 S.Ct. 1504, 1511, 84 L.Ed.2d 518 (1985)).
In an effort to evade the force of this principle, the appellant hauls two further arguments from her bag. First, she asseverates that Morse did not in fact eliminate her position, and that the district court's contrary finding, see Smith II, 901 F.Supp. at 43, is itself clearly erroneous. This asseveration leads down a blind alley.
When an employer defends an employment discrimination case on the ground of position elimination, the position may not, like a Dali painting, fade from one image to another only for the first image to reemerge at the blink of an eye. See Gallo v. Prudential Residential Servs., Ltd. Partnership, 22 F.3d 1219, 1226-28 (2d Cir.1994); LeBlanc, 6 F.3d at 846; Barnes v. GenCorp. Inc., 896 F.2d 1457, 1465 (6th Cir.), cert. denied, 498 U.S. 878, 111 S.Ct. 211, 112 L.Ed.2d 171 (1990). Yet, a position elimination defense is not defeated merely because another employee, already on the payroll, is designated to carry out some or all of the fired employee's duties in addition to his own, or because those duties are otherwise reallocated within the existing work force. See LeBlanc, 6 F.3d at 846; Barnes, 896 F.2d at 1465. The elimination of a position signifies the employer's belief that it can get by with one less helper; it does not necessarily convey a belief that the work the employee had been doing was superfluous and need not be performed at all.
Here, the undisputed evidence before the district court indicates that after Guimond dismissed Smith, the position that Smith had occupied-materials manager-fell into desuetude. There is no basis in the record for a suggestion that Lapanne or Hoffman assumed any of the appellant's former duties; those duties, which Paradis, Shevenell, and Gilday had performed during Smith's leave, continued to be performed by them (or, at least, by Paradis and Gilday). In short, the second round of the reorganization (which cost Smith her job) bore a striking resemblance to the first round (which gave Smith her promotion to materials manager). Given these facts, the district judge's determination that Morse eliminated the appellant's position is unimpugnable.
The appellant next endeavors to surmount the sharp escarpment of the clearly erroneous rule by casting a hook at the legal standard applied by the trial court. This is a theoretically sound way to climb the mountain, see, e.g., Reliance Steel Prods. Co. v. National Fire Ins. Co., 880 F.2d 575, 577 (1st Cir.1989) (explaining that appellate courts review questions of law de novo, even after a bench trial), but in this case the hook does not hold. The appellant's thesis is as follows. She says that Title VII prohibits an employer from dismissing an employee while she is on maternity leave even if the employer, in the process of rationalizing its work force, discovers that her position is redundant and eliminates it for that reason.
Refined to bare essence, this thesis suggests that, since Morse would not have discovered the redundancy at that time (if ever) but for the fact that Smith took a maternity leave, the leave brought about the firing.7 And the appellant attempts to drive this point home by citing Bond's testimony that “because” Smith was out on maternity leave, Morse was able to discover that her position was expendable-testimony which the appellant optimistically equates with an admission that Morse dismissed her “because” of her pregnancy. With respect, we believe that this argument, which seeks to apply a black-letter legal principle in a totally mechanical fashion, plays mischievously on the mendacity of language by substituting sound for sense.
It is settled under Title VII that an employer may not discharge an employee based on the categorical fact of her pregnancy. See Newport News Shipbuilding & Dry Dock Co. v. EEOC, 462 U.S. 669, 684, 103 S.Ct. 2622, 2631-32, 77 L.Ed.2d 89 (1983); Cumpiano, 902 F.2d at 153. By the same token, since a short-term inability to work is bound up with the very nature of pregnancy and childbirth, that disability is a pregnancy-related condition within the meaning of 42 U.S.C. § 2000e(k), and Title VII thus prohibits an employer from dismissing an employee in retaliation for taking an authorized maternity leave. Nevertheless, under the PDA, pregnancy does not confer total immunity.8 An employer may discharge an employee while she is pregnant if it does so for legitimate reasons unrelated to her pregnancy. See, e.g., Troupe v. May Dept. Stores Co., 20 F.3d 734, 738 (7th Cir.1994); Pearlstein, 886 F.Supp. at 268-69; see also Lipsett, 864 F.2d at 899 (holding that an employer may dismiss an employee who is in a protected class for a nondiscriminatory reason); Johnson v. Allyn & Bacon, Inc., 731 F.2d 64, 70 (1st Cir.1984) (similar). It follows, then, that an employer may discharge an employee while she is on a pregnancy-induced leave so long as it does so for legitimate reasons unrelated to her gravidity.
Harmonizing these principles leads to the following conclusions. Title VII mandates that an employer must put an employee's pregnancy (including her departure on maternity leave) to one side in making its employment decisions-but the statute does not command that an employer bury its head in the sand and struthiously refrain from implementing business judgments simply because they affect a parturient employee. See Troupe, 20 F.3d at 738 (holding that the PDA “requires the employer to ignore an employee's pregnancy, but ․ not her absence from work”); Crnokrak v. Evangelical Health Systems Corp., 819 F.Supp. 737, 743 (N.D.Ill.1993) (stating that “the PDA does not force employers to pretend that absent employees are present whenever their absences are caused by pregnancy”). At bottom, Title VII requires a causal nexus between the employer's state of mind and the protected trait (here, pregnancy). The mere coincidence between that trait and the employment decision may give rise to an inference of discriminatory animus, see St. Mary's, 509 U.S. at ----, 113 S.Ct. at 2747, but it is not enough to establish a per se violation of the statute (at least when, as now, the justification advanced by the employer in support of the employment decision is on its face legitimate and nondiscriminatory).9
To sum up, an employee (pregnant or not) runs a risk of suffering the ordinary slings and arrows that suffuse the workplace every day she goes to work and every day she stays away. Title VII is neither a shield against this broad spectrum of employer actions nor a statutory guaranty of full employment, come what may. Applying the PDA as the appellant asks would eliminate an employer's business necessity defense-long recognized under Title VII-and cripple industry's ability to manage workers in keeping with nondiscriminatory considerations. That is not the law. See Bowen v. Valley Camp of Utah, Inc., 639 F.Supp. 1199, 1204 (D.Utah 1986) (explaining that Title VII, as amended by the PDA, does not “preclude an employer from articulating legitimate nondiscriminatory reasons for terminating a woman while she was on maternity leave”); see generally Blackie v. Maine, 75 F.3d 716, 723 (1st Cir.1996) (suggesting, in retaliation case, that “[a] contrary rule would mummify the status quo”).
