concurring in part and dissenting in part.
The majority establishes a public-policy exception to the at-will employment doctrine and grants retroactivity to the deci*117sion announced today. The public-policy exception should be approved, but I would grant prospective rather than retrospective application of the newly created rule by applying it only to cases based on facts occurring after the announcement of this opinion. Accordingly, I respectfully dissent to the retroactive application of the public-policy exception. I concur in all other parts of the opinion.
The principle that no employee should be required to violate the law or falsify facts in order to maintain his employment is sound. In my view, the acts alleged by Paul Lorenz constitute grounds for a civil action and damages. While I agree that the rule announced today should be applied to the parties before the court, I do not agree that we should formulate new law creating liability for conduct that was not actionable in the past, and permit recovery under the public-policy exception for conduct that occurred prior to our announcement of the new rule. I would limit the application of the public-policy exception for cases predicated on conduct occurring after the announcement of this opinion.
I agree with the majority that James B. Beam Distilling Co. v. Georgia, — U.S. —, 111 S.Ct. 2439, 115 L.Ed.2d 481 (1991) (plurality opinion), which rejected “selective” prospective application of a new rule in civil cases, id. Ill S.Ct. at 2447, does not compel this court to dispense with the analysis of the prospective or retroactive issue under the three-prong test set forth in Chevron Oil Co. v. Huson, 404 U.S. 97, 106-07, 92 S.Ct. 349, 355-56, 30 L.Ed.2d 296 (1971). It has long been held that
[a] state in defining the limits of adherence to precedent may make a choice for itself between the principle of forward operation and that of relation backward. It may say that decisions of its highest court, though later overruled, are law none the less for intermediate transactions .... The choice for any state may be determined by the juristic philosophy of the judges of her courts, their conceptions of law, its origin and nature.
Great N. Ry. Co. v. Sunburst Oil & Ref. Co., 287 U.S. 358, 364-65, 53 S.Ct. 145, 148-49, 77 L.Ed. 360 (1932). The principle of state court control of the retroactivity of its decisions has been recently affirmed. See Beam Distilling, 111 S.Ct. at 2443; American Trucking Ass'ns v. Smith, 496 U.S. 167, 110 S.Ct. 2323, 2330, 110 L.Ed.2d 148 (1990) (plurality opinion) (“When questions of state law are at issue, state courts generally have the authority to determine the retroactivity of their own decisions.”). Three other states have either declined to follow or have distinguished Beam Distilling in resolving the selective prospectivity issue. Harper v. Virginia Dept. of Taxation, 410 S.E.2d 629, 631 (Va.1991), petition for cert. filed, 60 U.S.L.W. 3406 (U.S. Nov. 15, 1991) (No. 91-794); Devrnja v. West Virginia Bd. of Medicine, 408 S.E.2d 346, 349 (W.Va.1991); First Interstate Bank v. Larson, 475 N.W.2d 538, 545 (N.D.1991).
The Chevron factors were first reviewed by this court to analyze the retroactive application of an evidence standard announced by the United States Supreme Court in a parental termination proceeding. People in the Interest C.A.K., 652 P.2d 603, 607-09 (Colo.1982). C.A.K. recognized two justifications for denying retroactive application to a judicial ruling: First, the protection of parties who relied on the earlier state of the law, and second, the protection of societally important interests in stability. Id. at 608 (citing Thomas S. Currier, Time and Change in Judge-made Law: Prospective Overruling, 51 Va.L.Rev. 201 (1965)). In relation to the first justification, we recognized:
The reliance factor is more persuasive when the change in the law at issue concerns pre-litigation conduct that becomes the subject of later litigation, because most acts, once done, cannot be undone. Consequently, it would be inequitable to subject conduct to a standard that did not obtain when the conduct was engaged in.
Id. The Chevron analysis has three parts: (1) whether the decision at issue establishes new law, (2) whether retrospective application of the new rule would further or retard its operation, and (3) whether retrospective application of the new rule could *118produce substantial inequitable results. Chevron, 404 U.S. at 106-07, 92 S.Ct. at 355-56. Rather than weighing the three parts together, we treat the first part— whether a new rule has been established— as a threshold question. After the threshold has been met, we then balance the remaining two factors. Marinez v. Industrial Comm’n, 746 P.2d 552, 557 n. 4 (Colo.1987).
