(dissenting) — Relying on a dubious formulation of public policy, the majority today invalidates a company work rule designed to protect the lives of men and women employed as drivers in the unique and highly dangerous occupation of operating armored cars. I dissent.
In Thompson v. St. Regis Paper Co., 102 Wn.2d 219, 232, 685 P.2d 1081 (1984), this court recognized a "cause of action in tort for wrongful discharge if the discharge of the employee contravenes a clear mandate of public policy.” The Thompson court also recognized that this public policy exception to the common law terminable-at-will doctrine is a narrow one. Id. at 232; see also Wilmot v. Kaiser Aluminum & Chem. Corp., 118 Wn.2d 46, 53, 821 P.2d 18 (1991). Unfortunately, the majority ignores both the requirement that a clear mandate of public policy be contravened, and the requirement that the public policy exception be a narrow exception to the terminable-at-will doctrine. Then, the majority compounds this error by an expansive search for public policy to justify invalidating an employer’s definition of just cause for termination. Thompson has never been used by this court to pass judgment on a company’s definition of just cause and we should not do so now.
As the majority notes, the narrow public policy exception to the terminable-at-will doctrine has been applied when an employee is discharged for: (1) refusing to commit an illegal act, e.g., price fixing; (2) performing a public duty, e.g., jury duty; (3) exercising a legal right, e.g., filing *952a workers’ compensation claim; or (4) whistleblowing. Roe v. Quality Transp. Servs., 67 Wn. App. 604, 607, 838 P.2d 128 (1992) (citing Dicomes v. State, 113 Wn.2d 612, 618, 782 P.2d 1002 (1989)). The majority also observes that most of the Washington cases demonstrating a wrongful discharge in violation of public policy have fallen under the third and fourth categories listed—that is, instances where employees were fired for exercising a legal right or for reporting an employer’s misconduct.
After acknowledging this precedent, the majority ignores it and instead applies a "more refined analysis” to the present case because the majority implicitly recognizes that this court has never invoked public policy to find an employee’s discharge for violating a work rule is itself a violation of public policy. Majority at 940. As the majority observes, Loomis’ employee handbook states that violation of the rule forbidding a driver from leaving an armored vehicle will be grounds for termination. Employee Handbook at 10 (cited in majority at 934). Thus, the majority has applied a formerly narrow exception to the terminable-at-will doctrine to a situation where an employer provided just cause for termination and where the employment-at-will rule is inapplicable. The result of the majority’s analysis is that the public policy exception to employment-at-will now applies to a fifth, completely incompatible category; that is, where this court disagrees with an employer’s definition of just cause for termination, as set forth in the workplace rules.
This exercise in micromanagement by the majority is accomplished by embracing a four-part test promulgated by Henry Perritt. While I do not have serious objections to this test, I find it cumbersome and unnecessary to a determination of whether a discharge in violation of public policy has occurred. Moreover, I do not see how the Perritt test supports the majority’s conclusion that Gardner’s termination for leaving the Loomis truck constitutes a violation of public policy.
Perritt proposes that four elements must be satisfied *953before a plaintiff may prove a public policy wrongful discharge tort: (1) the existence of a clear public policy, (2) that discouraging the conduct in which he or she engaged would jeopardize the clear public policy, (3) that the public-policy-linked conduct caused the dismissal, and (4) the absence of an overriding employer justification for the dismissal. See Henry H. Perritt, Jr., Workplace Torts: Rights and Liabilities §§ 3.7-3.21 (1991), cited in majority at 940-41.
According to Perritt, whether clear public policy is implicated can be resolved in four basic ways, with diminishing effectiveness:
1. By identifying a specific provision of a statute, constitution, or administrative regulation
2. By synthesizing a policy from several different statutes or constitutional provisions
3. By identifying a right or mode of conduct covered by a traditional common law cause of action
4. By identifying a trade practice.
Perritt § 3.7, at 66-67.