Here, the district court found the requisite nexus lacking between the employer's mindset and the employee's gravidity. In the court's estimation, Morse discharged the appellant for nondiscriminatory reasons. The record permits that view of the facts. That the discharge took place while the appellant was on maternity leave possessed considerable evidentiary significance-but that circumstance neither transformed the character of the employer's action nor rendered it per se unlawful under Title VII. The district court therefore did not apply an erroneous legal standard.
III. THE BREACH OF CONTRACT CLAIM
We turn now to the appellant's partially tried breach of contract claim. At the close of her case, the trial court took this claim from the jury and directed a verdict in Morse's favor. The appellant assigns error.
A. Standard of Review.
The court of appeals reviews the grant of a motion for judgment as a matter of law de novo, applying the same legal principles that inform the trial court's ruling. See Rolon-Alvarado v. Municipality of San Juan, 1 F.3d 74, 77 (1st Cir.1993). Accordingly, we “examine the evidence and the inferences reasonably extractable therefrom in the light most hospitable to the nonmovant.” Fashion House, Inc. v. K mart Corp., 892 F.2d 1076, 1088 (1st Cir.1989). If the proof, eyed from this standpoint, permits a reasonable factfinder to reach only a conclusion favorable to the movant, then the court must remove the issue from the jury's consideration. See id.
While this approach does not allow the court to “consider the credibility of witnesses, resolve conflicts in testimony, or evaluate the weight of the evidence,” Wagenmann v. Adams, 829 F.2d 196, 200 (1st Cir.1987), neither does it pave the way for every case, no matter how sketchy, to reach the jury. Thus, “a mere scintilla of evidence is not enough to forestall a directed verdict, especially on a claim or issue as to which the burden of proof belongs to the objecting party.” Fashion House, 892 F.2d at 1088.
B. The Merits.
The parties-who concur on very little else-agree that New Hampshire law governs the breach of contract claim. Under that law, the at-will status of an employment relationship is “one of prima facie construction.” Panto v. Moore Business Forms, Inc., 130 N.H. 730, 547 A.2d 260, 267 (1988). That is to say, unless an employment relationship explicitly provides for a definite duration, it is presumed to be at-will. See Butler v. Walker Power, Inc., 137 N.H. 432, 629 A.2d 91, 93 (1993) (explaining that the at-will presumption “is a gap filler for determining duration when the parties' contract of employment is silent as to its expiration”). This is critically important when an employee challenges her ouster; an employer can give an at-will employee-even one who has been a stellar performer-her walking papers at any time, for any reason or no reason, unless a statute, a collective bargaining agreement, or some aspect of public policy proscribes firing the employee on a particular basis. See Panto, 547 A.2d at 267.
Of course, an employer and an employee may alter the at-will status of the employment relationship. See Butler, 629 A.2d at 93; Panto, 547 A.2d at 267. Such a modification sometimes may be accomplished if the employer makes a binding offer that the employee can accept by remaining on the job. See Panto, 547 A.2d at 265. Standard contract formation principles govern the creation and construction of such contracts. See id. at 264. Thus, the “offer must be so definite as to its material terms or require such definite terms in the acceptance that the promises and performances to be rendered by each party are reasonably certain.” Chasan v. Village Dist. of Eastman, 128 N.H. 807, 523 A.2d 16, 21 (1986) (quoting Restatement of Contracts § 32 (1932)).
Definiteness, like beauty, is frequently in the eye of the beholder. At best, it involves matters of degree. In the last analysis, the standard is reasonable certainty, not mathematical precision. See Sawin v. Carr, 114 N.H. 462, 323 A.2d 924, 926 (1974). The provisions of a contract need only be “sufficiently certain to allow claims of breach to be resolved readily, and to enable a reasonably certain computation of damages.” Panto, 547 A.2d at 264 (internal citations omitted); accord Phillips v. Verax Corp., 138 N.H. 240, 637 A.2d 906, 910 (1994); Sawin, 323 A.2d at 926.
In this instance, the appellant takes bits and pieces of various conversations that she had with Guimond and Bond, pastes them together, and argues that a rational jury, mulling the ensuing patchwork, could conclude that Morse offered to reinstate and promote her following her maternity leave. By continuing her employment in the wake of such promises, her thesis runs, she accepted the offer. The district court did not buy the patchwork, remarking in its ore tenus ruling that “the promises described by the evidence are of insufficient definiteness to be enforceable, do not modify the at-will employment relationship, [and are such] that any calculation of damages or any identification of breach would be impracticable if not impossible.” We agree with the lower court that the terms of the alleged contract are too indefinite to raise a jury question.
We start by attempting to decipher the true nature of the appellant's claim. Her lawyers tell us that the disjointed statements made to her (e.g., “don't worry, we will manage while you are on maternity leave, your job is secure,” “you will assume more responsibilities on your return,” you are “wanted back”) created a contract to reinstate her following the completion of her maternity leave. Yet, the appellant concedes that Bond's and Guimond's statements did not alter the durational component of the at-will employment relationship. A contract to reinstate an at-will employee to an at-will position (from which she could immediately be removed without cause) is no contract at all. See Light v. Centel Cellular Co., 883 S.W.2d 642, 645 n. 5 (Tex.1994) (holding that, as long as the at-will character of the employment relationship remains unchanged, any “promise made by either employer or employee that depends on an additional period of employment is illusory because it is conditioned upon something that is exclusively within the control of the promisor”); E. Allan Farnsworth, Contracts §§ 2.13, 2.14 (2d ed. 1990) (explaining that promises to maintain an at-will relationship are illusory); cf. Butler, 629 A.2d at 94 (terming an analytically equivalent argument “a thin reed”).
Nor is this the only shortcoming in the supposed contract for reinstatement. The evidence also fails to establish either the nature of the position Smith was to assume or her proposed rate of pay. These gaps seemingly foreclose a reasonably certain computation of damages.
Concluding, as we do, that the alleged contract for reinstatement is too indefinite to be actionable does not end this phase of our inquiry. In stark contrast to the reinstatement theory proffered by her counsel, the appellant's own testimony indicates that she understood the statements made to her as promises of employment “indefinitely,” and as constituting an abiding “commitment to a permanent position with F.W. Morse that would never end.” If, by this, she means to suggest a contract for lifetime employment, her claim also founders.
