Exxon Corporation v. Esso Workers' Union, Inc.

SELYA, Circuit Judge. This appeal tests the margins of an arbitrator’s ability to order the reinstatement, into a safety-sensitive job, of an employee who has failed a rehable drug test. After painstaking reflection, we conclude that a well defined and dominant public policy encourages employers to develop, establish, and enforce programs to prevent their employees from attempting to perform safety-sensitive work while under the influence of narcotics or other intoxicants. Moreover, once an employer has set such a program in place, it countermands public policy if courts too readily rescue employees who fail to satisfy programmatic standards from the predictable consequences of such violations. Hewing to this line, we refuse to enforce the arbitral award of which plaintiff-appellant Exxon Corporation (Exxon) complains. I. BACKGROUND The facts are essentially undisputed. Exxon operates a fuel terminal in Everett, Massachusetts and employs several truck drivers to supply petroleum to service stations and airports throughout New England. Exxon’s nemesis, the Esso Workers’ Union (the Union), appellee here, represents most of these drivers. Exxon and the Union entered into a collective bargaining agreement (the CBA) in *843 February 1990. The CBA establishes inter alia a five-step employee grievance procedure culminating in final and binding arbitration. Part 11 of the CBA covers employee discipline. Its first section provides that Exxon “shall post a list of offenses which it deems serious,” and its second section provides that Exxon “may discharge or otherwise discipline” any employee who commits a posted offense. The second section also stipulates that any employee who believes his suspension or discharge is without “just cause” may pursue a grievance. An appendix to the CBA catalogs the posted offenses. The list includes the following: 6. Alcohol Beverage/Habit-Forming or Illegal Drug or Any Dangerous Substance a. Being under the influence of an alcoholic beverage or drug on Company time or property. Testing positive on a drug test or refusal to submit to a drug test. b. Bringing onto Company property, or possessing, or using on Company time or Company property, an alcoholic beverage, illicit or unprescribed controlled substance, or any dangerous substance which the Company believes may impair the employee’s ability to properly perform duties in a safe and responsible manner. Exxon has implemented a comprehensive drug-free workplace program (the DFW program), embodied in a formal policy statement and the aforementioned list of posted offenses. The policy statement declares in part: Exxon Corporation is committed to a safe, healthy, and productive workplace for all employees. The Corporation recognizes that alcohol, drug, or other substance abuse by employees will impair their ability to perform properly and will have serious adverse effects on the safety, efficiency, and productivity of other employees and the Corporation as a whole.... Being unfit for work because of use of drugs or alcohol is strictly prohibited and is grounds for termination of employment. Exxon’s program is carefully tailored to meet the goals of the Drug-Free Workplace Act of 1988 (the DFW Act), 41 U.S.C. §§ 701-707 (1994). Exxon has made the program’s terms available to all employees; the program encourages employees voluntarily to report drug and alcohol problems; and the company not only provides rehabilitative services to employees who come forward, but also promises that “[n]o employee ... will be terminated due to the request for help in overcoming that dependency or because of involvement in a rehabilitation effort.” Exxon’s program reflects the company’s recognition that drug use during the performance of safety-sensitive tasks poses a significant threat to co-workers and to the public. Therefore, it subjects employees in these positions to random drug testing. In that regard, the program puts Exxon’s work force on notice of the company’s intention to conduct “[u]nannounced periodic or random [drug] testing” of employees who are working in certain designated safety-sensitive jobs. Albert A. Smith, a veteran Exxon employee, works in such a designated position. He is responsible for loading, driving, and unloading a five-axle tractor-trailer combination which, when fully loaded, carries 12,000 gallons of highly flammable motor fuel. He typically drives this rig through many of New England’s more densely populated areas. Exxon requires employees who occupy designated safety-sensitive positions—and Smith’s is plainly such a position 1 —to sign so-called compliance statements. Smith signed such a statement in 1989, thereby attesting that he had read and understood the parameters of Exxon’s DFW program, that he was not abusing alcohol or drugs, and that he was amenable to random drug testing. On August 21, 1990, Smith reported for duty. Without any forewarning, Exxon di *844 rected him to take a drug test. Smith submitted to the test and apparently drove his regular route that day. The test results were obtained