concurring.
INTRODUCTION
I fully concur with and join in the scholarly and thorough opinion of the majority. That opinion was exhaustively reviewed and contributed to by each concurring justice. This Court long and carefully considered the applicable law and the arguments of counsel and of our dissenting brother Bakes before rendering its decision. Ultimately, the unambiguous message of the Idaho Tort Claims Act’s language and of the relevant case law compelled our decision. In that sense, this is a judicially conservative decision; it adheres closely to the law as the legislature wrote it and to the federal case law to which the legislature looked. This decision changes nothing in the Act; rather, it restores the Act to what the legislature intended in 1971. With this case and henceforth the Act will function as it should, imposing on the government the same baseline accountability for acts of wrongdoing as private individuals bear, while at the same time sheltering the government from liability for “discretionary” policy decisions as well as for conduct involving the other enumerated exceptions to liability.
I write separately only to answer the arguments advanced by the dissent. In contrast with the majority opinion, the dissent’s desired activist result would cast aside the legislature’s clear intent as expressed in the statutory language and in the pre-1971 federal case law in favor of a perceived policy need to limit the government’s potential liability more than the government itself considered necessary. It hardly need be said that this is not our proper role. In order to reach a result contravening the dictates of law, the dissent of a necessity fundamentally misconceives (1) the import of I.C. § 6-903(a) and related state and federal case law, (2) the import of the “discretionary function” exception found in I.C. § 6-904(1) and of related case law, (3) well-established tort law concerning (a) an alleged tort-feasor’s duty to supervise a person with which the alleged tort-feasor has a special relationship and (b) respondeat superior liability, and (4) the appropriate circumstances for making decisions prospective-only in effect. I will address these arguments in the order the dissent makes them.
I. GOVERNMENTAL IMMUNITY
A. The “Parallel Function ” Test.
As with Dunbar v. United Steelworkers of America, 100 Idaho 523, 602 P.2d 21 (1979), cert. denied 446 U.S. 983, 100 S.Ct. 2963, 64 L.Ed.2d 839, the dissent derives from I.C. § 6-903(a) an exception to liability where the underlying role or “function” of the government is one without parallel in the private sector. Under this test, the court disregards the actual conduct of the *234alleged tort-feasor, United States v. Feres, 340 U.S. 135, 142, 71 S.Ct. 153, 157, 95 L.Ed. 152 (1950) (quoted in dissent, infra) (court looks to “status of both the wronged and wrongdoer” rather than to the “circumstances”), and disregards the fact that a private individual or entity would be potentially liable for the same alleged misconduct. Id. (no liability for alleged misconduct though private individuals and entities “would undoubtedly be held liable” for the same misconduct). The dissent’s advocacy of this test not only ignores (1) the plain meaning of the pertinent language of I.C. § 6-903(a), see majority op., supra, at p. 216, 723 P.2d at 760, and (2) the United States Supreme Court’s construction of similar language in the Federal Tort Claims Act which has prevailed from 1955 to present, but also the express rejection of the purported exception in I.C. § 6-903(a).
I.C. § 6-903(a) makes the state liable for tortious conduct “whether arising out of governmental or proprietary function. . . .”1 As the majority opinion explains, the Idaho legislature followed the lead of the Supreme Court in Indian Towing Co. v. United States, 350 U.S. 61, 76 S.Ct. 122, 100 L.Ed. 48 (1955); Rayonier, Inc. v. United States, 352 U.S. 315, 77 S.Ct. 374, 1 L.Ed.2d 354 (1957), and subsequent relevant Supreme Court and lower federal court decisions. See cases cited in majority opinion, supra, at pp. 218-219, 723 P.2d at p. 762-763. As the majority explains, the exception to immunity for “governmental function[s] of governing” which Dunbar enunciated, Dunbar, supra, 100 Idaho at *235546, 602 P.2d at 44, and which the dissent now defends, is the same exception to immunity for which the federal government argued in Indian Towing and again in Rayonier. See majority op., supra at p. 217, 723 P.2d at p. 761. The Supreme Court saw this argument for what it was, an ill-founded attempt to “resort to an alleged distinction, imported from the law of municipal corporations, between the Government’s negligence when it acts in a ‘proprietary’ capacity and its negligence when it acts in a ‘uniquely governmental’ capacity.” Rayonier, supra, 352 U.S. at 318-19, 77 S.Ct. at 376. As did the Supreme Court, the Idaho legislature declined to return to this “ ‘non-governmental’ — ‘governmental’ quagmire.” Indian Towing, supra, 350 U.S. at 65, 76 S.Ct. at 124.
To leave no doubt, the legislature put it in writing in I.C. § 6-903(a). Significantly, neither Dunbar nor the dissent even acknowledges the existence of the legislature’s language, “whether arising out of governmental or proprietary function.” The reason is obvious: this language evinces the legislature’s express rejection of the exception to liability advocated by Dunbar and now by the dissent. To read language out of the Act, particularly language unambiguously expressing the legislature’s intent, is judicial activism and interventionism at their utmost.
Blithely skipping over this language of 1.C. § 6-903(a) (which by itself settles the question of the status of the “parallel function” exception), the dissent focuses upon the language in the same section “where the governmental entity if a private person or entity would be liable for money damages under the laws of the state of Idaho____” the dissent persists in Dunbar’s contorted reading of this language, to-wit, gratuitously adding the requirement of a parallel function in the private sector. This the dissent does without regard to the rules of construction (1) that the 1971 legislature adopted the construction of the federal jurisdiction (the federal jurisdiction rejected such a construction), and (2) that the Act is to be construed with its beneficent purpose in mind.
There is one exception in the federal case law, upon which the dissent seizes. That exception is where the plaintiff belongs to the military. So held the Supreme Court in Feres, supra. Notably, however, the holding in Feres was that, because “the relationship of military personnel to the Government has been governed exclusively by federal law,” there was no federal jurisdiction under 28 U.S.C. § 1346(b) for an action based on local law. Feres, supra, 340 U.S. at 142-26, 71 S.Ct. at 157-59 (emphasis added). Thus, the “parallel function” language is dicta. Id.; O. Reynolds, The Discretionary Function Exception of the Federal Tort Claims Act, 82 Georgetown L.J. 81, 99 (1968) (Indian Towing refused to extend “Feres dictum that the act had made the Government liable only where analogous private liability was present____”)2
*236Regardless, the Feres doctrine applies only in the above context, and no other. See United States v. Shearer, 473 U.S. 52, 105 S.Ct. 3039, 3043, 87 L.Ed.2d 38 (1985) (“Although the Court in Feres based its decision on several grounds, ‘[i]n the last analysis, Feres seems best explained by the “peculiar and special relationship of the soldier to his superiors, the effects of the maintenance of such suits on discipline, and the extreme results that might obtain if suits under the Tort Claims Act were allowed for negligent orders given or negligent acts committed in the course of military duty.” ’ United States v. Muniz, 374 U.S. 150, 162 [83 S.Ct. 1850, 1858, 10 L.Ed.2d 805] (1963), quoting United States v. Brown, 348 U.S. 110, 112 [75 S.Ct. 141, 143, 99 L.Ed. 139] (1954).”); 28 U.S.C.A. § 2674 note 26, Military Service Claims (Feres doctrine limited to circumstance of military personnel suing the government). The majority in Dalehite v. United States, 346 U.S. 15, 73 S.Ct. 956, 97 L.Ed. 1427 (1953) attempted to expand it further (to a firefighting function), but this part of Dalehite was overruled in Indian Towing and in Rayonier. See majority op., supra, pp. 761-762; United States v. Varig Airlines, 467 U.S. 797, 104 S.Ct. 2755, 2765 n. 10, 81 L.Ed.2d 660 (“[T]he holding of Rayonier obviously overrules one element of the judgment in Dalehite," that element being the attempt to expand the parallel function test beyond the Feres context). The Supreme Court (not to mention all other federal courts) has continued to limit the Feres doctrine to its facts (involving military personnel as plaintiffs) up to the present. See majority op., supra, at pp. 218-219, 723 P.2d at pp. 762-763 (discussion of Indian Towing, Rayonier, Hatahley, Muniz, Varig, Lockheed, and cases cited in 28 U.S.C.A. § 2680 and 36 A.L.R.Fed. 240). All of these cases read the language to make the government liable for the same causes of action as private individuals, regardless of the existence or nonexistence of a “parallel function.” If this were not so, why did the Supreme Court hold the government liable in Lockheed Aircraft Corp. v. United States, 460 U.S. 190, 198, 103 S.Ct. 1033, 1038, 74 L.Ed.2d 911 (1983), where its “function” was that of the military, the quintessential “unique governmental function”? See majority op., supra, at p. 218,723 P.2d at p. 762. There is no parallel function test in the federal jurisdiction outside the Feres context (involving military personnel as plaintiffs), as anyone with a rudimentary familiarity of the federal case law would know. The dissent is unable to cite one case outside that limited context. One commentator explained:
The significance of Indian Towing’s interpretation of the Act’s liability section is clear: If the Government undertakes an activity, even one not required, it will be held to the same standard of reasonable care as a private person would be, even though private persons seldom engage in that particular activity....