Although tangentially related New Hampshire precedents exist, the state supreme court has not explicitly addressed the contours of contracts for lifetime employment. We are nonetheless confident that the court would adopt the prevailing view of such matters. See generally Kathios v. General Motors Corp., 862 F.2d 944, 949 (1st Cir.1988) (explaining that a federal court, called upon to determine state law in the absence of direct in-state precedent, may look, inter alia, to cases in other jurisdictions); Moores v. Greenberg, 834 F.2d 1105, 1107 (1st Cir.1987) (similar). That view regards such contracts as out of the ordinary, and insists that an offer for lifetime employment must be expressed in clear and unequivocal terms to be enforceable. See, e.g., Williamson v. Sharvest Mgmt. Co., 187 W.Va. 30, 415 S.E.2d 271, 274 (1992); Rowe v. Montgomery Ward & Co., 437 Mich. 627, 473 N.W.2d 268, 273 (1991); Vance v. Huff, 568 So.2d 745, 749 (Ala.1990); Shebar v. Sanyo Bus. Sys. Corp., 111 N.J. 276, 544 A.2d 377, 381-82 (1988); Degen v. Investors Diversified Servs., Inc., 260 Minn. 424, 110 N.W.2d 863, 866 (1961). Measured by this yardstick, the representations made by Morse do not stand sufficiently tall to confer lifetime employment. See, e.g., Williamson, 415 S.E.2d at 275-76 (finding employer's statement that it would “take care of” employee insufficiently definite to alter at-will employment); Skagerberg v. Blandin Paper Co., 197 Minn. 291, 266 N.W. 872, 874 (1936) (finding that the terms “permanent employment,” “life employment,” and “as long as the employee chooses” established only an at-will contract); Aberman v. Malden Mills Indus., Inc., 414 N.W.2d 769, 771-72 (Minn.Ct.App.1987) (concluding that the statement “we are offering you security” only indicated an at-will employment relationship).
IV. THE WRONGFUL DISCHARGE CLAIM
The district court terminated the appellant's wrongful discharge claim in advance of trial under the aegis of Fed.R.Civ.P. 56. The appellant presses her objection.
A. The Summary Judgment Standard.
The Civil Rules empower a court to grant summary judgment “if the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to a judgment as a matter of law.” Fed.R.Civ.P. 56(c). We have explored the nooks and crannies of this rule in a compendium of cases, see, e.g. McCarthy v. Northwest Airlines, Inc., 56 F.3d 313, 314-15 (1st Cir.1995); National Amusements, Inc. v. Town of Dedham, 43 F.3d 731, 735 (1st Cir.), cert. denied, 515 U.S. 1103, 115 S.Ct. 2247, 132 L.Ed.2d 255 (1995); Pagano v. Frank, 983 F.2d 343, 347 (1st Cir.1993); Wynne v. Tufts Univ. Sch. of Med., 976 F.2d 791, 793-94 (1st Cir.1992), cert. denied, 507 U.S. 1030, 113 S.Ct. 1845, 123 L.Ed.2d 470 (1993); United States v. One Parcel of Real Property (Great Harbor Neck, New Shoreham, R.I.), 960 F.2d 200, 204 (1st Cir.1992); Griggs-Ryan v. Smith, 904 F.2d 112, 115-16 (1st Cir.1990); Medina-Munoz v. R.J. Reynolds Tobacco Co., 896 F.2d 5, 7-8 (1st Cir.1990); Garside v. Osco Drug, Inc., 895 F.2d 46, 48-49 (1st Cir.1990), and it would serve no useful purpose to rehearse that jurisprudence here.
For the nonce, we think it is sufficient to repeat that “summary judgment's role is to pierce the boilerplate of the pleadings and assay the parties' proof in order to determine whether trial is actually required.” Wynne, 976 F.2d at 794. Thus, a Rule 56 motion may end the case unless the party opposing it can identify a genuine issue as to a material fact. In this regard, “genuine” means that the evidence on the point is such that a reasonable jury, drawing favorable inferences, could resolve the fact in the manner urged by the nonmoving party. See One Parcel, 960 F.2d at 204. By like token, “material” means that a contested fact has the potential to alter the outcome of the suit under the governing law if the dispute over it is resolved favorably to the nonmovant. See id.
When the summary judgment record is compiled the trial court must scrutinize it “in the light most hospitable to the party opposing summary judgment, indulging all reasonable inferences in that party's favor,” Griggs-Ryan, 904 F.2d at 115, but disregarding “conclusory allegations, improbable inferences, and unsupported speculation,” Medina-Munoz, 896 F.2d at 8. If no genuine issue of material fact is discernible, then brevis disposition ordinarily follows.
Because the summary judgment standard requires legal reasoning as opposed to differential factfinding, appellate review of summary judgment orders is plenary. See Pagano, 983 F.2d at 347; Garside, 895 F.2d at 48.
B. The Merits.
New Hampshire law controls Smith's pendent wrongful discharge claim. Under that law, even an at-will employee cannot be cashiered for a reason that offends public policy because such an employment decision “is not in the best interest of the economic system or the public good and constitutes a breach of the employment contract,” Monge v. Beebe Rubber Co., 114 N.H. 130, 316 A.2d 549, 551 (1974). The appellant urges that her severance offended the state's policy against gender-based discrimination. In the court below, Judge Stahl ruled that when a statutory remedy is available, New Hampshire courts would not entertain a complaint that an at-will employee had been wrongfully discharged in violation of public policy. Therefore, the appellant's common law claim for wrongful discharge failed because pregnancy discrimination is redressable under Title VII. See Smith I, slip op. at 9-10.
In reaching this conclusion, the district court drew heavily upon the teachings of Howard v. Dorr Woolen Co., 120 N.H. 295, 414 A.2d 1273 (1980). The appellant strives to convince us that a later New Hampshire case, Cloutier v. Great Atlantic & Pacific Tea Co., 121 N.H. 915, 436 A.2d 1140 (1981), defenestrates the district court's reading of Howard. We are not persuaded.