the following week; they revealed that Smith had cocaine in his bloodstream when tested. Although the test results could not indicate when Smith had used the cocaine or whether he had performed his job while still under its pernicious influence, Exxon decided that Smith posed a threat to public safety and fired him. The Union grieved Smith’s ouster. The grievance culminated in arbitration. The parties put two questions to the arbitrator: (1) Did Exxon have just cause to discharge Smith? (2) If not, what is the appropriate remedy? In September of 1992, the arbitrator found the results of the drug test to be reliable but nonetheless decided that Exxon wrongfully terminated Smith’s employment. The arbitrator acknowledged that Part 11 of the CBA gave Exxon the right to discharge Smith for committing a posted offense, but he reasoned that this right was subject to Part ll’s “just cause” provision. Concluding that dismissal was too extreme a punishment, the arbitrator settled upon a two-month suspension as an appropriate disciplinary measure, to be followed by Smith’s reinstatement if he passed a contemporaneous drug test. Exxon balked at the arbitrator’s award and sued in federal district court to set it aside. The parties cross-moved for summary judgment. The lower court granted the Union’s motion and affirmed the arbitral award. Unyielding in its commitment to prevent Smith from getting behind the wheel of a petroleum truck, Exxon appeals. Our review of the district court’s legal conclusions is plenary. See Prudential-Bache Securities, Inc. v. Tanner, 72 F.3d 234, 237 (1st Cir.1995). II. PRINCIPLES AFFECTING JUDICIAL REVIEW Collective bargaining agreements are designed to memorialize the terms and conditions of employers’ relationships with their unionized employees. These agreements typically contain grievance procedures that designate arbitration as the final dispute-resolution mechanism. “In such cases ... courts play only a limited role when asked to review the decision of an arbitrator.” United Paperworkers Int’l Union v. Misco, Inc., 484 U.S. 29, 36, 108 S.Ct. 364, 370, 98 L.Ed.2d 286 (1987). In large part, that role is ordained by the fact that “[i]n labor arbitration, matters of contract interpretation are typically for the arbitrator, not for a reviewing court.” El Dorado Technical Servs., Inc. v. Union General De Trabajadores, 961 F.2d 317, 319 (1st Cir.1992). As long as the arbitrator is arguably interpreting the CBA, a court cannot second-guess his decision. See id. (citing Misco, 484 U.S. at 38, 108 S.Ct. at 370-71); Dorado Beach Hotel Corp. v. Union De Trabajadores De La Industria Gastronomica, Local 610, 959 F.2d 2, 3-4 (1st Cir.1992). In such purlieus, a court’s task ordinarily is limited to determining whether the arbitrator’s construction of the collective bargaining agreement is to any extent plausible. See Misco, 484 U.S. at 36-38, 108 S.Ct. at 369-71. Policy spins this web of rules. Judicial deference to an arbitrator’s contract interpretation furthers “[t]he federal policy of settling labor disputes by arbitration [wdiich] would be undermined if courts had the final say on the merits of [arbitral] awards.” United Steelworkers v. Enterprise Wheel & Car Corp., 363 U.S. 593, 596, 80 S.Ct. 1358, 1360, 4 L.Ed.2d 1424 (1960). Through the medium of the CBA, the employer and the union bargain for the arbitrator’s interpretation, and a federal court must respect that bargain. See W.R. Grace & Co. v. Local Union 759, Int’l Union of United Rubber Workers, 461 U.S. 757, 765, 103 S.Ct. 2177, 2182-83, 76 L.Ed.2d 298 (1983). It follows, therefore, that a court should not tamper with an arbitral award “unless it can be shown that the arbitrator acted in a way for which neither party could have bargained.” Local 1445, United Food & Commercial Workers Int’l Union v. Stop & Shop Cos., 776 F.2d 19, 21 (1st Cir.1985). Public policy, however, has its own imperatives—and they occasionally conflict with the imperatives of contract interpretation. It is a fundamental rule that courts must refrain from enforcing contracts that *845 violate public policy. Collective bargaining agreements are simply a species of contracts and, as such, are not immune from the operation of this rule. “As with any contract ..., a court may not enforce a collective-bargaining agreement that is contrary to public policy.” W.R. Grace, 461 U.S. at 766, 103 S.Ct. at 2183; accord Misco, 484 U.S. at 42-43, 108 S.Ct. at 373-74. Because this refusal to enforce contracts which offend public policy is inured in judicial tradition, the question of what public policy demands is within the judicial, not the arbitral, domain. See Misco, 484 U.S. at 43, 108 S.Ct. at 373-74; W.R. Grace, 461 U.S. at 766, 103 S.Ct. at 2183-84. III. ANALYSIS In the district court, Exxon argued for reversal of the arbitral award on two grounds: first, that the arbitrator exceeded his authority; and second, that the award violates public policy. The district court rejected both arguments. See Exxon Corp. v. Esso Worker’s Union, Inc., 942 F.Supp. 703 (D.Mass.1996). Because courts ought not