... Any distinction under the Act between governmental and proprietary functions was flatly rejected. It had long been assumed that when a municipality acted in a public capacity, it enjoyed immunity from liability. When it acted in a proprietary capacity, however, it enjoyed no such immunity unless granted by statute. The Indian Towing rejection of the distinction indicated that no provision of the Act would be interpreted as including it. This rejection was generally hailed as a desirable step due to the conflicting and confusing applications of the local rules of municipal liability. Reynolds, supra, pp. 99-100 (emphasis added) (footnotes omitted).
Indian Towing was no aberration as the dissent asserts; it was reaffirmed in Rayonier and still controls the construction of the language we concern ourselves with in this case. See Varig, supra, 104 S.Ct. at 2764-65, 2765 n. 10.
The dissent’s claim that Varig “in effect” overruled Indian Towing (not to mention all the other cases mentioned above which are consistent with Indian Towing) is patently false. The Varig court actually restated without contradiction and thus reaffirmed Indian Towing’s holding, *237thereby rejecting the proposed parallel function exception and overruling a minor part of Dalehite:
The [Indian Towing\ Court rejected the Government’s assertion, reasoning that it would “push the courts into the ‘non-governmental’-'governmental’ quagmire that has long plagued the law of municipal corporations.” 104 S.Ct. at 2764.
The Court also reaffirmed Rayonier’s holding to the same effect. Id. at 2765 n. 10. Contradicting the dissent’s mistaken claim that Dalehite “was based entirely” on Feres’ reading of 28 U.S.C. § 2674, Varig proceeded to explain that the main part of Dalehite on which it was relying dealt with the discretionary function exception found in 28 U.S.C. 2680(a), while the main part of Indian Towing and Rayonier dealt with § 2674. Varig, supra, 104 S.Ct. at 2764-65. Varig only reaffirmed that part of Dalehite dealing with the discretionary function exception; the short section extending the Feres doctrine it left overruled. Id. The dissent utterly fails to acknowledge this.
Contrary to the dissent’s reading of Varig in relation to Indian Towing and its progeny, Fifth Circuit Judge John R. Brown recently noted:
The FTCA is not, as government counsel think and continue to urge, confined to the typical fender bender automobile intersectional collision between a postal truck and a citizen’s child-filled station wagon. We have still the significant, still valid, decisions in Indian Towing, Rayonier and Eastern Airlines which recognized FTCA liability in areas traditionally thought to have some government activity immunity____ Collins v. United States, 783 F.2d 1225, 1231 (5th Cir.1986) (Brown J., concurring) (footnotes omitted).
An important reason Indian Towing et al. rejected the parallel function test is its inherent unworkability and flexibility to the point of eliminating any predictability. See majority op., at pp. 218-219, 723 P.2d at pp. 763-765. Presumably the dissent concedes this point, since it fails to respond to it. And yet the dissent would keep in place a test which potentially equates the functions of the National Park Service (functions which include those found “unique” in Dunbar) with those of the operators of a wild animal park. The parallel function test has no guidelines to lend it any certainty or rationality.
One danger of this test is that it potentially exposes the government to liability for policy decisions affecting vast segments of the public, so long as a “parallel function” can be found in the private sector. For example, Dunbar disagreed with the Dalehite holding (that the government was immune) on the basis of the existence of a parallel function, 100 Idaho at 531, 602 P.2d at 29; yet Dalehite involved high-level policy judgments which resulted in the injury and deaths of hundreds of people. Such an application of the parallel function test would have resulted in mammoth liability for what was a policy decision regarding a large number of fertilizer plants. In the dissent’s haste to uphold an ill-founded decision, it risks unwittingly opening up the government to extraordinarily broad liability based on the arbitrary and unpredictable results of the parallel function test.
The dissent utterly fails to refute the argument that most of the express exceptions of the Idaho Act are rendered mere surplusage by a parallel function exception which would easily encompass them. See majority op., supra, at pp. 221-222, 723 P.2d at pp. 765-766. The dissent thus impliedly concedes this argument as well. The dissent mischaraeterizes the argument as an assertion that the parallel function exception “conflicts with the enumerated exceptions contained in I.C. § 6-904, thus swallowing the general rule and rendering it a nullity.” The majority opinion goes quite to the contrary. The parallel function exception does not “conflict” with the exceptions; it swallows them (along with the general rule of liability) whole. For example, are not the activities of the Idaho National Guard unique governmental functions without parallel in the private sector? *238Of course they are. So why did the legislature include at § 6-904(6) an exception to liability concerning the activities of the National Guard, when the parallel function test of § 6-903(a) has already eliminated such liability? The answer must be obvious even to the most unreceptive set of ears — because § 6-903(a) creates no such exception.
Certainly, I.C. § 6-903(a) is a “qualified general rule of liability” as the dissent explains, to the extent that a plaintiff must state a claim under which the government is potentially liable as “a private person or entity would be liable____” However, as just explained, the “qualification” posed in Dunbar and now by the dissent is so broad that it would render those express exceptions dealing with “traditional governmental functions” redundant, unnecessary, and mere surplusage. Any “analytical” difficulties in understanding the implications of the purported parallel function exception lie with the dissent.
In reality, the question of the existence of a “parallel function” exception is easily answered. The federal case law states that there is none in the Federal Act, and I.C. § 6-903(a) says that there shall be none in the Idaho Act. The majority opinion simply acquiesces to reality.