In Howard, the plaintiff alleged that he had been discharged because of age. The New Hampshire Supreme Court construed its seminal decision in Monge, 316 A.2d 549, “to apply only to a situation where an employee is discharged because he performed an act that public policy would encourage, or refused to do that which public policy would condemn.” Howard, 414 A.2d at 1274. A discharge due to age fell outside this “narrow category” inasmuch as the “proper remedy for an action for unlawful age discrimination is provided for by statute.” Id. (listing state and federal statutory remedies). In Cloutier, the court synthesized these cases, holding that to come within the judicially created public policy exception a plaintiff “must show that the defendant was motivated by bad faith, malice, or retaliation in terminating [her] employment,” 436 A.2d at 1143, and must also “demonstrate that [s]he was discharged because [s]he performed an act that public policy would encourage, or refused to do something that public policy would condemn,” id. at 1144. Cloutier did not answer, however, the question of whether such a cause of action lies where, as here, the public policy at stake is codified in a statute that itself provides a private right of action to remedy transgressions.10
A recently decided case makes the import of the state supreme court's earlier decisions pellucid and speaks directly to the question that confronts us here. In Wenners v. Great State Beverages, Inc., 140 N.H. 100, 663 A.2d 623 (1995), the plaintiff relied on a section of the Bankruptcy Code to establish a public policy against the termination of his employment. See id. at 625. The court held that “[w]hile a plaintiff may not pursue a common law remedy where the legislature intended to replace it with a statutory cause of action,” a wrongful discharge action could proceed if the relevant statutory provision did not provide a private cause of action for its violation. Id. (internal citations omitted). We deem this holding to be dispositive of Smith's contention.11
Title VII not only codifies the public policy against gender-based discrimination (including, but not limited to, pregnancy discrimination) but also creates a private right of action to remedy violations of that policy and limns a mature procedure for pursuing such an action. Under Wenners, the existence of such a remedy precludes the appellant, in the circumstances of this case, from asserting a common law claim for wrongful discharge. It follows that the district court acted impeccably in granting summary judgment on this claim.12
We need go no further. On the factbound Title VII claim, this case presents a close question. In the end, however, we must uphold the district court's judgment because the standard of review is generous and there is enough evidence in the record to support the trier's findings. On the two common law claims, our task is appreciably easier; both claims raise questions of law, not of fact, and the district court-albeit in the person of two different district judges-correctly resolved them.
Although I am compelled by the deference due a district court's findings of fact to concur in the final result, I write separately because I am troubled by the analysis used in deciding the Title VII claim. The majority applauds the district court's failure to fully analyze Smith's claims as “prudential.” I, however, am convinced that Smith produced direct evidence of intentional discrimination and that the district court was obligated to fully analyze plaintiff's case under the framework of Price Waterhouse v. Hopkins, 490 U.S. 228, 109 S.Ct. 1775, 104 L.Ed.2d 268 (1989). Additionally, I think that the majority mischaracterizes the law relevant to the causation requirement under Title VII and Morse's position-elimination defense. Its opinion could erroneously be viewed as an invitation to use that defense as a cover for discrimination against women who take or plan to take maternity leave.
I. The District Court's Analytical Process
The basic facts are undisputed. My first concern arises from the district court's abbreviated analysis of plaintiff's claim. The Supreme Court has established two analytical frameworks that courts reviewing Title VII claims must follow. Where the evidence produced at trial is “direct,” the Price Waterhouse framework applies.13 See Fields v. Clark Univ., 966 F.2d 49, 51-52 (1st Cir.1992), cert. denied, 506 U.S. 1052, 113 S.Ct. 976, 122 L.Ed.2d 130 (1993); Cumpiano v. Banco Santander P.R., 902 F.2d 148, 152 (1st Cir.1990); Jackson v. Harvard Univ., 900 F.2d 464, 467 (1st Cir.), cert. denied, 498 U.S. 848, 111 S.Ct. 137, 112 L.Ed.2d 104 (1990).
If the evidence of discrimination is indirect or circumstantial, the burden-shifting framework of McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973), governs. See McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973); Texas Dep't of Community Affairs v. Burdine, 450 U.S. 248, 101 S.Ct. 1089, 67 L.Ed.2d 207 (1981); St. Mary's Honor Ctr. v. Hicks, 509 U.S. 502, 113 S.Ct. 2742, 125 L.Ed.2d 407 (1993). These basic rules have been followed, as they must, by this Circuit. See, e.g., Cumpiano, 902 F.2d at 152; Jackson, 900 F.2d at 467; Chamberlin v. 101 Realty, 915 F.2d 777, 782 n. 7. (1st Cir.1990).
Yet, the district court found that gender discrimination played no part in the decision to terminate the plaintiff's employment without determining whether there was direct evidence under Price Waterhouse or even mentioning McDonnell Douglas. See Jackson, 900 F.2d at 467 (holding that a finding of direct evidence renders the McDonnell Douglas framework inapplicable). The majority compounds this analytical omission by praising the district court for its “directness” and for having “largely bypassed any differential direct evidence/circumstantial evidence tamisage.” A district court's decision to circumvent the analytical processes Supreme Court and circuit precedent require should be criticized, not praised.
This is particularly true where Title VII cases are concerned. The discrimination that plaintiffs like Kathy Smith face in the workplace is frequently as subtle as it is invidious. It is in recognition of this hard truth that the Supreme Court established an analytical process which district courts, in my opinion, are required to follow. See, e.g., McDonnell Douglas, 411 U.S. at 801, 93 S.Ct. at 1823-24 (“[I]n the implementation of [employment] decisions, it is abundantly clear that Title VII tolerates no ․ discrimination, subtle or otherwise.”); see also Price Waterhouse, 490 U.S. at 271, 109 S.Ct. at 1802. The Court's jurisprudence stands for the principle that the unlawfulness of the employment actions typically challenged in Title VII cases is best exposed through a process of inquiry. See, e.g., Burdine, 450 U.S. at 255 n. 8, 101 S.Ct. at 1094 n. 8 (“In a Title VII case, the allocation of burdens and the ․ prima facie case [requirement] [are] intended progressively to sharpen the inquiry into the elusive factual question of intentional discrimination.”). Because I stand by that principle, I would ordinarily suggest a remand in a case such as this.
I have come to the conclusion, however, that remand would not be meaningful in this case. This does not mean that I agree with the district court's finding that the evidence produced by Smith was not compelling. I concur in the result because I am bound by Supreme Court and circuit precedent. And in this area, that precedent, unfairly in my opinion, imposes too heavy a burden on plaintiffs trying to prove the ultimate issue in discrimination cases: that the employer intentionally discriminated against her on the basis of a Title VII-protected trait. I believe that Smith has produced enough evidence to meet her initial burden under Price Waterhouse or McDonnell Douglas, but agree that it would have been plausible for a factfinder to conclude that Morse proved its position-elimination defense by a preponderance of the evidence or, alternatively, that the facts established were insufficient to show pretext. Although it did so without adhering to the process Title VII requires, the district court decided the ultimate issue in the case and, although I disagree with it, I cannot say that decision was clearly erroneous.