trespass unnecessarily into the uncertainties of the public policy terrain, we begin by discussing Exxon’s more case-specific argument. A. The Arbitrator’s Authority. The key to this issue lies in Part 11 of the CBA. One section of Part 11 provides that Exxon “may discharge or otherwise discipline” employees who commit posted offenses—“may,” in this context, “means has a right to,” according to the definition contained in the CBA—and another section provides that employees may challenge discharges which Exxon has imposed without “just cause.” Exxon asseverates that the arbitrator should have equated the “right to discharge” language with the “just cause” language; because Exxon reserves the right to discharge employees who commit posted offenses, this thesis runs, it perforce has just cause to discharge such employees. But the arbitrator teased another meaning out of Part 11. He concluded that the language which permits Exxon “to discharge or otherwise discipline” an employee who commits a posted offense furnishes Exxon with a range of disciplinary options, and that this range is in turn subject to an independent application of the just cause barometer. On this reading of Part 11, the arbitrator ruled that Exxon did not have just cause to cashier Smith merely because he tested positive for drugs. 2 Although Exxon’s interpretation of the CBA may be somewhat less strained, judges have no roving writ to construe the contract language in the way that they think best. Rather, a court’s proper province is to determine whether the arbitrator’s reading is plausible, albeit not the reading the court might choose. See El Dorado, 961 F.2d at 320 (“When the language of the underlying contract, taken in context and with due regard for the surrounding circumstances, is fairly susceptible to differing meanings, a reviewing court must not meddle with the arbitrator’s rendition.”). In this instance, the arbitrator’s interpretation survives that indulgent scrutiny. The proof of the pudding is found in Crafts Precision Indus., Inc. v. Lodge No. 1836, Etc., 889 F.2d 1184 (1st Cir.1989). There, the employer had dismissed an employee for insubordination. The CBA listed insubordination as “one ‘example[]’ of conduct [that] may result in suspension, or immediate discharge,” and also included a clause reserving for the employer the exclusive right to discipline employees. Id. at 1184-85. In a refrain that echoes the argument which Exxon *846 makes here, the employer argued that these two clauses, in conjunction, gave it an absolute right to discharge an employee for insubordination and urged the arbitrator to equate this right to discharge with the CBA’s “just cause” provision. The arbitrator interpreted the right to discharge as distinct from just cause to discharge and instead reinstated the employee. On appeal, we upheld the award because the challenged language was open to several interpretations, and the arbitrator’s position reflected one such (plausible) iteration. See id. at 1185. Because Crafts is a fair congener, precedent compels us to conclude that the arbitrator’s interpretation of the disputed language here is within the pale and that the arbitrator did not exceed his authority in this respect. B. Public Policy. Exxon’s second claim of error can most usefully be discussed in three segments. 1. Framing the Inquiry. Misco is the watershed case in respect to judicial review of an arbitration award which is challenged on public policy grounds. There, the company employed Cooper as a night-shift machinist whose duties involved the operation of a dangerous piece of equipment. One night, police arrested him in the company parking lot, having discovered him “in the backseat of [a] car with marijuana smoke in the air and a lighted marijuana cigarette in the front seat ashtray.” 484 U.S. at 33, 108 S.Ct. at 368. The company then fired him for breaking its rule against possession of illicit drugs on business premises. The union grieved Cooper’s discharge, and an arbitrator ordered his reinstatement. The company sued and the federal district court annulled the award based on public policy. The Fifth Circuit affirmed, holding that Cooper’s reinstatement “would violate the public policy ‘against the operation of dangerous machinery by persons under the influence of drugs or alcohol.’” Id. at 35, 108 S.Ct. at 369 (quoting Misco Inc. v. United Paperworkers Int’l Union, AFL-CIO, Quachita Local 651, 768 F.2d 739, 743 (5th Cir.1985)). The Supreme Court reversed, ruling that a court may set aside an arbitrator’s award on public policy grounds only when “the contract as interpreted would violate ‘some explicit public policy’ that is “well defined and dominant.’ ” Id. at 43, 108 S.Ct. at 373 (quoting W.R. Grace, 461 U.S. at 766,103 S.Ct. at 2183-84). Neither common sense nor “general considerations of supposed public interests” are suitable vehicles for identifying public policy; rather, courts must glean public policy from laws and legal precedents. Id. (quoting W.R. Grace, 461 U.S. at 766, 103 S.Ct. at 2183-84). Because the lower courts had predicated their perceptions of public policy on intuition rather than positive law, the judgment could not stand. Misco teaches that, though courts may set aside arbitral awards which contravene public policy, they may do so only in a narrow class of cases, marked by a special set of circumstances. See id. at 43, 108 S.Ct. at 373-74. To determine whether a particular case fits within the confines of this class, courts must employ a two-tiered analytic approach. First, since a generalized sense of public policy provides an insufficient basis upon which to annul an arbitral award, an inquiring court must review existing statutes, regulations, and judicial decisions to ascertain whether they establish a well defined and dominant public policy. If positive law does not give rise to such a policy, the inquiry is at an end. See id. at 43-44, 108 S.Ct. at 373-74. If, however, the court finds that such a policy exists, it must then proceed to the second step of the pavane and determine whether the arbitral award clearly violates the discerned public policy. 3 See id. at 44, 108 S.Ct. at 374. 2. Identifying the Public Policy. There is a plenitude of positive law to support the *847 existence of a well defined and dominant public policy against the performance of safety-sensitive jobs while under the influence of drugs or other intoxicants. See Gulf Coast Indus. Workers Union v. Exxon Co., 991 F.2d 244, 252-53 (5th Cir.1993) (collecting cases). Gulf Coast itself is a representative ease. There, the court set aside an arbitral award which proposed to reinstate in a safety-sensitive position an employee who had tested positive for drug use after admitting to his employer that he had a drug problem but representing (falsely, as matters turned out) that he was obtaining treatment and abstaining from substance abuse. The court amply illustrated the proposition that numerous statutes, regulations, and judicial opinions “pronounce the emphatic national desire to eradicate illicit drugs from the workplace,” particularly in safety-sensitive occupations. Id. at 250; see also Exxon Corp. v. Baton Rouge Oil, 77 F.3d 850, 855-56 (5th Cir.1996) (again finding a well defined and dominant public policy against the performance of safety-sensitive jobs while under the influence of drugs). The Third Circuit has addressed the same issue in a trilogy of cases (all featuring an employer related to the appellant here). In Exxon Shipping Co. v. Exxon Seamen’s Union, 993 F.2d 357 (3d Cir.1993) (Exxon I), the court invoked public policy in refusing to enforce an arbitral award which directed the employer to reinstate a helmsman who had tested positive for drug use after his ship ran aground. Id. at 364. The court relied in part on a series of Coast Guard regulations, declaring them to be “part of a broader public policy against operation of common carriers under the influence of drugs,” and found that policy adequately evinced by an array of drug-testing regulations. Id. at 361-62 (citing 14 C.F.R. part 121, Appendix I (1992) (Federal Aviation Administration drug-testing program); 49 C.F.R. part 219 (1991) (Federal Railroad Administration drug-testing program); 49 C.F.R. part 391 subpart H (1991) (Federal Highway Administration drug-testing program)). In Exxon Shipping Co. v. Exxon Seamen’s Union, 11 F.3d 1189 (3d Cir.1993) (Exxon II), the court continued on the same course. It set aside as contrary to public policy an arbitral award reinstating an employee who reported to work inebriated. The court declared “that an owner or operator of an oil tanker should not be compelled to reinstate to a ‘safety-sensitive’ position an individual who has been found to be intoxicated while on duty on that vessel.” Id. at 1194. Finally, in Exxon Shipping Co. v. Exxon Seamen’s Union, 73 F.3d 1287 (3d Cir.), cert. denied, — U.S. -, 116 S.Ct. 2515, 135 L.Ed.2d 203 (1996) (Exxon III), the court reinstated an employee who had refused to submit to a drug test, finding that the CBA did not require the employee to take the test. Even then, the court reaffirmed its earlier finding that there exists a “broad public policy against permitting an individual to operate a vessel while under the influence of drugs or alcohol.” Id. at 1292. This chorus has many voices. Several other courts likewise have identified a well defined and dominant public policy against the performance of safety-sensitive jobs by persons under the influence of intoxicants. Thus, in Union Pacific R.R. Co. v. United Transp. Union, 3 F.3d 255, 262 (8th Cir. 1993), the court used public policy as a lever to set aside an arbitral award reinstating a railroad brakeman who had tested positive for drug use after a switching accident. The court had “no difficulty in concluding that there exists a well-defined and dominant public policy against a railroad’s employment of individuals whose impaired judgment due to the use of drugs or alcohol could seriously threaten public safety.” Id. at 261. Similarly, in Delta Air Lines, Inc. v. Air Line Pilots Ass’n Int'l, 861 F.2d 665, 674 (11th Cir.1988), the court defenestrated an arbitral award presuming to