B. The Discretionary Function Exception.
The dissent’s discussion of the discretionary function exception is more notable for what it fails to address than for what it does address. The dissent ignores the majority opinion’s analysis of the plain meaning of the exception; thus, it impliedly concedes that the two clauses of the exception establish a distinction between operational activities — those “in reliance upon or the execution or performance of” statutory or regulatory policy (which are afforded immunity only when carried out with “ordinary care”), and planning activities — those involving policy judgment and decision (which are afforded unqualified immunity). The dissent declines to examine the statutory language at all. If it did, it would find that ordinary care indeed is required in some circumstances. The dissent ignores the majority’s discussion of the federal legislative history; thus, it impliedly concedes that that history supports the planning-operational distinction. The dissent ignores the majority opinion’s discussion of all pre-1971 federal case law with the single exception of Dalehite, supra; thus, it impliedly concedes that the pre1971 federal case law aside from Dalehite (which case the majority opinion well explains, supra, at pp. 228-229, 723 P.2d at pp. 772-773) establishes the planning-operational distinction. Finally, the dissent (1) ignores the majority opinion’s long list of cases holding that a probation/parole officer’s supervision of a probationer/parolee is operational, thereby requiring ordinary care, and (2) fails to cite one on-point case to the contrary; thus, it impliedly concedes that every on-point case holds there to be no unqualified immunity under the discretionary function exception.
The cases the dissent does cite are considerably off-point. As the annotation it refers to explains, those cases generally deal “with the personal, civil liability of a policeman, a sheriff, or some other peace officer, or his bond, for injury or damage suffered by a third person because of the officer’s failure to enforce the law or arrest a lawbreaker.” Annot., 41 A.L.R.3d 700, 701 (footnotes omitted). Obviously, none of these eases involve a governmental official charged with the supervision of a person of known dangerous propensities. As the majority opinion notes, supra, at pp. 225-226, 723 P.2d at pp. 769-770, a multitude of decisions have held that the government is potentially liable in the instant context.
The dissent attempts to construe Sterling’s allegations to accommodate its authority. A review of the allegations as summarized in the majority opinion, supra, at p. 213, 723 P.2d at p. 757, belies the dissent’s Procrustean effort. Sterling did not simply allege that the probation officer was negligent for failing to arrest Bloom or revoke his probation. She alleged that the probationer failed to properly supervise Bloom and to properly enforce the order and agreement of probation. The probation officer could *239have and should have taken any number of steps short of arrest or revocation (the latter of which Sterling’s complaint acknowledged he had no authority to do). He could have ordered Bloom not to drive for recreational purposes, and not to drive a vehicle without written permission and without insurance, and to report regularly. He could have overseen Bloom’s conduct to prevent these violations. Sterling’s allegations are that he did none of this. Her allegations fall neatly within the confines of a permissible action against a governmental official or entity.
Naturally, Sterling’s claim would be barred by the holding of Chandler Supply Co., Inc. v. City of Boise, 104 Idaho 480, 660 P.2d 1323 (1983), but then so also would every claim brought against a governmental entity, whether it involved planning, policy-oriented activities or activities on the operational level. Justice Bakes, the author of Chandler, still offers no authority to support its unsupported holding. His dissent explains the holding of Chandler as follows:
This Court has traditionally been called upon in past decisions to draw the distinction between discretionary and ministerial action by governmental entities____ The holding in Chandler Supply Co. v. City of Boise, 104 Idaho 480, 660 P.2d 1323 (1983), is nothing more than a recognition of those prior determinations of whether or not challenged action constitutes discretionary action will be made in the same manner as it has always been made by this Court.
The dissent does not elaborate on exactly what the distinction is between discretionary and ministerial activity. One thing is certain, ministerial activity does not equate in Justice Bakes’ mind with activities putting policy into operation, since Chandler held that operational as well as planning activities are immunized. Id. at 485, 660 P.2d at 1328. As the majority opinion notes, to except from liability both planning and operational activities is to except all governmental activities. Majority op., supra, at p. 227, 723 P.2d at p. 771. If the dissent equates discretionary activities with “traditional governmental functions,” then it derives meaning from the discretionary function exception which is not there, it ignores United States Supreme Court authority to the contrary, and it ignores the pronouncement of I.C. § 6-903(a) that the government is potentially liable for “governmental” functions. See Majority op., supra, at p. 232, 723 P.2d at p. 776. In short, Chandler remains as it was, unsupported and unsupportable, unexplained and unexplainable — all of which makes it eminently overrulable.
Surprisingly, the dissent searches for support in United States v. Varig Airlines, 467 U.S. 797, 104 S.Ct. 2755, 81 L.Ed.2d 660 (1984). That decision is quite in line with earlier federal precedent and with the majority opinion. In fact, today’s majority opinion is consistent with the recent application of Varig in Collins v. United States, 783 F.2d 1225 (5th Cir.1986) (unanimous opinion).3
The Collins court accurately explained the context of Varig:
The two cases before the Court in Varig involved in-flight fires and consequent destruction of aircraft that the FAA or its predecessor had certified under a post-check certification program. In fact, the planes did not meet FAA standards and may never have been inspected at all. Plaintiffs alleged that the FAA would have discovered the defects had it conducted a plane-by-plane inspection of the aircraft instead of delegating responsibility for satisfying FAA standards to airplane manufacturers with periodic spot checks to encourage compliance. Id. at 1228.
Varig did not involve allegations of negligent inspection of aircraft or enforcement of the conditions of certification; rather, the Varig plaintiffs challenged the FAA’S *240policy decision of setting the conditions for certification themselves.
Varig’s context makes clear that when it speaks of the unqualified immunity of the government as a “regulator,” it means the government as the passer of regulations (in Varig, those regulations being the certification and inspection requirements of the two aircraft types). As the Collins court noted:
[T]his [reference to the role as regulator] does not mean that all regulatory acts are discretionary acts. Following Dalehite and Varig, we are instructed that there is discretion under § 2680(a) when “there is room for policy judgment and decision____” The exception therefore comprehends (1) “the initiation of programs and activities” and (2) “determinations made by executives or administrators in establishing plans, specifications or schedules of operations.” The exception extends to “acts of subordinates in carrying out the operations of government in accordance with official directions.” Dalehite, 346 U.S. at 35-36 [73 S.Ct. at 967] (emphasis added). Complemented with the learning from Varig we know that the United States is not liable for damages based on challenges necessarily directed at an agency’s discretion, if it exists, in determining the extent to which it will regulate or the manner in which this will be done, even if the regulations being enforced are mandatory. Throughout an analysis of the discretionary function exception we must be mindful that a purpose of the exception is to prevent judicial second-guessing of agency decisions that involve considerations of social, economic, or political policy. 104 S.Ct. at 2765.
In the context of a challenge to a subordinate’s conduct, Dalehite and Varig demonstrate that the subordinate’s conduct is not within § 2680(a) just because the conduct is regulatory in nature. Section 2680(a) protects only discretionary conduct; not all regulatory conduct is discretionary conduct. Discretionary conduct requires room for policy analysis and judgment. Dale-kite, 346 U.S. at 36 [73 S.Ct. at 968]. If an agency has the authority to do so, it may entrust to a subordinate’s discretion decisions on the extent or manner in which regulations, even mandatory regulations, will be enforced where the decision to be made entails balancing considerations of social, economic, or political policy. This does not necessarily mean, though, that the employee whose conduct is being challenged must have made a policy decision: if an employee acts in accordance with official directions, the conduct is within § 2680(a). Id. In both cases, § 2680(a) applies because challenging the employee’s conduct also challenges the agency’s discretion to implement a particular regulatory scheme. Vang, 104 S.Ct. at 2768. 783 F.2d at 1229-30 (emphasis added in second paragraph) (footnote omitted).
Sterling challenged not the Board of Corrections or the criminal court’s policy decisions concerning Bloom, but rather the implementation of that policy, which must be done “in accordance with official directions.” The dissent ignores the latter qualification, but it is key to the scope of immunity under the Federal Act. To ignore this qualification in order to extend immunity to all conduct which is regulatory in nature would be, as the Collins court observed, a radical and unwarranted extension of Varig. Id. at 1228.