Initially, it is significant that most Washington cases finding a public policy violation have identified a single statute that clearly sets forth the relevant policy. See Bravo v. Dolsen Cos., 125 Wn.2d 745, 888 P.2d 147 (1995) (firing of nonunion employees for striking violated policy of RCW 49.32.020, which prohibits employers from interfering with the collective action of employees); Wilmot, 118 Wn.2d 46 (firing of employees for seeking workers’ compensation benefits violated policy set forth in RCW 51.48.025, which prohibits employers from discharging employees who file worker compensation claims); Bennett v. Hardy, 113 Wn.2d 912, 784 P.2d 1258 (1990) (firing of 60-year-old employee for hiring an attorney to protect her from age discriminatory practices violated policy of RCW 49.44.090, which makes age discrimination an unfair labor practice); and Hayes v. Trulock, 51 Wn. App. 795, 755 P.2d *954830 (firing of employee for complaining to officials about employer’s refusal to pay overtime violated RCW 49.46.100), review denied, 111 Wn.2d 1015 (1988).
By contrast, none of the three policies discussed in this case—helping law enforcement, encouraging good Samaritans, and protecting human life—is embodied in any single, specific legal provision. Instead, the plaintiffs and the majority attempt to satisfy the clear public policy element of the Perritt test by synthesizing policies from a number of laws which are only tangentially related to the public policy which the majority wants to find.
The majority itself rejects the plaintiffs’ efforts to find a public policy of encouraging citizens to help law enforcement. While the statutes plaintiffs cite encourage citizens to come to the aid of law enforcement, they do so "only under very limited circumstances.” Majority at 942. The majority finds it more accurate to say that the statutes support a public policy encouraging citizens to cooperate with law enforcement when requested or required, and then states as follows:
Public policy is not furthered by encouraging citizens to jump into the midst of every criminal situation. Citizens have not had law enforcement training, and their involvement in many situations can create additional risks of harm to those involved. A limited, albeit clear, public policy can be found in the cited statutes, but Plaintiffs give an overexpansive reading of those statutes in their attempt to present a general policy encouraging citizens to help in law enforcement.
Majority at 942. In other words, the majority concludes that the limited statutory policy of cooperating with law enforcement does not justify this court’s invalidation of the Loomis work rule. I agree.
My disagreement with the majority comes when it synthesizes two other public policies— encouraging good Samaritans and protecting human life—in exactly the same manner as the plaintiff found his public policy of helping law enforcement—by an overexpansive reading of case law and statutes. *955The majority first explains that the good Samaritan policy cannot be found in the rescue doctrine, which allows a rescuer to recover damages from the rescued person when the rescuer was injured in the course of the rescue. Maltman v. Sauer, 84 Wn.2d 975, 976, 530 P.2d 254 (1975). This doctrine applies only if the rescued person negligently caused the dangerous situation that invited the rescue. As the majority observes, this limited doctrine "is not squarely based on a public policy of encouraging citizens to help those in need . . . Majority at 943.
The majority finds, however, that RCW 4.24.300 and State v. Hillman, 66 Wn. App. 770, 832 P.2d 1369, review denied, 120 Wn.2d 1011 (1992), "weakly demonstrate that society places some value on citizens who help others in need.” Majority at 943. RCW 4.24.300 releases emergency caregivers from liability when the victim is negligently injured by the caregiver in the course of rendering emergency transportation or care. Hillman upholds the exceptional sentence of a defendant who killed a good Samaritan. Hillman, 66 Wn. App. at 777. From this support, the majority concludes that plaintiffs have proven the existence of a broad good Samaritan policy.
I find it difficult to see how affirming the exceptional sentence in Hillman is "squarely based on a public policy of encouraging citizens to help those in need . . . .” See Majority at 943. Will citizens be encouraged to act as good Samaritans if they know that those who kill them will be punished severely? More plausibly, Hillman is aimed at punishing the wrongdoer, not at encouraging the actions of a good Samaritan.
I also fail to see how a statute dealing with the release of emergency caregivers from liability establishes a clear public policy of encouraging citizens to go to the rescue of others or to act as good Samaritans. Here again, "[a] limited, albeit clear, public policy can be found” in the cited statute, but plaintiffs "give an overexpansive reading of the statutes in their attempt to present a general policy” encouraging citizens to act as good Samaritans. See Majority at 942.