reinstate a pilot who had flown an aircraft while obviously drunk. The court described this as a “rare example of an award the enforcement of which would violate clearly established public policy which condemns the operation of passenger airliners by pilots who are under the influence of alcohol.” Id. at 671. By like token, the district court in Georgia Power Co. v. International Bhd. of Elec. Workers, Local 81, 707 F.Supp. 531, 538-39 (N.D.Ga.1989), aff'd, 896 F.2d 507 (11th Cir.1990), recognized the public policy against performance of safety-sensi *848 tive jobs by persons under the influence of drugs and set aside an arbitral award aimed at reinstating an employee who had tested positive for drug use. We agree with these courts. In our judgment, society has achieved a broad national consensus that persons should not be allowed to endanger others while laboring under the influence of drugs. This consensus is made manifest by positive law and translates into a well defined and dominant public policy— indeed, a national crusade—counselling against the performance of safety-sensitive tasks by individuals who are so impaired. One subset of this policy is that persons who are under the influence of narcotics or other intoxicants should not be permitted to operate commercial vehicles on public highways. This conclusion is fortified by our knowledge that the legislatures of those states through which Smith must drive a petroleum tanker-truck have uniformly criminalized the operation of motor vehicles by persons who are under the influence of alcohol or controlled substances. See Mass. Gen. Laws Ann. ch. 90 § 24(l)(a)(l) (West 1997) (criminalizing the operation of “a motor vehicle while under the influence of intoxicating liquor, or of marijuana, narcotic drugs, depressants or stimulant substances”); R.I. Gen. Laws § 31-10.3-31(a) (1996) (making it “illegal for any person driving any commercial motor vehicle ... to operate or control any such vehicle while under the influence of alcohol, drugs, toluene, or any other [controlled] substance”); id. § 31-27-2(a) (criminalizing the driving of “any vehicle ... while under the influence of any intoxicating liquor, drugs, toluene, or any controlled substance”); Conn. Gen.Stat. Ann. § 14-227a(a) (West 1997) (similar); N.H.Rev.Stat. Ann. § 265:82(I)(a) (1995) (similar); Vt. Stat. Ann. tit. 23, § 1201(a) (1995) (similar); Me.Rev. Stat. Ann. tit. 29-A, § 2411(1) (West 1996) (similar). We find further evidence of this policy in Congress’ enactment in 1991 of the Omnibus Transportation Employee Testing Act (the Testing Act), now codified in 49 U.S.C. § 31306 (1994). The Testing Act instructs the Secretary of Transportation to promulgate regulations “that establish a program requiring motor carriers to conduct preemployment, reasonable suspicion, random, and post-accident testing of operators of commercial motor vehicles for the use of alcohol or controlled substances.” Id. § 31306(b)(1)(A). In response, several Department of Transportation agencies have promulgated regulations designed to promote the public policy against performance of safety-sensitive tasks by persons who use drugs. For example, the Federal Aviation Administration has devised a program which requires preemployment drug testing as well as periodic drug testing of employees in safety-sensitive positions. See 14 C.F.R. Part 121, Appendix I (1996). The Coast Guard has promulgated regulations in order “to minimize the use of intoxicants by merchant marine personnel and to promote a drug free and safe work environment.” 46 C.F.R. § 16.101(a) (1996). The Federal Railroad Administration has adopted regulations crafted to “prevent accidents and casualties in railroad operations that result from impairment of employees by alcohol or drugs.” 49 C.F.R. § 219.1(a) (1996). The Federal Transit Administration’s regulations now require each recipient of a subsidy “to implement an anti-drug program to deter and detect the use of prohibited drugs by covered employees.” 49 C.F.R. § 653.3 (1996). Last, but surely not least, the Federal Highway Administration’s regulations have been tailored “to help prevent accidents and injuries resulting from the misuse of alcohol or use of controlled substances by drivers of commercial motor vehicles.” 49 C.F.R. § 382.101 (1996). Congress’ strongest statement against the performance of safety-sensitive tasks while under the influence of drugs is embodied in the DFW Act, which instructs federal agencies to award contracts or grants only to those employers who promise to provide a drug-free working environment by: (1) publishing a statement informing employees that use of drugs is prohibited in the workplace; (2) establishing a “drug-free awareness program;” (3) providing employees with drug counseling and rehabilitation services; (4) adopting and imposing penalties on employees who violate the terms of the “drug-free awareness program;” and (5) furnishing em *849 ployees with copies of the employer’s statement against on-the-job drug use. 41 U.S.C. §§ 701(a)(1), 702(a)(1). At this point in American history, few elements of public policy command the consensus that attaches to the policy against the use of controlled substances by those whose work potentially imperils others. Judicial decisions, agency regulations, and legislative enactments combine to form a solid phalanx of positive law evidencing a well defined and dominant public policy against the performance of safety-sensitive tasks while under the influence of drugs. Thus, Exxon has satisfactorily negotiated the first step of the public policy pavane. 