In Collins, the plaintiffs alleged that government officials had negligently failed to properly classify a dangerous mine despite their statutory and regulatory obligation to do so. The Collins court held such negligence to be outside the ambit of the discretionary function exception to liability:
We disagree with the government that the questioning of any regulatory act necessarily involves judicial second-guessing. Determining whether a decision involves policy at all is not the same as scrutinizing the correctness of that policy decision. To the extent that an employee refuses to carry out a mandatory statute or regulation, one which *241leaves him no discretion in the manner or extent to which it will be enforced, the employee is simply disobeying a flat command. Id. at 1231 (emphasis added).
Sterling similarly alleges that Bloom’s probation officer disobeyed the commands of the agreement and order of probation. This he lacked the discretion to do. Accordingly, even under the holdings of Dalehite and Varig, viewed apart from the rest of federal case law (which of course we should not do), Sterling states a claim for negligent failure to carry out the policy of Bloom’s probation “in accordance with official directions.”
The extent to which the dissent misconstrues the case law on the discretionary function is revealed in its use of the following quote:
In the difficult cases it [the planning/operational distinction] will only “prove to be another example of a distinction ‘so finespun and capricious as to be almost incapable of being held in the mind for adequate formulation.’ ” Smith v. United States, 375 F.2d at 246. Dissenting op. of Bakes, J., infra.
The language the Smith court was quoting was from Indian Towing. Unfortunately for the dissent, what the Indian Towing Court was referring to was not the planning-operational distinction, but was the distinction between governmental and nongovernmental functions, the very distinction which the Chandler holding appears to advocate. Here as elsewhere, when analyzed, the authority to which the dissent resorts cuts against its position rather than for it.
II. TORT LAW
As with the authority the dissent relies on concerning governmental immunity, the cases and commentaries concerning tort law upon which it relies, far from supporting its position, actually undermine it even further. Furthermore, the arguments the dissent makes in Part II contradict the record and our standard of review.
A. New Causes of Action.
The dissent begins by contending that the majority has interpreted the Idaho Tort Claims Act as creating a new cause of action — the negligent supervision of a probationer. This, the dissent argues, the Act forbids. The dissent misconstrues the majority opinion.
Nowhere does the majority hold that the Act creates the tort of negligent supervision, and in no way can the Act be viewed as creating this tort or any other tort. What the majority opinion does say about Idaho’s Act, and what the United States Supreme Court has said about the Federal Act, is simply that with the enactment of these two Acts, the Idaho legislature and Congress both intended to make governmental entities responsible for their conduct in certain circumstances, whereas in the past they had enjoyed absolute immunity.4 Thus, it is easy to understand why the Supreme Court found this change in the law by Congress “novel and unprecedented,” Rayonier, supra, 352 U.S. at 319, 77 S.Ct. at 377, given the fact that up to the enactment of the Federal Act, federal government entities enjoyed absolute immunity. This same reasoning applies to the Idaho Act and the Idaho legislature’s intent. Clearly, the majority has not created a new tort out of the Idaho Act, but merely made the state liable for the same torts as are private persons and entities. The dissent’s characterization of what the majority has said on this point is therefore erroneous.
The dissent’s reliance upon I.C. § 6-903(f) also is groundless. That section is clear and straightforward. It states in per*242tinent part: “[Njothing in this act shall enlarge or otherwise adversely affect the liability of an employee or a governmental entity. Any immunity or other bar to a civil lawsuit under Idaho or federal law shall remain in effect.” What the first sentence of § 6-903(f) does is prohibit the creation of causes of action based on the fact that the defendant is a governmental entity. In other words, the viability of a cause of action will not be dependent upon whether the defendant is a private individual or a public entity. The second sentence simply makes clear that Idaho’s Act does not preempt other specific grants of immunity. For example, I.C. § 20-231 specifically immunizes public entities and public employees from injuries resulting from a decision either to place a prisoner on parole or to revoke a prisoner’s grant of parole. Thus, pursuant to § 6-903(f), that specific grant of immunity is valid regardless of the scope of liability under the Act.
The dissent misunderstands both what the majority and the Supreme Court have said regarding the changes wrought by the respective tort claims acts, and also what § 6 — 903(f) states. The novelty of the acts is that they made formerly immune governmental entities responsible for their conduct; tort law itself remained unchanged. If there be anything “novel and unprecedented” relative to this issue, it rests solely in the dissent’s reasoning process.
B. Duty.
The majority opinion more than adequately explains the basis for the duty that the probation officer owed to those people foreseeably endangered by Bloom. The cases and commentaries that the dissent relies upon are either unpersuasive or not on point. I will address each of them in turn.
The dissent begins its analysis by citing Prosser & Keaton, The Law of Torts, §§ 37, 53 (5th ed. 1981). Conveniently unmentioned is section 56, from which the majority, at pp. 224-225, 723 P.2d at pp. 768-769, extensively quotes, and which unequivocally states that with respect to relationships that are custodial in nature, the individual who has custody of another has a duty to control his or her charge and “to guard other persons against his dangerous propensities.” Prosser, supra, § 56. Alas, the dissent can derive no support from Dean Prosser and Professor Keaton.
The dissent next cites Jacobson v. McMillan, 64 Idaho 351, 132 P.2d 773 (1943), for support. Because this is a prior decision of this Court, albeit now 43 years old, a close look is warranted. In Jacobson, as the dissent states, Dan O’Connor attempted to shoot his wife. A criminal complaint was filed and the husband was arrested by defendant McMillan, a sheriff, and taken before a probate judge for a preliminary hearing. The judge ordered O’Connor committed to the Kootenai County jail pending trial on charges of assault with a deadly weapon. Although an information was filed, no trial was held and no bond was ever furnished for O’Connor’s release. The commitment order was never rescinded. Sheriff McMillan, believing O’Connor was suffering from a mental defect, transferred O’Connor outside the county to the State Hospital in Blackfoot for observation. There, O’Connor was released to the custody of the hospital Superintendent, Cromwell. While still in the custody of Cromwell, O’Connor was permitted to go out of the hospital unattended and onto the hospital grounds. As a result, O’Connor escaped from the hospital. Several months later, with shotgun in hand, O’Connor returned to his wife’s residence and, in attempting to shoot her, wounded Jacobson who had been specifically hired to protect the wife from O’Connor.
Jacobson brought suit against both Sheriff McMillan and Superintendent Cromwell and their respective sureties. In his complaint, Jacobson alleged that his injuries were the proximate result of the negligence of the sheriff and the superintendent whose acts were in direct violation of the court order committing O’Connor to the county jail. Defendants moved to dismiss the complaint, which the trial court grant*243ed. On appeal, this Court in a 3-2 vote affirmed.
The majority’s holding in Jacobson is not based upon any conclusion that there was no duty owed by Sheriff McMillan to plaintiff Jacobson, as the dissent asserts. That the Jacobson majority contains some language to this effect in the opinion is true, but it is clear from reading the majority and dissenting opinions that the language was mere dicta.
The majority began its analysis by stating that Jacobson knew of O’Connor’s “affection with homicidal insanity.” Id. at 357, 132 P.2d at 776. The majority further stated: “Possessed of such knowledge, appellant [Jacobson] took employment and voluntarily placed himself in a position, subjecting himself to any attack that might be made by O’Connor.” Id. at 357, 132 P.2d at 777 (emphasis added). The majority held that this decision to work for O’Connor’s wife was the “proximate cause of appellant’s injuries.” Id. (emphasis added). The majority went on to say that Sheriff McMillan and Superintendent Cromwell “were in no way responsible for appellant’s presence at the home of Grace Work O’Connor [the wife] and neither directly nor indirectly subjected appellant to the danger from which he suffered.” Id.