II. Direct Evidence Under Price Waterhouse
In light of my concurrence in the majority's ultimate holding on Smith's Title VII claim, issues pertaining to the nature of the evidence Smith produced at trial are, admittedly, moot. Nevertheless, I want to explain my belief that Smith produced direct evidence and that Price Waterhouse controls this case. This is important for two reasons. First, the availability of direct evidence determines whether a case should be analyzed under Price Waterhouse or McDonnell Douglas. Direct evidence renders the McDonnell Douglas framework inapposite and imposes a heavier burden of proof on the employer. Fuller v. Phipps, 67 F.3d 1137, 1141 (4th Cir.1995).
Second, the determination of whether the evidence produced at trial is direct, though cast in procedural terms, affects the substantive outcome in Title VII cases. See Deborah C. Malamud, The Last Minuet: Disparate Treatment After Hicks, 93 Mich.L.Rev. 2229, 2229 (1995) (“Title VII jurisprudence cloaks substance in the ‘curious garb’ of procedure.”). This observation is of less import in Smith's case because, at the time the events giving rise to Smith's suit occurred, the law provided that an employer shown to have unlawfully discriminated could avoid Title VII liability by demonstrating by a preponderance of evidence that the adverse employment decision would have been the same even if discrimination had played no role. Lam v. Univ. of Hawai'i, 40 F.3d 1551, 1564-65 (9th Cir.1994). In other words, direct evidence of discrimination, without more, was not enough to impose liability on Morse. Id.
Under today's applicable law, however, a plaintiff producing direct evidence of discrimination under Price Waterhouse may have a Title VII remedy. Id. at 1565 n. 24. The Civil Rights Act of 1991 “modified the Price Waterhouse scheme” and made “mixed-motives treatment more favorable to plaintiffs.” Fuller, 67 F.3d at 1142; see Civil Rights Act of 1991, Pub.L. 102-166, § 107, 105 Stat. 1071, 1073 (1991) (codified at 42 U.S.C. § 2000e-2). Section 107 of the Act provides that Title VII is violated whenever an employer takes sex or pregnancy into account, regardless of whether other considerations independently explain the adverse employment decision. Id.; see 42 U.S.C. § 2000e-2(m) (“[A]n unlawful employment practice is established when the complaining party demonstrates that race, color, religion, sex, or national origin was a motivating factor for any employment practice, even though other factors also motivated the practice.”). Prevailing mixed-motives plaintiffs, at the very least, are now entitled to declaratory and injunctive relief and attorney's fees. See Kerr-Selgas v. Am. Airlines, 69 F.3d 1205, 1210 (1st Cir.1995) (citing 42 U.S.C. § 2000e-5(g)(2)(B)) (where an employer in a mixed-motives case proves that it would have made the same decision, the prevailing plaintiff is entitled to attorney's fees, and declaratory and injunctive relief, but not damages or reinstatement). Thus, what constitutes direct evidence is a critical issue for Title VII plaintiffs.
The majority makes repeated references to “smoking gun” evidence. Using this term only obscures the fact that this Circuit has yet to clearly define what constitutes direct evidence of gender discrimination. On prior occasions we have held that “[d]irect evidence is evidence which, in and of itself, shows a discriminatory animus.” See, e.g., Jackson, 900 F.2d at 467. But, this reasoning is circular and does not further understanding of the term. Justice O'Connor, in her concurring opinion in Price Waterhouse, defined the term in the negative, explaining that direct evidence “exclude[s] ‘stray remarks in the workplace,’ ‘statements by nondecisionmakers', or ‘statements by decisionmakers unrelated to the decisional process itself.’ ” Price Waterhouse, 490 U.S. at 277, 109 S.Ct. at 1804-05 (O'Connor concurring).
I contend that the evidence Smith produced at trial was direct and, therefore, warranted full application of the Price Waterhouse framework. The evidence shows that Smith was pregnant, and requested and received unpaid maternity leave. After being on leave several weeks, Smith notified Morse's general manager, Guimond, that she wanted to return to work on May 15, 1989, a week earlier than planned. Guimond approved the earlier start time and assured Smith that her job was secure. She also asked Smith whether she intended to have additional children; Smith indicated that she did.
On May 2, 1989, the day after this conversation occurred, Guimond also questioned Vendasi, Smith's sister, about Smith's future childbearing plans. Smith confronted Guimond about this behavior and the rumor that she would not be returning to work because she had decided to stay home with her child. Guimond denied any knowledge about the rumor and reiterated that Smith's job was secure; she repeated this guarantee two days later. Despite these assurances, Guimond terminated Smith on May 11, 1989, one week after their last conversation and four days before Smith was slated to return to work. Guimond requested permission to tell people that Smith failed to return to work because she decided to stay home to care for her child, but Smith refused to give it.
There is precedent holding that statements like those Guimond made to Smith and Vendasi constitute direct evidence. For example, in the Eighth Circuit, statements made by an employer can be direct evidence of discrimination, if made during a key decisional process. In Beshears v. Asbill, 930 F.2d 1348, 1354 (8th Cir.1991), the court held that an employer's oral statement, “older employees have problems adapting to changed and new policies,” was direct evidence of age discrimination. 930 F.2d at 1354. Two years later, the court expanded its Beshears holding to include written statements. Radabaugh v. Zip Feed Mills, Inc., 997 F.2d 444, 449-50 (8th Cir.1993), held that written statements included in corporate planning documents were also direct evidence of discrimination.
Other circuits have included statements made outside of the decisional process in the definition of direct evidence. In 1994, the Seventh Circuit held that post-discharge statements made by a supervisor were direct evidence of age bias, even though they were not reflective of an express intent to discriminate. See Robinson v. PPG Indus., Inc., 23 F.3d 1159, 1165 (7th Cir.1994). Similarly, the Eleventh Circuit has held that statements made by an employer to third parties are direct evidence of discriminatory animus. In EEOC v. Beverage Canners, Inc., 897 F.2d 1067, 1070 (11th Cir.1990), the court found that racially biased statements made by a supervisor to workers in his plant were direct evidence of racial animus and a hostile environment under Title VII.
Guimond's statements to both Smith and Vendasi fall well within the definition of direct evidence established by cases such as Beshears and Beverage Canners. Guimond was solely responsible for Morse's personnel decisions. Her questions about Smith's childbearing plans were neither stray nor random and evinced a concern about future pregnancy. Additionally, Guimond began asking questions about Smith's childbearing plans during what she admits was a key decisional period. Finally, the facts show that the timing of the decision to terminate Smith was suspicious. Cf. Troupe v. May Dep't Stores, 20 F.3d 734, 736 (7th Cir.1994); Josey v. Hollingsworth Corp., 996 F.2d 632, 639 (3d Cir.1993). Within two weeks of learning about Smith's plans to have more children, Guimond decided to terminate Smith, even though she had repeatedly assured Smith that her job was secure.