Today’s decision establishes a six-factor test to determine whether an employee's discharge violates public policy and grants an exemption to the well recognized at-will employment rule.1 To succeed in establishing a prima facie case for the tort of wrongful discharge in violation of public policy “a plaintiff must prove (1) that he refused to perform an action (2) ordered by his employer (3) which would violate a specific statute (4) whose terms are more than a broad general statement of policy and (5) that his termination resulted from his refusal,” Cronk v. International Rural Elec. Ass’n, 765 P.2d 619, 622 (Colo.1988), as well as (6)
that the employer was aware, or reasonably should have been aware, that the employee’s refusal to comply with the employer’s order or directive was based on the employee’s reasonable belief that the action ordered by the employer was illegal, contrary to clearly expressed statutory policy relating to the employee’s duty as a citizen, or violative of the employee’s legal right or privilege as a worker.
Maj. op. at 109. The majority has engraft-ed the sixth step on the Cronk rule, creating a new principle of law that could not have been foreseen. Since the Chevron threshold has been met, we must now balance the remaining factors.
The second factor in Chevron, whether retrospective application of the new rule would further or retard its operation, was derived from an earlier criminal case, Linkletter v. Walker, 381 U.S. 618, 85 S.Ct. 1731, 14 L.Ed.2d 601 (1965), which held:
Once the premise is accepted that we are neither required to apply, nor prohibited from applying a decision retrospectively, we must then weigh the merits and demerits in each case by looking to the prior history of the rule in question, its purpose and effect, and whether retroactive operation will further or retard its operation. We believe that this approach is particularly correct with reference to the Fourth Amendment’s prohibitions as to unreasonable searches and seizures. Rather than “disparaging” the Amendment we but apply the wisdom of Justice Holmes that “[t]he life of the law has not been logic: it has been experience.”
Id. at 629, 85 S.Ct. at 1737 (quoting Oliver W. Holmes, The Common Law 5 (Howe ed. 1963)). At issue in Linkletter was whether the application of the exclusionary rule to the states in Mapp v. Ohio, 367 U.S. 643, 81 S.Ct. 1684, 6 L.Ed.2d 1081 (1961), was to be applied retroactively. In examining the purpose of Mapp, the Court stated,
Mapp had as its prime purpose the enforcement of the Fourth Amendment through the inclusion of the exclusionary rule within its rights. This, it was found, was the only effective deterrent to lawless police action. Indeed, all of the cases since Wolf [v. Colorado, 338 U.S. 25, 69 S.Ct. 1359, 93 L.Ed. 1782 (1949),] requiring the exclusion of illegal evidence have been based on the necessity for an effective deterrent to illegal police action. We cannot say that this purpose would be advanced by making the rule retrospective. The misconduct of the police prior to Mapp has already occurred and mil not be corrected by releasing the prisoners involved.
Linkletter, 381 U.S. at 636-37, 85 S.Ct. at 1741-42 (citation omitted, emphasis added). The Court concluded Mapp was only to be *119applied prospectively. Id. at 640, 85 S.Ct. at 1743.2
Here, the general purpose of the public-policy exception to at-will employment is to deter employers from using economic power to coerce employees to violate statutory law. The exception is primarily designed to benefit the public rather than employees. See Harrison, The Price of the Public Policy Modification of the Terminable-at-Will Rule, 34 Lab.L.J. 581, 581 (1983) (public-policy exception is designed to “counteract what can be termed the ‘externalities’ of the traditional terminable-at-will rule”). The purpose of the sixth prong is to limit frivolous litigation by discouraging an employee from bringing a public-policy wrongful discharge suit without proof that the defendant employer had actual or constructive knowledge that the employee refused to perform an order he reasonably believed to be illegal. Neither purpose would be furthered by applying today’s decision to litigants asserting claims based upon conduct occurring prior to the announcement of this opinion. We may properly address the issue in this case, but to grant retrospective application of the law announced today to cases based on prior conduct would amount to judicial legislation. The misconduct of other parties has already occurred and cannot be corrected by retroactive application of the public-policy exception. The majority states that “the retroactive application of the public-policy exception will have the salutary effect of prohibiting an employer from discharging an employee for refusing to engage in conduct clearly illegal or detrimental to the public good.” Maj. op. at 114. While this may be true, prospective application will have exactly the same effect without creating new claims where no cognizable claim existed before.