The majority concluded that “[t]he acts of the sheriff and the superintendent, in permitting O’Connor to escape and be at large, were only a remote cause in comparison with the negligence of appellant, which could not have been reasonably foreseen or anticipated____” Id. at 360, 132 P.2d at 778 (emphasis added). Thus, the majority’s reason for affirming the dismissal of the complaint was based upon principles of assumption of risk and contributory negligence, which, at that time and until the advent of comparative negligence in 1971, operated as a complete bar to recovery; the Court’s decision was not based upon the absence of a duty. In closing, the majority stated:
[A]ppellant (according to his complaint) had every reason to believe that he would be attacked by O’Connor, if the latter should escape or be released and found appellant at the O’Connor residence. He hired to take the risk.
Appellant knew of the dangers of employment, when he entered the service, and while he did not know and could not foresee the escape of O’Connor from custody, still his employment, in anticipation of some such exigency, was for the purpose of guarding against just such a contingency as arose.
It may be urged that the action of respondents, sheriff and superintendent, in permitting O’Connor to escape, so increased or augmented the dangers which appellant risked as to render them legally liable for negligence. The answer, however, to such a suggestion is, that, under appellant’s complaint, there would have been no such danger to increase or augment, had appellant not undertaken the hazardous task he assumed. Id. at 360-61, 132 P.2d at 778 (emphasis added).
It is interesting to note that Jacobson has been cited to only once by this Court. In Weaver v. Pacific Finance Loans, 94 Idaho 345, 347, 487 P.2d 939, 941 (1971), Chief Justice Donaldson cited Jacobson for the proposition that in order for plaintiff’s allegations to merit recovery, the “defendant’s conduct [must be] shown to be the proximate cause of plaintiff’s injury. Jacobson v. McMillan, 64 Idaho 351, 132 P.2d 73 (1943).” (Emphasis added.) This is exactly what Jacobson stands for, and no more than that.
Thompson v. County of Alameda, 27 Cal.3d 741, 167 Cal.Rptr. 70, 614 P.2d 728 (1980), is the next case the dissent cites. It, too, provides no support. First of all, Thompson involved the alleged negligent release by the county of a juvenile offender — not the continuing supervision of a probationer or parolee. Our facts, as the dissent knows, involve the latter. Second, and more importantly, Thompson’s rationale for its holding is not applicable here. In Thompson, parents sued Alameda County for the wrongful death of their five-year-old son, alleging that the county had *244acted recklessly in releasing from custody a juvenile delinquent who was known to have dangerous and violent propensities toward young children, and who, twenty-four hours after being released to the custody of his mother, murdered the plaintiffs’ son. The California Supreme Court rejected the plaintiffs’ argument, stating that Alameda County had no affirmative duty to warn the plaintiffs of the juvenile delinquent’s dangerous tendencies upon its decision to release the juvenile. The court reasoned that “plaintiffs’ decedent was not a known, identifiable victim, but rather a member of [a] large amorphous public group of potential targets.” Id. 167 Cal.Rptr. at 80, 614 P.2d at 738.
I have no quarrel with Thompson. It denied liability because there was no “special relationship ” between the county and either the victim or the person committing the harm. (Here there is such a relationship between the Board and the perpetrator.) The Thompson court, however, approvingly quoted and applied §§ 315-319 of the Restatement (Second) of Torts. Said the Thompson court:
Likewise in Tarasoff we were concerned with the duty of therapists, after determining that a patient poised a serious threat of violence, to protect the “foreseeable victim of that danger.” (Tarasoff, supra, 17 Cal.3d [425] at p. 439, 131 Cal.Rptr. 14, 551 P.2d 334.) In reaching the conclusion that the therapists had a duty to warn either “the endangered party or those who can reasonably be expected to notify him, ...” (id. at p. 442, 131 Cal.Rptr. at p. 27, 551 P.2d at p. 347), we relied on an exception to the general rule that one owes no duty to control the conduct of another. (Id., at p. 435, 131 Cal.Rptr. 14, 55 P.2d 334; see Rest.2d Torts (1965) §§ 315-320.) As declared in section 315 of the Restatement such a duty may arise if “(a) a special relation exists between the actor and the third person which imposes a duty upon the actor to control the third person’s conduct OR (b) a special relation exists between the actor and the other which gives the other a right to protection.” Id. 167 Cal.Rptr. at 75-76, 614 P.2d at 733-34 (emphasis added).
This reasoning is applicable here, and argues for the existence of a duty where a special relationship exists between an actor (such as a probation officer) and a third person in need of control (such as a probationer).
The California Supreme Court has continued to quote from and apply § 315 of the Restatement. See Peterson v. San Francisco Community College District, 36 Cal.3d 799, 205 Cal.Rptr. 842, 685 P.2d 1193, 1196 (1984); Davidson v. City of Westminster, 32 Cal.3d 197, 185 Cal.Rptr. 252, 649 P.2d 894, 897 (1982).
The dissent states that Thompson as well as cases from Washington, New York, Kansas, and Florida stand for the proposition that except where there are sufficient personal contacts by the police or probation officers with a particular individual upon which that individual relies for safety or protection, there is no legal action. This is simply false. As the above quote from Thompson states, not only does a duty arise where there is a special relationship between a government official and the victim, a duty also exists where there is a special relation between the government official and the third person who commits the harm.5
That duty is reflected in Restatement § 315, which Thompson, Peterson, and Davidson all quoted and applied. The majority opinion establishes the same general duty in its quotation from Dean Prosser on page 25: “The general duty which arises in many relations to take reasonable precautions for the safety of others may include the obligation to exercise control over the conduct of third persons.” For this proposition Dean Prosser cites in a footnote none other than § 315.
*245Section 319, which the majority opinion adopts, merely delineates the specific special relationship between those charged with the supervision of persons with dangerous propensities and their charges. It hardly can be argued that California has rejected § 319 when it has expressly adopted § 315, of which § 319 is but a subcategory. The California court, like all others of which I am aware to date, recognizes that a duty arises out of certain special relationships with third persons. Just such a relationship existed between Bloom and his probation officer.
The New York case cited to by the dissent, Sorichetti v. City of New York, 65 N.Y.2d 461, 492 N.Y.S.2d 591, 482 N.E.2d 70 (1985), and the Washington case, Chambers-Castanes v. King County, 100 Wash.2d 275, 669 P.2d 451 (1983) similarly fail to support it. Both involved only allegations that police officers had failed to properly respond to the plaintiffs’ call for help; neither case involved any third party at all. In addition, in Chambers-Castanes, the Washington court stated that the plaintiffs had alleged facts, which, if found to be true, would correctly place them in a special relationship with the police and sent the case back for trial. Chambers-Castanes, supra, 669 P.2d at 457.