This evidence of discrimination is direct and clear even if it does not reach the status of a smoking gun. That some inferences must be drawn from what was said and done to reach this conclusion does not make Smith's evidence indirect. As the Seventh Circuit recognized in its 1991 decision, Visser v. Packer Eng'g Assoc., Inc., 924 F.2d 655, 659 (7th Cir.1991), “all knowledge is inferential.” Because judges are not mind-readers and cannot reach into the mind of a Title VII defendant, a certain amount of inference-drawing is necessary in any case, whether the evidence is direct or indirect. The ultimate issue in disparate treatment cases-whether the employer intended to discriminate-cannot be established by purely direct evidence. See Charles A. Sullivan, Accounting For Price Waterhouse: Proving Disparate Treatment Under Title VII, 56 Brook.L.Rev. 1107, 1138 (1991) (“ ‘[D]irect evidence’ of intent cannot exist, at least in the sense of evidence which, if believed, would establish the ultimate issue of intent to discriminate.”); Tyler v. Bethlehem Steel Corp., 958 F.2d 1176, 1183-84 (2d Cir.), cert. denied, 506 U.S. 826, 113 S.Ct. 82, 121 L.Ed.2d 46 (1992).
Rather than adhering to the colorful but meaningless requirement of a smoking gun, I think we should adopt a definition of direct evidence in Title VII cases which satisfies the minimum negative requirements Justice O'Connor set out in Price Waterhouse: “exclude[s] ‘stray remarks in the workplace,’ ‘statements by nondecisionmakers', or ‘statements by decisionmakers unrelated to the decisional process itself.’ ” Price Waterhouse, 490 U.S. at 277, 109 S.Ct. at 1804-05 (O'Connor concurring). In accord with the Civil Rights Act of 1991, this definition preserves the mixed-motives case as a viable option in Title VII suits. Cf. Michael A. Zubrensky, Despite The Smoke, There Is No Gun: Direct Evidence Requirements In Mixed-Motives Employment Law After Price Waterhouse v. Hopkins, 46 Stan.L.Rev. 959, 969 (1994). It lowers the high hurdle of “smoking gun” evidence to reasonable limits so that plaintiffs in employment discrimination cases can receive all the protections Title VII was intended to give.
Even if my definition of Price Waterhouse direct evidence is rejected, however, it is irrefutable that Smith made out a prima facie case of discrimination under McDonnell Douglas: that after being directly so asked, she expressed an intention to become pregnant in the future; that her performance at work was more than satisfactory; that she was terminated after repeated assurances that her job was “secure;” and that her duties continued to be performed by comparably qualified individuals. See Cumpiano, 902 F.2d at 153; Lipsett v. Univ. of P.R., 864 F.2d 881, 899 (1st Cir.1988).
Smith's reiteration of these facts on appeal complied with Supreme Court and circuit precedent. Smith proved that she was fired even though she was an excellent manager and that her duties continued to be performed by other employees. In my view, this is all McDonnell Douglas' prima facie case burden requires. See, e.g., Byrd v. Ronayne, 61 F.3d 1026, 1031 (1st Cir.1995) (“[T]he required prima facie showing is not especially burdensome.”) (citing Woodman v. Haemonetics Corp., 51 F.3d 1087, 1091 (1st Cir.1995)). The district court should have shifted to the McDonnell Douglas framework before finding Smith's evidence deficient.
III. Causation Under Title VII
In addressing the question of causation in disparate treatment cases, the majority stresses that a “coincidence” between pregnancy leave and an employment decision does not prove intentional discrimination. It may not in all cases, but it arguably did in this case. The majority's discussion of causation completely disregards this possibility. Its blanket contention that pregnancy does not give plaintiffs “total immunity” from adverse employment actions ignores the extent to which maternity leave gives employers an opportunity to discharge women who take maternity leave or who express an intention to have one or more children.
The evidence arguably shows that the position Smith held would have been eliminated even if Morse had not considered her pregnancy or intention to become pregnant in the future. It does not necessarily follow from this, however, that Smith would have been fired had Morse not considered her maternity leave or desire to have more children. In their conversations before Smith took maternity leave, Bond, Morse's president, and Guimond discussed eliminating the materials manager position, but not Smith. The record shows both that Bond initially intended to retain Smith because of her excellent skills and that he admitted that Smith would still be employed at Morse had she not taken maternity leave.
Had Smith refused to disclose or even lied about her intention to have more children, she would probably still have a job at Morse. The facts show that Guimond was very concerned about the disruption Smith's absence would cause and suggest that she would have taken steps to avoid such disruption in the future. The majority completely ignores the probability that Smith's expressed desire to have more children was the motivating factor in her discharge and that her temporary absence on maternity leave gave her employer an opportunity to find a reason to discharge her. I contend that the evidence Smith produced was sufficient to establish intent and causation.
The two examples the majority gives to illustrate the need for a causal connection between pregnancy and the adverse employment action challenged in disparate treatment cases are both inapposite and unfair. Footnote 9 of the court's opinion analogizes Smith's dismissal during maternity leave to an employee who is discharged while on religious leave because heroin is discovered in her desk. It is true that in both cases the employee's absence enabled the employer to make the discovery resulting in discharge. But here the analogy breaks down.
The possession of heroin is illegal; its presence in the employee's desk was a fact that could not be refuted (although an explanation might be made). The employer did not have to make any determination as to the quality of the employee's work or her capabilities. She had to be fired. In the case of maternity leave, however, an employer would have to make a judgment as to whether eliminating the position made good business sense. Considerations such as the employee's prior performance and future childbearing plans would be part of the employer's position-elimination decision. At least in part, that decision would be “because of” pregnancy, present and future. It could not be made in the vacuum the majority's hypothetical presupposes.