3. The Interface. Confirming the existence of a well defined and dominant public policy is only half the battle. To abandon an arbitral award as contrary to public policy, a court must find that the award clearly violates the identified policy. See Misco, 484 U.S. at 43, 108 S.Ct. at 373-74; Prudential-Bache, 72 F.3d at 241. In this instance, the Union contends that, even if there is a cognizable public policy against the performance of safety-sensitive work by individuals who are under the influence of drugs, reinstating Smith would not insult such a policy because there is no evidence that Smith was in the grip of cocaine while driving his petroleum truck. According to the Union, the positive result of Smith’s random drug test “merely” indicates the presence of cocaine in his bloodstream; it does not necessarily signify that Smith was under the influence of the narcotic either at the time of the test or at the time he drove his rig. 4 The Union casts this argument so narrowly that it misses the mark. Relying upon job-relatedness as the sole determinative factor in permitting employers to discharge employees who test positive for drug use would force employers to wait for some other consequential indication that drugs are affecting work performance. Typically, this other indication will be an accident. See, e.g., Union Pacific, 3 F.3d at 256-57; Exxon I, 993 F.2d at 358-59; Amalgamated Meat Cutters, Local Union 540 v. Great W. Food Co., 712 F.2d 122, 123-24 (5th Cir.1983). The notorious mishap involving the Exxon Valdez, which produced vast environmental devastation, highlights the core problem associated with this “wait-and-see” approach. If we have learned anything from such catastrophes, it is that employers must act affirmatively to avoid drug-related accidents rather than wait passively for such accidents to happen. We conclude, therefore, that the well defined and dominant public policy which we have identified does not require an employer to await the occurrence of an accident before discharging an employee who tests positive for drug use. In this sense, the public policy is not as closely cabined as the Union implies. It is the Union’s failure to recognize this aspect—and, thus, to appreciate the full breadth of the discerned public policy—that is fatal to its argument and crucial to our decision. The pertinent public policy dictates not only that employees refrain from performing safety-sensitive jobs while under the influence of drugs, but also that employers develop (and enforce) programs designed to discourage such activity. This added dimension is most apparent in the DFW Act and in the Testing Act. The impact of the latter statute is made manifest by the proliferation of governmental regulations which mandate regular drug testing for employees in safety-sensitive positions. See, e.g., 14 C.F.R. Part 121, Appendix I (1996) (codifying Federal Aviation Administration’s drug-testing program); 46 C.F.R. §§ 16.101-16.500 (1996) (codifying Coast Guard’s chemical testing program); 49 C.F.R. §§ 219.1-219.715 (1996) (limning Federal Railroad Administration’s drug-testing procedures); 49 C.F.R. §§ 653.1-653.83 (1996) (delineating Federal Transit Administration’s drug-testing proce *850 dures); 49 C.F.R. §§ 382.101-382.605 (1996) (describing, inter alia, Federal Highway Administration’s drug-testing procedures). This statutory and regulatory mosaic bears witness that the same public policy which countervails the performance of safety-sensitive tasks while under the influence of drugs also encourages (and, in some cases, requires) employers to implement and enforce drug-free workplace programs which include mandatory drug testing of those in safety-sensitive posts. Consistent with this enhanced understanding of the discerned public policy, we hold that forcing an employer to reinstate an employee who tests positive for drug use pursuant to a test that the employer administers as part of a drug-free workplace program would undermine that policy. It makes no sense to construe public policy as encouraging—and in some cases mandating—employers to establish and enforce drug-testing programs, yet to preclude them from taking decisive action against those employees who test positive. The Union warns that this holding is wholly unprecedented. But the demands of public policy are dynamic rather than static. Modem society’s widespread recognition of, and increasingly aggressive response to, the growing drug problem is a harbinger that public policy may make progressively greater demands on industry. Moreover, the Union’s claim that we are blazing a