In evaluating the third factor in Chevron, whether retrospective application of the new rule could produce substantial inequitable results, I agree with the majority that it would not be inequitable to apply the public-policy exception to Lorenz’s claim for relief. The fact that we apply today’s rule to the case before us does not dispense with the necessity for the Chevron analysis of inequity. To be comprehensive, such a general inequity review must not be limited to a one-sided examination of the retroactive effect of a rule. We cannot be certain of the outcome of similar public-policy cases which are not before us. The majority predicts that application of the employer knowledge prong will “enure[ ] to the sole benefit of the employer.” Maj. op. at 114. I question whether this inevitably will be the case. Consider, for example, a hypothetical case in which an employee has written a memorandum to her employer explaining that she refuses to obey orders on the basis that obeyance would violate statutory law. With such a declaration, the employee’s case would be stronger rather than weaker under today’s rule. What is relevant to Chevron is not which type of similarly situated party may benefit from retroactive application of a new rule, but rather whether similarly situated parties on either side would suffer inequitable detriment.
While the majority admits that the additional evidentiary requirement may result in “injustice or hardship” to employees, it declines to ascribe any retroactive significance to this supposed fact. See maj. op. at 114. In essence, the majority holds that the probability of inequity to employees is irrelevant. Under such a rule, the flexibility of the Chevron retroactivity analysis would be greatly diminished. Analysis would be limited to a one-sided equity appraisal. I would not endorse such a result. Without speculating on the effect today’s rule may have on pending cases, it can surely be said that the public-policy exception enunciated today significantly changes *120the evidentiary requirements for both plaintiffs and defendants in public-policy wrongful discharge cases. Depending on the facts of each case, there is a likelihood of serious adverse consequences to parties who have relied on the prior state of the law, whether we are speaking of employers who have made termination decisions or employees who have decided to bring wrongful discharge suits. Furthermore, retroactive application of the public-policy exception, by substantially changing the rules of the game, may result in added litigation expense for bo.th parties. Prospective application would avoid such inequity. Application of the public-policy exception should be limited to cases based on conduct occurring after this opinion is announced. We have recognized the inequity of subjecting past conduct to a standard that did not exist at the time the conduct was engaged in. See C.A.K., 652 P.2d at 608. Retrospective application of a rule that changes settled employment law may interfere with stability in an area “where society attaches particular importance to stability.” Id.; Ground Water Comm’n v. Shanks, 658 P.2d 847, 849 (Colo.1983). I would conclude that the balance of the second and third Chevron factors tilts toward application of today’s formulation of the public-policy exception to the at-will employment doctrine only to those employees whose wrongful discharge claims are based on facts occurring after the announcement of this opinion. Therefore, I respectfully dissent to retroactive application of the public-policy exception to the at-will employment doctrine.
I am authorized to say that Chief Justice ROVIRA and Justice YOLLACK join in this dissent.
. The majority treats Cronk v. Intermountain Rural Electric Ass'n, 765 P.2d 619 (Colo.App.1988), as the relevant decision for retroactive application. Today’s test includes a sixth prong that substantially changes Cronk. This change in the law was not clearly foreshadowed. Indeed, the majority cites no authority for requiring employer knowledge of an employee's reasonable belief that a refused order was illegal. See maj. op. at 109.
. The Linkletter approach has been rejected in criminal law. New rules must be applied to all cases pending on direct review, Griffith v. Kentucky, 479 U.S. 314, 323, 328, 107 S.Ct. 708, 713, 716, 93 L.Ed.2d 649 (1987), but not necessarily to cases pending on collateral review, Teague v. Lane, 489 U.S. 288, 288, 310, 109 S.Ct. 1060, 1060, 1074, 103 L.Ed.2d 334 (1989) (plurality opinion) (“Unless they fall within an exception to the general rule, new constitutional rules of criminal procedure will not be applicable to those cases which have become final before the new rules are announced.”).