*956Assuming that the cited statutes and case law establish a broad public policy of encouraging rescues by good Samaritans, I agree with the majority that this public policy is too weak to justify interfering with the Loomis work rule. However, I do not see how Hillman and RCW 4.24.300 clearly prove the existence of such a good Samaritan policy, and I would conclude that this policy fails to satisfy even the first requirement of Perritt’s four-part test.
Nor do I see any stronger support for the majority’s second public policy—protecting human life. As evidence thereof, the majority first cites federal case law holding that the Fourth Amendment’s protection against warrant-less searches is waived where the search is necessary to prevent physical harm to the officers or others. The majority then turns to Washington statutes and case law, observing (1) that the illegal use of force becomes lawful when done to protect persons from injury; (2) that homicide is justifiable if committed in defense of oneself or others; and (3) that in a prosecution for any crime other than homicide, it is a defense that the actor was threatened with immediate death or serious injury. Majority at 944. From these legal precepts, the majority discovers a public policy of encouraging persons to save others from life-threatening situations that it says satisfies Perritt’s clear public policy element.
Again, I disagree that such a clear public policy exists. To borrow from the majority’s earlier criticism of the plaintiffs’ proffered public policy of aiding law enforcement: the statutes and cases relied upon are "not squarely based on a public policy of encouraging citizens to help those in need.” Majority at 943. Rather, a waiver of Fourth Amendment protections involves a weighing of a criminal defendant’s right to privacy against the right of third parties to safety from physical harm. Likewise, statutes preserving the right of a person to act in self-defense are aimed at protecting such persons from criminal prosecution—not at "encouraging citizens to help those in need.” *957Id. at 15. In finding a clear public policy of saving others from harm, the majority is guilty of the same over expansive synthesis of law for which it earlier criticized the plaintiffs.
Instead of the questionable analysis employed by the majority here, I would follow the reasoning of the Court of Appeals in another public policy termination case, Roe v. Quality Transp. Servs., 67 Wn. App. 604, 838 P.2d 128 (1992). In Roe, the plaintiff argued that she was wrongfully discharged from her job for refusing to submit to a drug test. She contended that a clear mandate of public policy favoring employee privacy could be found in the state constitution, statutes, and the common law tort of invasion of privacy. Roe, 67 Wn. App. at 607. Before analyzing these legal provisions, the Court of Appeals cited this court’s earlier directive that it look to " 'the letter or purpose of a constitutional, statutory, or regulatory provision or scheme’ or in '[p]rior judicial decisions . . .’ when attempting to find and articulate a clear mandate of public policy.” Id. at 607 (citing Thompson, 102 Wn.2d at 232). The Court of Appeals then rejected the plaintiffs’ contention that art. I, § 7 of the Washington Constitution provided the needed clear mandate of public policy. While the court acknowledged that the provision creates a right of privacy, it also observed that, in Washington, this constitutional provision has been construed as a restraint on government and not as a restraint on private individuals. Roe, 67 Wn. App. at 608.
The court next considered whether RCW 49.44.120, which prohibits private or public employers from requiring employees to take lie detector tests, was a valid source of public policy. The Court of Appeals noted that a West Virginia court had found by analogy that pubic policy would prohibit drug testing by private employers, but added that West Virginia’s standard for recognizing a public policy exception to the terminable-at-will doctrine is "less rigorous than the clear mandate of public policy [that Washington] courts require.” Roe, 67 Wn. App. at *958609. The Court of Appeals did not find Washington’s lie detector statute a proper source of public policy. The court further found that none of the other privacy statutes plaintiff cited suggested a legislative intent to announce public policy in the area of drug testing. Plaintiff then contended that the common law tort of invasion of privacy provided the needed mandate, but the Court of Appeals concluded that the cases cited raised primarily personal interests rather than the competing interests of the private employer and the employee. Id. at 610.