new trail is not entirely accurate. At least two recent cases track the expanding public policy on which we rely. These cases note, albeit in dicta, that employers must not be compelled to reinstate personnel who violate the terms of a comprehensive drag-free workplace program. In Baton Rouge Oil, the Fifth Circuit reversed as contrary to public policy an arbitral decision awarding back pay to an employee in a safety-sensitive position who had tested positive for cocaine during a random drug test. The court held that allowing the employee to collect back pay would contravene public policy despite the absence of any evidence that he actually had performed his job while drug-impaired. See Baton Rouge Oil, 77 F.3d at 856. In so holding, the court noted the absurdity of reinstating such an employee: It is undisputed that Chube [the employee] occupied a safety-sensitive position. It is also undisputed that Chube tested positive for cocaine use while occupying that position, and thereby endangered the safety of other employees. We think that the public policy exception ... must be read not only to prohibit the prospective placement of an employee into a position where he is a danger to his company and to fellow employees (i.e., order of reinstatement into a safety-sensitive position), but also to prohibit a retrospective approval of the conduct. ... Id. The Third Circuit echoed these sentiments in Exxon III while upholding an arbitral award which reinstated an employee who refused to take a drug test. The court premised this ruling on the arbitrator’s conclusion that, under the terms of the collective bargaining agreement, the company lacked cause to insist upon a drug test. See Exxon III, 73 F.3d at 1295-96. En route to this determination, however, the court observed that “[a] clearly defined and cautiously administered program of drug testing ... is the natural corollary to ... a strong public policy that precludes allowing intoxicated or drug-impaired seamen to remain in safety-sensitive positions aboard oil tankers.” Id. at 1294. The court went on to proclaim that the “right to test employees for alcohol or drug use ... is critical to achieving the objective” of preventing drug-impaired individuals from performing safety-sensitive jobs. Id. The court’s ensuing discussion left no doubt that, if a drug test was validly requested, reinstating an employee who boycotted it would undermine public policy. See id. at 1294-95. Baton Rouge Oil and Exxon III reinforce the proposition that a comprehensive and finely-tuned DFW program which includes a drug-testing component is a natural corollary to the ringing public policy against performance of safety-sensitive jobs by individuals who are under the influence of narcotics or other intoxicants. It follows that, if an employer elects to establish such a program and *851 properly preserves its right of implementation in the collective bargaining agreement, thwarting the employer’s efforts to enforce the program’s standards would countervail the basic public policy. The Union intimates that the public policy we have identified, if it persists at all, can be vindicated by some other disciplinary measure, short of termination. This intimation misses the point. The arbitrator has said in effect that the CBA, properly construed, requires Exxon to reinstate Smith—and it would insult public policy for a court to enforce a contract that requires the ongoing employment in a safety-sensitive capacity of a worker who has scorned the employer’s drug-free workplace program. 5 This case is emblematic of the proposition. In terms of public policy, it would be grossly counterproductive to impede Exxon’s efforts at fully implementing its DFW program by forcing it to reinstate an employee who blatantly violated the program’s terms. Indeed, Smith’s utter disregard for Exxon’s DFW program is one feature which distinguishes this case from Misco. 6 Unlike the employer in Misco, Exxon maintains a comprehensive DFW program which is delicately calibrated to further the public policy against job performance while under the influence of drugs and other intoxicants. Smith transgressed the terms of this program three times over: failing to report his drug use to Exxon, falsely representing that he abjured illicit drugs, and testing positive for drug use. Given this threefold violation, Exxon acted reasonably in selecting discharge as the most appropriate means of eliminating the threat that Smith poses to the public. In the bargain, Exxon’s action was also a necessary means of ensuring the integrity of its DFW program. Forcing Exxon to reinstate, into a safety-sensitive position, an employee who lacks any meaningful commitment to its DFW program would hamstring its well-directed attempts to implement public policy. The Union tries to retrieve yet one more arrow from its quiver. Under the terms of its DFW program, Exxon treats employees who test positive for drug use more harshly than employees who voluntarily come forward and reveal that they are experiencing problems. During oral argument, the Union attempted to distort