dissenting:
I disagree with the majority’s conclusion that the rule of law established in Padilla v. Industrial Commission, 696 P.2d 273 (Colo.1985), should be given retroactive effect. Padilla established a new rule of law because it decided an issue of first impression whose resolution was not clearly foreshadowed. I also conclude that the purpose and effect of retroactively applying that holding may retard its operation and the inequity imposed by retroactive application may impose injustice or hardship. I would reverse the judgment of the court of appeals and remand the case to the Director of the Division of Labor to hear evidence on the second and third factors for retroactivity set forth in People in Interest of C.A.K., 652 P.2d 603 (Colo.1982).
In C.A.K., we adopted the standard for retroactive application of decisions in civil cases set forth by the United States Supreme Court in Chevron Oil Co. v. Huson, 404 U.S. 97, 92 S.Ct. 349, 30 L.Ed.2d 296 (1971). In C.A.K., we described the factors involved as follows:
*1229First, the decision, if it is not to be applied retroactively, must establish a new rule of law. Second, the merits of each case must be weighed by looking to the purpose and effect of the rule in question and whether retrospective operation will further or retard its operation. Third, the inequity imposed by retroactive application must be weighed to avoid injustice or hardship.
C.A.K., 652 P.2d at 607. As the majority correctly points out, the question of re-troactivity arises only when a decision establishes a new rule of law. At 1224. For a decision to establish a new rule of law, it must either overrule clear past precedent on which the litigants may have relied or decide an issue of first impression whose resolution was not clearly foreshadowed. Chevron, 404 U.S. at 106, 92 S.Ct. at 355.
In Padilla, this court held that the statutory authority of the Director of the Division of Labor to reopen claims applied in cases resolved by settlement agreements as well as cases resolved by adjudication, and a settlement agreement between the parties could not abrogate that authority. The majority in Padilla relied on the “beneficient purpose” of the Workmen’s Compensation Act, the language of the reopening provisions, and the prior cases of this court.
The majority concludes that the resolution of the issue presented in Padilla was clearly foreshadowed. For the reasons hereinafter set forth, I disagree.
First, the fact that Padilla was not clearly foreshadowed is shown by the language of the opinion itself. The majority quotes a portion of the Padilla opinion where the court stated that “several prior decisions of this court suggest that such conclusion must be implied from the policy and structure of the Act.” Padilla, 696 P.2d at 279 (emphasis added). I do not believe that the holding in Padilla was clearly foreshadowed where prior case law merely suggested or implied the outcome. Although the Padilla court characterized its conclusion as “inescapable,” I did not then and do not now believe that conclusion was warranted.
Next, the conclusion that the holding in Padilla was not clearly foreshadowed is evident from the fact that courts in other jurisdictions had conflicting views on the issue at the time Padilla was decided. As set forth in Padilla, some courts agree with the majority’s conclusion that “settled workers’ compensation claims may be reopened on the ground of changed circumstances.” Padilla, 696 P.2d at 280. Courts in other jurisdictions, however, “have reached contrary conclusions.” Padilla, 696 P.2d at 280.
I also believe that Marinez v. Industrial Commission, 746 P.2d 552 (Colo.1987), supports my conclusion that Padilla established a new rule of law. In Marinez, this court held that Engelbrecht v. Hartford Accident & Indemnity Co., 680 P.2d 231 (Colo.1984), applied retroactively. However, in reaching that holding, the court concluded that Engelbrecht established a new rule of law because it decided an issue of first impression whose resolution was not clearly foreshadowed.
In Engelbrecht, we held that cost-of-living increases to federal social security disability benefits were not “periodic disability benefits” and therefore did not require corresponding reductions in state workers’ compensation benefits under section 8-51-101(1)(c), 3B C.R.S. (1986). To reach this conclusion, we examined the language of the statute and determined that it was subject to two interpretations. Therefore, we looked to the intent of the General Assembly. To discern that intent, we looked at the purposes of the Workmen’s Compensation Act: namely, to protect employees who suffer injuries, to give injured workers a reliable source of compensation, and to prevent double awards. Based on the language of the statute and previous decisions discussing the purposes of the act, we held that reductions in workers’ compensation benefits were not required under section 8-51-101(1)(c), 3B C.R.S. (1986), when cost of-living increases were made in federal social security disability benefits. This court in Padilla based its decision on the *1230same grounds: namely, the language of the reopening provision in the statute, prior cases of this court, and the “beneficent purpose” of the Workmen’s Compensation Act.