Similarly, the cases the majority cites to support its view obscure the causation issue and unfairly compare Smith to employees who are placed on probation because of poor attitudes or who are discharged because of unexcused absences. Cases such as Troupe v. May Dep't Stores Co., 20 F.3d 734 (7th Cir.1994), Crnokrak v. Evangelical Health Systems Corp., 819 F.Supp. 737 (N.D.Ill.1993), and Johnson v. Allyn & Bacon, Inc., 731 F.2d 64 (1st Cir.), cert. denied, 469 U.S. 1018, 105 S.Ct. 433, 83 L.Ed.2d 359 (1984), involved discharge, not position elimination. In Troupe, the employee's pre-maternity leave dismissal was motivated by her tardiness and frequent absences. Crnokrak involved a plaintiff who was terminated after returning from maternity leave later than originally expected, whereas Johnson dealt with an employee who lacked supervisory skills and who was fired after being placed on probation because of a poor work attitude. The one position-elimination case the majority cites, Pearlstein v. Staten Island Univ. Hosp., 886 F.Supp. 260 (E.D.N.Y.1995), is similarly inapposite; it involved adoption, not pregnancy, and an employee who gave short notice of her need for maternity leave. And in that case, the evidence showed that the plaintiff was accidentally overpaid, that her employer was experiencing financial difficulties, and that she had received no assurances about the security of her job.
These cases do not directly address the causation issues presented here. In contrast to Pearlstein, the evidence in this case shows that Smith received repeated assurances about her job, that the raise she received before taking maternity leave was intentional, and that Smith's termination was not due to economic hardship. Additionally, the evidence does not show that Smith was fired for a poor attitude, that she had ever been on probation, or that she lacked supervisory skills. The fact that Smith received regular promotions and that few people at the Morse plant exceeded her level of education or experience belies any suggestion that Smith's performance and skills were below par.
Finally, Smith received permission for her maternity leave, shortened the duration of that leave, and was fired before she could return to work, not before she left. Smith's maternity leave, thus, did not pose a problem for Morse in the same way that the Troupe employee's unexpected illness or the Crnokrak plaintiff's extended leave did for their employers. The crux of Morse's defense, after all, is that Smith was fired because her absence had no effect whatsoever on Morse's operations.
My point is simple: just as pregnancy does not fully shield plaintiffs from adverse employment actions, business judgment or necessity does not totally immunize employers from Title VII's sanctions. The majority's discussion of causation understates this important point. I believe that, more often than not, a correlation between pregnancy and position elimination during maternity leave will exist. It is naive to think that an employer would not take an employee's pregnancy or intention to become pregnant in the future into consideration during the process of determining whether the employee's position should be eliminated.14
IV. The Position-Elimination Defense
The majority upholds the district court's finding that Morse made out a position-elimination defense on two grounds: that Morse reduced its management-level staff and that Smith's duties were shifted to employees who were already on the Morse payroll. Though I concur in the holding that Morse arguably proved the facts necessary to rebut Smith's gender discrimination claim, I think the scope of the position-elimination defense is considerably more narrow than the majority's interpretation of the facts suggests. That a company is able to manage in the absence of one of its key employees will not always be proof of a nondiscriminatory purpose, contrary to what the court's opinion implies. Were that so, every woman who took maternity leave would do so at risk of losing her job.
Moreover, the conclusion that Morse reduced its management staff is not supported by the evidence. Morse did not, as the court's exposition of the facts suggests, reduce its management team from seven to three. The majority reached this conclusion by eliminating Bond and Guimond from its final count, even though they each donned one of the two hats formerly worn by Darryl Robinson, Damar's founder and chief officer. It also erroneously included Smith in Damar's original management team, even though she did not have a management title at that time. And it failed to include the two assistant manager positions in its final count, even though the individuals holding those slots did have management titles. If the individuals excluded from the majority's calculations are added, the size of Morse's management team was the same at the end as it was in the beginning-seven.15
The facts demonstrate that Morse mainly reorganized its management team. It consolidated positions and eliminated titles, but did not decrease the size of its management. Because it would have been plausible for the district court to interpret this reorganization as position elimination, I concur in the court's holding. I do not agree, however, that reorganizations of the sort Morse carried out will be enough to rebut claims of intentional discrimination in every case. For me, whether the district court was clearly erroneous in its findings on this issue was a very close call.
The court's holding that Smith was not replaced, that her duties were merely transferred to other Morse employees, is based on our holding in LeBlanc v. Great Am. Ins. Co., 6 F.3d 836 (1st Cir.1993), cert. denied, 511 U.S. 1018, 114 S.Ct. 1398, 128 L.Ed.2d 72 (1994). LeBlanc holds that a position-elimination defense is not defeated by the claim that an employee was only “replaced” because “another employee [was] assigned to perform the plaintiff's duties in addition to other duties, or [because] the work [was] redistributed among other existing employees already performing related work.” 6 F.3d at 846; see also Barnes v. GenCorp., Inc., 896 F.2d 1457, 1465 (6th Cir.), cert. denied, 498 U.S. 878, 111 S.Ct. 211, 112 L.Ed.2d 171 (1990).
To the extent that Morse's defense comports with LeBlanc at all, it does so on the basis of the first prong, not the second. In analogizing Morse's first reorganization to the reorganization which occurred after Smith's firing, the majority opinion gives the impression that LeBlanc 's second prong, the “related work” requirement, can be satisfied by demonstrating that a plaintiff's duties were simply transferred to someone working in the same company. I disagree. I contend that LeBlanc 's related-work requirement cannot be met unless the employer proves that it shifted the plaintiff's duties to employees who were already performing some of the plaintiff's duties or, at least, duties that were very similar. This did not occur in this case.
In the first reorganization, Smith was promoted to materials manager and asked to officially assume some of the duties she had already been performing because of the inadequacies of other managers. Smith at that time assumed duties which, in my opinion, constituted related work under LeBlanc. In contrast, the second reorganization did not shift Smith's responsibilities to managers who had already been performing her job. After Smith was fired, those managers took on what were essentially new duties; the majority's own contention that Paradis and Shevenell were far more experienced than Smith and responsible for the technical aspects of Morse's business bears this out. That they performed those duties for some period before Smith was fired was only because Smith was on maternity leave. The nonpregnancy-based explanation for their additional responsibilities did not kick in until after Smith's firing.
If Title VII's protections against pregnancy-based discrimination are to have any force, the relevant period of inquiry for determining whether the duties formerly performed by a plaintiff were assumed by someone already performing related work under LeBlanc should not be during a maternity leave. The relevant period of inquiry must be before that leave began. Using the time period when the woman is on maternity leave creates a perverse incentive to discriminate against pregnant women by firing them when they are not at their jobs and when it will almost always be true that someone else is performing their duties. In this case, if Smith had not become pregnant and taken maternity leave, she would still be a valued Morse employee.