The Court of Appeals then made a statement of particular relevance to the present case: "Thompson requires that we 'find’ not 'create’ public policy and further requires that the existence of such public policy be 'clear.’ ” Id. at 610. The court held that the trial court had properly "exercis[ed] the caution that Thompson requires” and affirmed the dismissal of the case. Id. at 610.
The majority’s approach here is closer to a creation of policy than a discovery thereof, and what it creates is scarcely clear. None of the common law or statutory provisions it cites applies squarely to the facts in this case. Loomis cites Roe in its brief and offers a criticism of plaintiffs’ argument that now applies to the majority opinion:
[I]t is nonsensical to argue . . . that a party may choose a statute which is not affected by the facts of the case . . . strip the statute of its limiting terms and then analyze the remains to determine its broadest conceptual purpose . . . and then extrapolate back to a conclusion that since the general conceptual purpose is impacted by said facts, the statute likewise is necessarily impacted.
Def.’s Opp’n Br. at 15.
The next step in the Perritt test is for the plaintiff to show public policy is jeopardized by the employer’s rule. To satisfy this factor a plaintiff must show that other means for promoting the public policy are inadequate. Majority at 945 (citing Perritt § 3.14, at 77). While the ma*959jority recognizes this second element, apparently the Perritt analysis fails to support the majority’s analysis, for it declines to address this required showing and instead cites a case adopting a Model Penal Code provision that allows the use of force when the actor believes such force is necessary for another’s protection. State v. Penn, 89 Wn.2d 63, 66, 568 P.2d 797 (1977). The majority then analyzes Gardner’s belief that another person was in danger and finds that belief reasonable.
The cite to Penn and the Model Penal Code suggests how far afield the majority is traveling to reach its result. More importantly, the majority’s discussion of Gardner’s subjective belief does not satisfy the showing required by the Perritt test that other means for promoting the policy would be inadequate.
Those other means are set forth in the majority opinion’s statement of facts. Instead of leaving the armored car when emergencies arise, the driver can summon help by using the truck’s two-way radio, public address system, and sirens. Indeed, it is likely that in most instances these measures would be more effective in saving lives than would be the driver’s actions in leaving an armored car. The majority’s earlier criticism of people engaging in law enforcement is equally pertinent to a situation where a driver leaves the safety of an armored vehicle to confront an uncertain and dangerous situation: "Public policy is not furthered by encouraging citizens to jump into the midst of every criminal situation. Citizens have not had law enforcement training, and their involvement in many situations can create additional risks of harm to those involved.” Majority at 942.
Loomis argues persuasively that its work rule promotes rather than conflicts with a policy of saving lives:
The armored car business operates in a very dangerous world, where the goal of those few citizens who necessitate the very use of armored cars is to get the people inside the car to open it up to the people outside the car. It is not hyperbole to state that, in such situations, the armored car, with its bulletproof *960glass, sirens and two-way radios, is as much a lifeline for the driver/guards and their custodian/partners who are out of the armored truck as it is a secure depository for valuable commodities.
Def.’s Opp’n Br. at 7.
The majority concludes summarily and rather emotionally that if employers are allowed to terminate their employees for saving persons from life-threatening situations when the employee "appears to be the only hope of rescue,” then the policy encouraging "all citizens” to engage in such conduct shall be jeopardized. Majority at 946. This conclusion takes no notice of the fact that armored car companies and their drivers are relatively few in number, or that their rule forbidding a driver from leaving a truck is actually intended to save lives. I agree with Loomis that "[wjhen one looks past the limited and emotionally charged facts of the instant case, it becomes clear that Loomis’ maligned work rule actually serves the interests of society, is consistent with public policy and therefore cannot be the basis of a claim for wrongful termination in violation of public policy.” Def.’s Opp’n Br. at 9.
Next, the majority finds that the causation element of Perritt’s test is met, rejecting Loomis’ argument that this element is not satisfied because Gardner was fired for leaving the truck and not for his subsequent conduct in the hostage situation. The majority states that this argument lacks merit despite its apparent acceptance of the distinction earlier in its opinion. See Majority at 940 ("Loomis argues it did not fire Gardner in retaliation for his saving a hostage—it fired him solely because he left the truck in violation of the work rule. This claim is supported by the fact that Loomis did not discipline Gardner’s partner, who was just as involved with the situation as Gardner.”)