Other similarities between Padilla and Engelbrecht are apparent from a reading of our discussion of Engelbrecht in Mari-nez. In both cases, the issue before the court was one of first impression. Marinez, 746 P.2d at 557; at 1226. In both cases, prior to our decision, insurers operated according to their understanding of the statute’s requirements. In both cases the prevalent, if not uniform, practice of insurers before the decision was contrary to the interpretation given to the statute by this court. I believe that in this case, as in Marinez, we should find that a new rule of law was established.
Finally, I believe the reaction of the General Assembly in amending the Workmen’s Compensation Act within two months of the Padilla decision to abrogate that decision supports my conclusion that it was not clearly foreshadowed. The majority dismisses this point by stating that
“[t]he responsibility for determining whether a rule of law will be granted retrospective or prospective application is an issue for the courts to resolve.” Thus, even if we accepted the ... position that the 1985 amendments expressed a legislative policy as to the retroactive effect of Padilla, any such legislative determination would not be binding upon us.
At 1228 n. 11 (quoting Ground Water Comm’n v. Shanks, 658 P.2d 847 (Colo.1983)).
This rejection of the argument misconstrues the effect of the 1985 amendments. These amendments should not be construed as a pronouncement by the General Assembly that Padilla is not to be given retroactive effect; that question is one for the courts to resolve. These amendments are, however, a strong indication that the General Assembly did not think the conclusion reached by the majority in Padilla correctly reflected legislative intent and was clearly foreshadowed. On the contrary, these amendments demonstrate that the General Assembly assumed that final settlements in workers’ compensation cases would be binding, a conclusion opposite the holding in Padilla.
Not only did the General Assembly assume that final settlements were binding, but also insurance companies, self-insured businesses, and employees were of the same opinion. The workers’ compensation settlement provisions had existed since the enactment of the Workmen’s Compensation Act in 1919, without a decision that such agreements were not final and binding until Padilla was decided in 1985.
Because I conclude that Padilla was not clearly foreshadowed and was therefore a new rule of law, I must next examine the second and third factors described in C.A. K. The second factor in the analysis requires a weighing of the merits of Padilla by looking to the purpose and effect of the rule to decide whether retrospective operation will further or retard its operation.
The effect of the Padilla opinion, as I stated in my dissent there, ignores the fact that settlement of disputed issues of liability and the extent of injury not only serves a useful purpose for the claimant, employer, and insurance carrier, but also allows efficient use of public resources. Padilla, 696 P.2d at 282 (Rovira, J., dissenting).
Furthermore, in Padilla, the majority opinion relied in part on the “beneficent purposes” of the Workmen’s Compensation Act in reaching its conclusion. Padilla, 696 P.2d at 279. The “beneficent purposes” of the act include providing a reliable source of compensation to injured workers, and the financial stability of the fund from which injured employees receive benefits is an “important” concern. Marinez, 746 P.2d at 558-59.
As pointed out in the majority opinion, the Director of the Division of Labor and the Industrial Claim Appeals Panel assumed that Padilla applied retroactively in this case. The issue of retroactivity was not analyzed until the case reached the court of appeals. Therefore, no evidence was heard regarding the financial impact of retroactive application of Padilla on the *1231“important” concern of providing a reliable source of compensation to injured workers. Accordingly, I would remand the case to the Director of the Division of Labor for evidence on this question and a decision on the second factor in the retroactivity analysis.
The final consideration is the inequity resulting from retroactive application. The majority does not reach this question, and the court of appeals dismisses this factor, stating:
[TJhere is no evidence concerning the extent to which agreements of the kind at issue here were being employed by parties to workmen’s compensation proceedings prior to Padilla, nor is there any estimate as to the number of claims that a retrospective application of Padilla would make eligible for reopening. Indeed, the employer here has failed even to provide an estimate of the additional cost to it that the retroactive application of Padilla would require in this case. Under this state of the evidence, there is simply no factual basis for concluding that substantial inequity would result from a retroactive application of Padilla. This record, therefore, provides no reasonable ground for limiting that decision’s applicability.
Loffland Bros., 754 P.2d 768, 770 (Colo.App.1988). This conclusion is unfounded because, as pointed out above, the issue of the retroactivity of Padilla was not considered by the Director of the Division of Labor or the Industrial Claim Appeals Panel. The first time the retroactivity of Padilla was considered was by the court of appeals. Because the retroactivity issue was not considered at any level with a fact-finding function, there was no opportunity to present evidence regarding any inequity that would be caused by the retroactive application of Padilla. Therefore, I would remand the case to the Director of the Division of Labor for evidence on the question of inequity caused by the retroactive application of Padilla.
Accordingly, I respectfully dissent.
I am authorized to say that ERICKSON, J., joins me in this dissent.