Furthermore, the dissent neglects to cite another 1983 Washington case, Petersen v. State, 100 Wash. 421, 671 P.2d 230 (1983), wherein the Washington Supreme Court expressly adopted and applied § 315 of the Restatement (Second) of Torts. In Petersen, the court held that a state psychiatrist who had diagnosed as schizophrenic an individual whose vehicle subsequently struck and injured the plaintiff at an intersection had a duty to take reasonable precautions to protect a person such as the. plaintiff who might foreseeably be endangered by the individual’s drug-related mental problems. Id. at 236-37. If further support needs to be provided for the virtually universal acceptance of § 315, the Washington Court provided it, citing to Lipari v. Sears, Roebuck & Co., 497 F.Supp. 185, 188 (D.Neb.1980); Bradley Center, Inc. v. Wessner, 161 Ga.App. 576, 287 S.E.2d 716 (1982), aff'd 250 Ga. 199, 296 S.E.2d 693 (1982); McIntosh v. Milano, 168 N.J.Super. 466, 403 A.2d 500 (1979). Petersen, supra, 671 P.2d at 236.
In Sorichetti, the New York Court of Appeals affirmed a jury verdict in favor of the plaintiff, stating that the facts in the case showed that a “ ‘special relationship’ existed between the police and [the plaintiff] such that the jury could properly consider whether the police conduct satisfied the duty of care owing to [plaintiff].” Sorichetti, supra, 492 N.Y.S.2d at 597-98, 482 N.E.2d at 471. The New York Court purported in no way to restrict liability to special relationships between the police and the victims. It did not need to deal with any special relationship existing between the police or a parole or probation officer and the person doing the harm because no such relationship was alleged or argued. Consequently, no such discussion was necessary to resolve the issue before it.
A fair reading of Sorichetti and Chambers-Castanes, which both recognized liability on the part of governmental conduct where special relationships exist, reveals that they are harmonious with § 315, and, far from supporting the dissent’s position, actually undermine it further. That fact is further validated by the Washington Supreme Court’s subsequent adoption of § 315 in Petersen, of which § 319 is a subcategory. The same can be said for the Kansas and Florida cases, Everton v. Willard, 468 So.2d 936 (Fla.1985), and Dauffenbach v. City of Wichita, 233 Kan. 1028, 667 P.2d 380 (1983).
In Everton, the Florida Supreme Court had before it the issue of “whether to enforce the law by making an arrest is a basic judgmental or discretionary governmental function that is immune from suit, regardless of whether the decision is made by the officer on the street, by his sergeant, lieutenant or captain, or by the sheriff or chief of police.” Everton, supra, 468 So.2d at 937 (emphasis added). The court *246answered the issue in the affirmative, relying upon Trianon Park, supra, which held that governmental entities are immune when they are exercising their discretionary power to enforce compliance with laws.
Significantly, in arriving at this conclusion, the court additionally stated: “if a special relationship exists between an individual and a governmental entity, there could be a duty of care owed to the individual.” Id. at 938 (emphasis added). No such relationship, however, existed in Ever-ton: the police had not accepted the specific responsibility for the protection of the victim, and there was no evidence that the State had any custodial relationship or other responsibility for the tort-feasors’ behavior. Accordingly, rather than rejecting the Restatement (Second) of Torts §§ 315-319, Everton is a proper application of those sections. Furthermore, the holding in Everton is grounded in discretionary immunity and not in any rejection of the tort of negligent supervision.
In Dauffenbach, the plaintiff sued two police officers and the City of Wichita for the excessive use of force in arresting the plaintiff. The facts conclusively show that the police officers in the case had no “special relationships” with the plaintiff. In fact, like Sorichetti, supra, and ChambersCastanes, supra, there was no third party involved at all in Dauffenbach. Accordingly, as is the case with Everton, Sorichetti, Thompson, Chambers-Castanes, and Petersen, Dauffenbach is but an application of §§ 315-319. The cases upon which the dissent relies, far from sustaining it, unequivocably contradict its position.
The dissent next argues at page 263, 723 P.2d at page 807 that the majority today creates “a new tort of negligent supervision.” This is incorrect. As the majority notes, this Court in Doe v. Durtschi, supra, recognized that there could be a cause of action for the negligent retention and supervision of a school teacher who later commits a tort on a third person, relying in part upon the Restatement (Second) of Torts, § 449. There is no benefit in belaboring the obvious, other than to point out that the majority opinion today has created no “new tort law”; the majority opinion on the issue of duty is merely an application of universally accepted concepts of tort law — which are found both in Idaho and elsewhere. The majority’s recognition of the tort of negligent supervision is but the application of Durtschi and the overwhelming majority of case law from other jurisdictions.
In footnote 14, p. 264, 723 P.2d at p. 808, the dissent concludes its argument on this point by stating that this Court has rejected the tort of “negligent supervision” in Pedigo v. Rowley, 101 Idaho 201, 610 P.2d 560 (1980). This, too, is incorrect. In Pedigo, the defendant filed a claim for contribution against a parent, arguing that the parent’s negligent supervision of the plaintiff-child contributed to the injuries of the child. The Court held that the doctrine of parental immunity barred the action, not that there was no such tort. Id. at 205, 610 P.2d at 564.
Furthermore, as counsel pointed out at oral argument, Pedigo does not stand for the proposition that where a negligently supervised child commits a tort against a third party, there is no action against the parent by the third party. In Pedigo the plaintiff-child was floating on an air mattress in the lake and did not hurt anyone when she was run over by a boat; there is simply no allegation that the child hurt anyone due to her parent’s negligent supervision. Here we are dealing with a fact pattern where a negligently supervised probationer physically committed a tort upon another person. Thus, not only does Pedigo not stand for what the dissent says it stands for, it also is not factually analogous.
C. Proximate Cause.
The dissent’s arguments on proximate cause need not be discussed at length. The dissent has forgotten that for purposes of this appeal, the defendant — the State of Idaho — has admitted that the acts of its probation officer were both negligent and the proximate cause of the plaintiff’s damages. In its brief, the defendant states: *247“To the extent that on a Motion for Judgment on the Pleadings, the Defendant admits the allegations within the Plaintiff’s complaint, the Defendant agrees with the Plaintiff’s statement of the facts.” Respondent’s Brief, p. 1 (emphasis added). In appellant’s brief and in the record before us, it has been alleged that the probation officer’s admitted negligence “proximately caused” the plaintiff's injuries. Accordingly, for the purposes of this appeal, there is no issue of proximate cause, and the dissent’s discussion on the matter serves merely to confuse and obfuscate the issues.
Contained within the dissent’s discussion of proximate cause, however, are allegations that amount to nothing more than scare tactics. First is the allegation that “[t]he effect of the majority opinion today is that every probationer who violates any term of his probation must be arrested, or the state will be liable in tort for any subsequent violations which that probationer may incur.” This is patently untrue. As previously explained, Sterling did not allege that the state was required to arrest Bloom, but merely to use “ordinary care” in his supervision. If the probation officer did supervise Bloom with “ordinary care” (which obviously does not necessarily involve arrest), then there will be no liability. The dissent apparently has forgotten that the result of today’s opinion is not to hold the state liable for the allegedly negligent acts of its probation officer in this case, but to hold that the state may be liable depending upon a determination of the facts by a jury. The majority opinion, of course, intimates no view about the defendant’s culpability.
The dissent’s next assertion is equally off-base. It is that parole and probation officers will be unable to exercise discretion in managing their charges, and that today’s opinion “substitutes the tort law system for planning and policy decisions made by the Idaho legislature.”6 In no way does today’s opinion substitute Idaho’s tort law system for the present parole and probation system. The Idaho legislature made no decision on the supervision of Bloom, which is the only question before us. As for the Board, its policy decisions regarding probation and parole remain immune under the “discretionary” function exception; only the implementation of policy decisions carries with it the requirement of ordinary care. In short, far from creating chaos, the majority opinion today brings order to the chaos and confusion that was created by Chandler, supra, and Dunbar, supra; far from requiring new legislation, the legislature can be assured that the incorrect interpretation and emasculation of its Tort Claims Act perpetrated by Dunbar and Chandler have been corrected, and the wrongs created by Chandler and Dunbar have been remedied.