Exxon’s distinction between these two types of employees by suggesting that, since Exxon does not discharge the latter (i.e., employees who voluntarily report drug abuse), it lacks sufficient reason to discharge the former (i.e., employees who are “caught” by random drug testing). This argument is deeply flawed. Exxon encourages employees to report their drug use so that the company can transfer such workers to jobs that do not implicate public safety while they undergo rehabilitation. These employees do not pose a threat to the public because, by reporting their drug abuse, they provide Exxon with the opportunity to implement safety precautions. The actions of these employees are radically different from the actions of employees who, like Smith, attempt to conceal their drug use. These duplicitous employees pose a real and serious threat: by failing to report their problem, they deny Exxon the opportunity to take precautions to safeguard the public. On this basis, we believe it is reasonable—and fully consistent with the identified public policy—for Exxon to offer a measure of job security as an incentive for voluntary reporting, while cutting all ties with employees who do not accept the incentive and who subsequently are caught. *852 We need go no further. Because Smith thumbed his nose at Exxon’s DFW program, his reinstatement clearly would violate the well defined and dominant public policy against performance of safety-sensitive jobs while under the influence of drugs. Hence, the federal courts must refuse to enforce the arbitral award. Reversed. Footnotes [1] In an earlier, unrelated case which involved a hauler who, like Smith, failed a random drug test, we described a somewhat similar job as entailing "work of a kind where, one suspects, there might be old practitioners, and there might be bold practitioners—but there would likely be lew (if any) old, bold practitioners.” Jackson v. Liquid Carbonic Corp., 863 F.2d 111, 112 (1st Cir.1988). [2] According to the arbitrator: The just cause standard requires that the Company prove by the preponderance of the evidence that the employee committed the offense and that the level of discipline was warranted. In this case the Company's actions were automatic: if an employee in a designated position tests positive, s/he is terminated. The Company’s presumption is that the employee is a danger to public safety and the only remedy is to excise that danger. The Company’s self-imposed narrowness in its choice of remedy fails to meet the just cause standard. There was no evidence that Company drivers had any record of dangerous driving due to ingesting illicit drugs. In the case of Smith, there was no record of any discipline or any signs or indications of a drug-related problem during his nearly twenty years with the Company. [Emphasis supplied.] [3] The Misco Court provided an apt illustration of how the second-stage inquiry operates. It noted that, even assuming the existence of the public policy perceived by the court of appeals, reinstating Cooper did not necessarily frustrate that policy because there was no showing that Cooper had used marijuana while on the job. The Court thought that "the assumed connection between the marijuana gleanings found in Cooper’s car and Cooper’s actual use of drugs in the workplace is tenuous at best and provides an insufficient basis for holding that his reinstatement would actually violate the [perceived] public policy." 484 U.S. at 44, 108 S.Ct. at 374. [4] Although the arbitrator found that the drug test reliably indicated the presence of cocaine in Smith’s system (a finding that the Union does not contest on appeal), the test results could not pinpoint when Smith was under the drug’s influence. This uncertainty arises from the fact that the manner in which cocaine metabolizes within a person’s body depends upon a myriad of factors, many of which (e.g., the potency and purity of the drug ingested, the drug-user’s tolerance, food consumption, and psychological condition) were not known to Exxon. [5] Moreover, the alternative remedy selected by the arbitrator—a two-month suspension, followed by a one-time drug test—does not hold out much promise for the safety of either the public or Smith’s fellow employees. Smith's failed drug trust evinces his inability or unwillingness to conform to the strictures of the DFW program. If he were returned to a safety-sensitive position, as the arbitrator suggests, there would be no sound reason for believing that the leopard had changed his spots. [6] Another distinguishing feature is temporal in nature. Misco arose out of an incident that occurred in January 1983. Judicial review did not end until the Supreme Court spoke in 1987. Here, however, Smith failed the drug test in the summer of 1990, and judicial review is still ongoing. As our discussion of the emerging public policy reveals, see text supra, Misco predates both the Testing Act and the DFW Act. This chronological reality highlights the broader fact: public policy in respect to drugs in the workplace has matured greatly in the decade since Misco was decided.