This inconsistency aside, the majority proceeds to demonstrate the flaw in Loomis’ causation argument with a questionable example of its own. The majority reasons that if the truck were on fire, Gardner would have to leave *961it to save his life. If Loomis then fired Gardner for leaving the burning truck, public policy would clearly be violated. It thus follows, according to the majority, that "Gardner’s reasons for exiting the truck must be taken into account-when determining whether his discharge was because of the public-policy-linked conduct.” Majority at 947.
Initially, the majority’s example implicates entirely different policy concerns—those ^relating to self-preservation—and is therefore of little use in resolving this case. Furthermore, an employee’s reasons for violating a work rule do little to alter the thread of causation between the violation and a subsequent termination therefor. The question which must be answered under this part of the Perritt test is whether Gardner was fired for helping another in need or because he exited the vehicle in contravention of a work rule designed for employee safety. Since Loomis did not discipline Gardner’s partner it is clear that Gardner was not terminated for coming to the aid of bank personnel.
The final element of the Perritt test addresses the employer’s reasons for terminating the employee to determine whether those reasons justify a violation of public policy. The majority recognizes that a driver who leaves an armored car severs the partner’s lifeline to safety and renders both employees more vulnerable to harm. The majority also acknowledges that the work rule is strictly enforced to lessen the risk of robbers using a ploy to get the driver out of the truck. Loomis addressed these risks in criticizing Gardner’s self-preoccupation:
Plaintiffs’ analysis completely fails to consider the interests and safety of Gardner’s former co-employees. Is it fair to compromise their safety and security because one individual seeks to violate a work rule whenever he deems it necessary to do a good thing? In the armored transport industry a multitude of ruses are used by criminals to lure driver/guards out of their trucks thereby cutting the custodian/partner’s sole lifeline.
Def.’s Opp’n Br. at 8.
Another justification of the Loomis rule mentioned by *962the majority concerns the difficulty of insuring armored vehicles. What insurer will cover a loss if the truck is robbed while left unattended?
After listing these justifications, the majority concludes that the good Samaritan policy is not of sufficient importance to "warrant interfering with an employer’s workplace and personnel management.” Majority at 948. The majority concludes, however, that the "narrow public policy” encouraging citizens to rescue persons from life-threatening situations does warrant such interference. Majority at 949. The majority adds that its decision does nothing to invalidate Loomis’ work rule forbidding drivers from leaving their trucks on pain of termination. "Our holding merely forbids Loomis from firing Gardner when he broke the rule . . . .” Majority at 949-50.
Does the majority really believe that its decision will have no effect on Loomis’ work rule? It seems obvious to me that following the majority’s decision, the Loomis rule will be called into question whenever a driver leaves a truck. If a driver leaves thinking that a passerby is in mortal danger, but no such danger is present, will his actions become those of a mere good Samaritan, thereby enabling Loomis to fire him? Or, will his belief that there was life-threatening danger be sufficient to waive the termination rule? Will that belief require substantiation of the supposed danger by witnesses, or will the driver’s word alone be sufficient? The majority’s statement that the Loomis rule will be unaffected by its decision does not withstand scrutiny.
Moreover, the majority takes no notice of the new role that its opinion will thrust upon the courts of this state. Under the guise of a claim that the public policy exception to the terminable-at-will doctrine should apply, courts will now be forced to analyze an employer’s work rules to determine whether they provide proper cause for termination. This type of micromanagement of business is a complete misapplication of the public policy exception, and can hardly be the result intended by the Thompson *963court’s acceptance of a narrow exception to the at-will rule.
There is no question that the Loomis rule forbidding Gardner from leaving the truck is based on the substantiated conclusion that a driver’s departure endangers his own and/or another’s life. I therefore cannot conclude that the Loomis rule violates a public policy in favor of saving lives and would uphold the enforcement of that rule in this case. Mr. Gardner’s termination does not fit within the narrow parameters of a wrongful discharge in violation of public policy.