D. Course and Scope of Employment.
The dissent’s final argument in Part II is that the state is immune because its probation officer’s negligent conduct placed him outside the scope of his employment. Again, this argument is irrelevant, since the Board conceded Sterling’s allegation that the probation officer was acting within the scope of his employment. Aside from this conclusive fact, the dissent’s argument misconstrues employer-employee law.
Were the dissent’s argument valid, there would never be employer liability for the negligent acts of its employees, because, by the dissent’s definition, those negligent acts would place the employee outside the scope of his employment. In short, the dissent’s argument would lead to the conclusion that the doctrine of respondeat superior is void. A review of a wealth of case law shows the fallacy of such views.
The universal rule of respondeat superi- or is that an employer is liable for negligent acts or omissions of its employee committed in the scope of his or her employment. Scrivner v. Boise Payette Lumber Co., 46 Idaho 334, 268 P. 19 (1928); *248Smith v. Thompson, 103 Idaho 909, 911, 655 P.2d 116, 118 (Ct.App.1982); see also Williams v. Alyeska Pipeline Service Co., 650 P.2d 343 (Alaska 1982); State ex rel. City of Havre v. District Court of Twelfth Judicial District in and for Hill County, 187 Mont. 181, 609 P.2d 275 (1980), cert. denied, 449 U.S. 875; Nabors v. Harwood Homes, Inc., 77 N.M. 406, 423 P.2d 602 (1967). The test for whether an employee was acting within the scope of his employment when he or she committed a tort is the “right to control reserved by the employer over the functions and duties of the agent.” Van Vranken v. Fence-Craft, 91 Idaho 742, 747, 430 P.2d 488, 493 (1967); Koch v. Elkins, 71 Idaho 50, 57, 225 P.2d 457, 462 (1950). In other words, the doctrine of respondeat superior is applicable during the period of time in which the principal has the right to control the employee’s actions. McCauley v. Ray, 80 N.M. 171, 453 P.2d 192 (1968). This analysis focuses on whether the act or omission is of a kind that the employee was hired to do or supposed to do, and whether the act or omission occurred substantially within the authorized limits of time and space. Stanfield v. Laccoarce, 284 Or. 651, 588 P.2d 1271 (1978).
Applying this test, it is clear that Sterling has adequately alleged facts that, if found to be true, would place Bloom’s negligence within the course and scope of his employment. The alleged negligence — a failure to supervise Bloom in a reasonable way — involves conduct for which the probation officer was specifically hired. The mere fact that this conduct may be negligent does not take it outside the course and scope of the probation officer’s employment, as both Idaho case law and case law from other jurisdictions point out.
It is clear that the dissent’s novel view of respondeat superior is not the law of Idaho or of any other jurisdiction which I could discover. Therefore, its arguments on this point are without merit.7
III. PROSPECTIVE APPLICATION
The dissent’s last argument is that today’s decision should only apply prospectively, apparently excluding even Sterling from the benefits of her appeal. Such an argument contradicts controlling law on this matter. In Jones v. Watson, 98 Idaho 606, 608, 570 P.2d 284, 286 (1977), this Court stated that “the determination of whether an overruling decision will be applied retroactively or prospectively, is a matter left to state courts for determination on a case-by-ease basis.” Quoting from Warwick v. State ex rel. Chance, 548 P.2d 384, 393 (Alaska 1976), the Court continued: “ ‘A state supreme court has unfettered discretion to apply a particular ruling either purely prospectively, purely retroactively, or partially retroactively, limited only “by the juristic philosophy of judges ... their conceptions of law, its origin and nature.” The decision is not a matter of law but a determination based on weighing the merits and demerits of each case.’ ” Jones, supra, 98 Idaho at 608, 570 P.2d at 286 (emphasis added). In making this determination, three criteria are considered: “1) the purpose of the decision, 2) reliance on the prior rule of law, and 3) the effect upon the administration of justice.” Id. at 609, 570 P.2d at 287.
In Thompson v. Hagan, 96 Idaho 19, 25, 523 P.2d 1365, 1371 (1974), this Court noted three different approaches to retroactivity:
The first approach is the traditional rule which is derived from the concept that courts do not pronounce new law, but only discover the true law. Under this approach there are no new decisions, but only clarifications of the true law which makes a decision applicable to both past and future cases. The second approach *249is the prospective rule. Under this rule a decision is effective only in future actions, and does not affect the rule of law in the case in which the new rule is announced. The third approach is the modified prospective rule which is a combination of the traditional and prospective rules. Under the modified prospective rule, the new decision applies prospectively and to the parties bringing the action resulting in the new decision; or to the parties bringing the action and all similar pending actions. (Emphasis added.)
After considering each approach, the Court adopted the third approach in Thompson, holding that the effect of its ruling — a decision that Idaho’s guest statute was unconstitutional — would apply to the parties of the case, all pending actions at the date of this decision, and all actions arising in the future. Id.
In Dawson v. Olson, 94 Idaho 636, 639, 496 P.2d 97, 100 (1972), this Court acknowledged the “plaintiffs’ legitimate interest in compensation after undertaking the effort and expense of bringing the issue before us.” Accord Sims v. State, 94 Idaho 801, 803, 498 P.2d 1274, 1276 (1972). In Dawson, we specifically stated:
In the torts area, the likelihood that an injured party will undertake extended and costly litigation without hope of compensation, simply to establish a new doctrine, is slight unless the injury is recurring in nature. Smith arose from a single incident, an automobile accident on a bridge maintained by the Department of Highways. Only by applying the decision to the plaintiffs at bar could their legitimate interest in compensation, after trying and appealing the issue, have been satisfied. Dawson, supra, 94 Idaho at 639 n. 6, 496 P.2d at 100 n. 6 (emphasis added).
In Smith v. State, 93 Idaho 795, 808, 473 P.2d 937, 950 (1970) — the case upon which the dissent hangs its argument — this Court applied its holding (that the doctrine of sovereign immunity should be abolished) to the parties to the appeal, and to all causes of action arising subsequent to 60 days after the adjournment of the First Regular Session of the Forty-First Idaho State Legislature.
Thus, it is clear that the dissent has cited not one case that would countenance its apparent proposal of declining to apply the majority’s holding to the parties to this appeal. Reason also defies that which the dissent would do.
First, the action by this Court in Smith —abolishing the judicial doctrine of sovereign immunity — is far more dramatic than the reinterpretation and setting aright of that which the Idaho legislature had provided. Thus, if, under the circumstances of Smith, this Court was willing to apply its result to the parties of that case, then a fortiori, this Court should be willing to apply its holding today to the parties of this case. In contrast with Smith, today’s decision does not change a judge-made doctrine; it restores a statute to its original intent. To fail to apply our holding at least to the instant parties would be to prolong the thwarting of the legislature’s will. Second, as the Dawson Court put it: “Only by applying the decision to the plaintiff at bar could [her] legitimate interest in compensation, after trying and appealing the issue, have been satisfied.” Dawson, supra, 94 Idaho at 639 n. 6, 496 P.2d at 100 n. 6.
Third, the overruling of Dunbar and Chandler can hardly be viewed as unexpected. Over one year ago, I unequivocably stated that Dunbar and Chandler did not command a majority of this court, quoting from the dissenting opinions filed in Chandler by Justices Donaldson and Bistline. Merritt v. State, 108 Idaho 20, 27, 696 P.2d 871, 878 (1985) (Huntley, J., dissenting). From that date until now, it was clear to all those who could add two plus one that both Chandler and Dunbar were terminally ill.