William James once said that an idea's “validity is the process of its valid-ation.” Accordingly, I concur in the outcome reached in this case, but not the process employed, because I disagree with the view of pregnancy discrimination cases taken by the majority. I think it only plausible that gender was not the motivation for the adverse employment action taken against Smith, not “true.” And I agree only that position elimination can be a defense in Title VII cases, not that it will be a defense in every case. For me, the process employed in reaching a result, which includes the hypotheticals drawn and examples given, matters.
1. During this same time frame, the company eliminated the engineering manager's position. However, Bickford remained with Morse in a lesser capacity.
2. The company reprimanded the employee and trial testimony tended to establish that Morse had not authorized the comments.
3. The Civil Rights Act of 1991, Pub.L. 102-166, § 102, 105 Stat. 1071, 1073 (1991) (codified at 42 U.S.C. § 1981a(c)(1)), authorizes trial by jury in Title VII cases. Since the events that form the basis of the appellant's claim occurred prior to the effective date of the 1991 Act, she had no right to a jury trial on her Title VII claim. See Landgraf v. USI Film Prods., Inc., 511 U.S. 244, ----, 114 S.Ct. 1483, 1487, 128 L.Ed.2d 229 (1994) (holding that the 1991 Act is not retroactive). By like token, the Price Waterhouse framework for proof of “mixed-motive” discrimination that we describe in Part II(B), infra, is somewhat changed under the 1991 Act. See Fuller v. Phipps, 67 F.3d 1137, 1142 (4th Cir.1995).
4. Mesnick is a case brought under the Age Discrimination in Employment Act (ADEA), 29 U.S.C. §§ 621-634, rather than under Title VII. The same burden-shifting framework applies in both instances; therefore, ADEA cases have solid precedential value in Title VII litigation. Hence, we cite herein interchangeably to Title VII and ADEA cases, often without distinguishing between them.
5. To cite an example, Damar split the responsibility for manufacturing between two managers (Shevenell and Paradis), a situation that, in appellant's own phrase, caused daily “chaos.”
6. Bond testified that he purchased Damar to acquire its engineering talents. Paradis and Shevenell were highly trained and experienced engineers, while Smith had no such credentials. When Morse discovered that it could function with one less manager, the decision to retain Paradis and Shevenell, and dismiss Smith, seems quite plausible.
7. We note in passing that the appellant's reasoning is hopelessly circular. Morse demonstrated a firm commitment to downsizing and actively sought ways to streamline its operations. Consequently, there is no basis for surmising that Morse would have failed to realize that the materials manager's position was superfluous whether or not Smith took a maternity leave.
8. We stress that this case is brought pursuant to, and is governed by, Title VII. If the recently enacted Family and Medical Leave Act of 1993, P.L. 103-3, 107 Stat. 6 (1993) (codified at 29 U.S.C. §§ 2601-2654) were applicable, a different set of rules would obtain.
9. Say, for example, a Jewish employee, in charge of maintaining corporate records, stays home for a week to observe Passover. In her absence, her employer rummages through the file drawers that she maintains in search of a particular memorandum. The employer finds a packet of heroin. The employer would not have had the occasion to look through the file drawers but for the fact that the employee was on religious leave; he would simply have asked the employee for the memo. In such circumstances, we think it is clear that the employer can fire the employee for introducing drugs into the workplace without violating Title VII's ban on religious discrimination.
10. In Cloutier, the defendant argued that there must be a statutory expression of a public policy, and that a generalized assertion of a public policy (loosely based on a federal statute) is insufficient as a matter of law to meet the public policy prong of a wrongful discharge claim. See Cloutier, 436 A.2d at 1144-45. The court disagreed, observing that it had “not restrict[ed the] holding in Howard to situations involving a public policy enunciated in a statute. Public policy exceptions giving rise to wrongful discharge actions may also be based on non-statutory policies.” Id. at 1144. This language means no more than that a plaintiff can utilize a statutory provision to prove the existence of a public policy; it does not address the more sophisticated issue of whether a plaintiff may rely on a statute that provides a remedy for its violation.
11. To the extent that either Kopf v. Chloride Power Electronics, Inc., 882 F.Supp. 1183, 1189-90 (D.N.H.1995), or Godfrey v. Perkin-Elmer Corp., 794 F.Supp. 1179, 1187 (D.N.H.1992), hold otherwise, Wenners consigns them to the scrap heap.
12. We acknowledge some apparent tension between this ruling and our earlier opinion in Chamberlin v. 101 Realty, Inc., 915 F.2d 777, 786-87 (1st Cir.1990). We set Chamberlin to one side for a pair of reasons. First, the parties there did not raise the issue of statutory preclusion, and the panel did not address that issue. Second, Wenners makes a dispositive difference. When the highest court of a state disposes of an issue of state law contrary to the resolution of the issue theretofore suggested by a federal court, the latter ruling must give way. See Williams v. Ashland Eng'g Co., 45 F.3d 588, 592 (1st Cir.) (permitting relaxation of stare decisis principles when “controlling authority, subsequently announced,” undermines an earlier decision), cert. denied, 516 U.S. 807, 116 S.Ct. 51, 133 L.Ed.2d 16 (1995).
13. The plurality opinion in Price Waterhouse does not itself require direct evidence of discrimination. The reference to direct evidence appears in Justice O'Connor's concurrence in that case. See, e.g., 490 U.S. at 270-74, 109 S.Ct. at 1801-03. This court first adopted Justice O'Connor's conclusion that direct evidence is required in mixed-motives cases in Jackson v. Harvard Univ., 900 F.2d 464 (1st Cir.1990), cert. denied, 498 U.S. 848, 111 S.Ct. 137, 112 L.Ed.2d 104 (1990).
14. I am, of course, aware that the Family and Medical Leave Act of 1993, P.L. 103-3, 107 Stat. 6 (1993) (codified at 29 U.S.C. §§ 2601-2654) addresses a number of the concerns I raise. That Act, however, does not apply in pre-1993 cases and does not, moreover, correct the problems I perceive in the majority opinion's analysis and posture towards Smith's discrimination claim.
15. Post-acquisition of Damar, Morse's upper-level management team included the following seven people: Bond (president); Guimond (general manager); Paradis (machining); Shevenell (sheet metal); Bickford (engineering); Seeger (sales); and Smith (materials). I do not include Lane and Hickman in this number because they were fired almost immediately after Damar's acquisition, partially due to their poor performance. After Smith was fired, Morse's upper-level management team still included seven individuals: Bond (president); Guimond (general manager); Paradis (operations); Shevenell (manufacturing); Seeger (sales); Lapanne (assistant manager); and Hoffman (assistant manager).
SELYA, Circuit Judge.