In summary, the dissent’s argument on this matter is without authority or reason. Because this issue was not raised on appeal, it was not briefed. It was for this reason that the majority opinion did not *250address the issue. Suffice it to say, however, that the opinion applies at least to the parties to this case, as in Smith, supra. Any other result would produce the type of injustice that the people of Idaho have had to endure under the now defunct opinions of Dunbar and Chandler8
CONCLUSION
Today’s decision restores balance to a system that had been seriously disrupted and skewed. The importance of this balance is essential not only because principles of fairness require it, but also because the very principles of democracy demand it, I close with that which I have previously stated:
Recent decades have seen dramatic growth in the body of constitutional law as citizens have sought protection of their individual and collective rights through the nation’s courts. This may be partly a result of the growth of government, which has great potential to overlook the interests of individuals and minority groups as it increases in size and complexity. It is also likely due to an increasing awareness of the relationship of government to its citizens. We recognize the need for a continuing balance between the interests of government and the interests of individuals.
In a democracy, the government is permitted to assume its role of governing only at the sufferance and with consent of those governed. Thus, when the government acts arbitrarily or unfairly in its dealings with the people, the basic foundations of the democracy are weakened. To the extent a government wrongs its citizens and permits that wrong to go without redress, it loses a part of its license to govern.
It is not surprising that recent decades have also witnessed a surge of developments in the area of governmental liability in tort. Both constitutional law and governmental liability law represent an attempt to reconcile the tension between
the power of government and the rights and integrity of the individual. There is only slight conceptual difference between the due process clause of our state or federal constitution and the principle of governmental accountability for its torts. In the former, citizens are assured that the state cannot take from them their rights or property without due process, and property may not be taken in eminent domain unless the citizen receives just compensation. In the latter, government or its agents may not injure citizens or their property in the course of government activities without affording them a proper remedy.
Because the doctrine of governmental immunity in tort is so much involved with the fundamental relationship of government to its citizens, it is important that courts and legislatures alike give careful consideration to all laws which deal with it. The judiciary should be particularly conscious of its determinations, and view them in light of the underlying considerations at stake. What may be at stake is the citizens’ respect for, and trust in government. Certainly the judiciary should take care to discern the objectives of the people as they are expressed through their elected representatives, and enacted into law. Where the legislature has acted to resolve the problems created by governmental immunity, judicial action should be all the more considered. In the words of Justice Cardozo: “The exemption of the sovereign from suit involves hardship enough where consent has been withheld. We are not to add to its rigor by refinement of construction where consent has been announced.” [Anderson v. John L. Hayes Const. Co., 248 N.Y. 140, 147, 153 N.E. 28, 29 (1926).]
Huntley, Prologue, Sovereign Immunity and Reemergence of the Governmental/Proprietary Distinction: A Setback in Idaho's Governmental Lia*251bility Law, 20 Idaho L.Rev. 197-98 (1984).
BISTLINE, J., concurs.. Such language is noticeably absent from the statutes of Ohio and Florida, on which jurisdictions the dissent relies. (The Washington case cited by the dissent is discussed infra, n. 2). Even when this critical distinguishing factor is ignored, the two cases cited by the dissent still are to no avail to it.
Shelton v. Industrial Commission, 51 Ohio App.2d 125, 367 N.E.2d 51 (1976), contains no hint of an exception based on the unique or unparalleled nature of the government’s underlying function or role. The Shelton court held just as this Court holds today that the Ohio Act “provides a remedy for existing duties where the state was previously immune from suit and a private party under similar circumstances would be liable. ” Id. at 367 N.E.2d 54 (emphasis added). The difference between Shelton and the instant case is that the Shelton court found no duty on the part of the state running to the plaintiffs. Id. Here, as discussed subsequently, the state’s special relationship with Bloom created a duty to those Bloom foreseeably might injure.
Trianon Park v. City of Hialeah, 468 So.2d 912, 917 (Fla.1985), a 4-3 decision, likewise held that under the Florida Act, "the identical existing duties for private persons apply to governmental entities.” As with the Shelton court, the Trianon Park court found there to be no duty on the part of the governmental entities involved. Id. at 919-23. In addition, the court’s holding that no duty existed was grounded on its holding that the governmental functions involved were "discretionary.” Id. In the subsequent case of Everton v. Willard, 468 So.2d 936, 938 (Fla.1985), the Florida court (1) explained Trianon Park as involving “discretionary” functions, and (2) held that in some circumstances police officers are subject to liability even though their function is uniquely governmental.
The three state cases upon which the dissent relies fail to support his argument; in contrast, one can easily find a host of state decisions construing their acts as does today’s majority. See, e.g., First Insurance Co. of Hawaii, Ltd. v. International Harvester Co., 66 Hawaii 185, 659 P.2d 64, 67 (1983) (rejects the confusing distinction between governmental and proprietary functions; holds government liable as a private person would be); Ryan v. State, 134 Ariz. 308, 656 P.2d 597, 598 (1982) (State argued that "governmental immunity should continue in those instances where the services provided are uniquely governmental in nature. This is the old proprietary-governmental distinction in a bright new word-package."); Jennings v. State, 566 P.2d 1304, 1313 (Alaska 1977) (failure to sign a road is actionable although a traditional governmental function); City of Pittsburg v. Commonwealth, 468 Pa. 174, 360 A.2d 607, 610 n. 4 (1976) (“This Court has recognized the difficulty in applying the governmental-proprietary distinction and has abandoned the distinction in dealing with tort liability.'); Johnson v. State, 69 Cal.2d 782, 73 Cal.Rptr. 240, 447 P.2d 352, 362-63 (1968) (State potentially liable for negligence of parole officer — filling a traditional governmental function); see also cases cited in the majority op., supra, at p. 226, 723 P.2d at p. 770.
At any rate, as Justice Bakes acknowledged in his dissent to Doe v. Durtschi, 110 Idaho 466, 716 P.2d 1238 (1986), since the 1971 legislature largely adopted the Federal Tort Claims Act, we primarily must look to the pre-1971 federal case law. The message of that casé law is loud and clear: there is no exception granting immunity for uniquely governmental functions.
. The Supreme Court of Washington in dicta appeared to adopt Feres’ doctrine of immunity for the analogous context of a member of the Washington Air National Guard suing the state. Edgar v. State, 92 Wash.2d 217, 595 P.2d 534, 538-39 (1979), cert. denied 444 U.S. 1077, 100 S.Ct. 1026, 62 L.Ed.2d 760. This hardly constitutes an adoption of a parallel function exception for other contexts as the dissent suggests. In fact, in Chambers-Castanes v. King County, 100 Wash.2d 275, 669 P.2d 451, 452 (1983), the Washington Supreme Court held that a governmental entity was potentially liable even when engaged in the function of law enforcement, a traditional governmental function without parallel in the private sector if there ever was one. The Edgar dicta, like the Feres dicta, subsequently has not been cited outside the context of military personnel as plaintiffs. In fact, it has not been cited at all.
At any rate, like Feres, the Edgar language referring to parallels in the private sector was dicta (and received no headnote from the publisher), since the court held that the plaintiff as an "at will" employee would have no action in the private sector for being wrongfully suspended from his employment. Edgar, supra, 595 P.2d at 540. Thus, the Edgar court, like this Court, read its tort claims act to require "a person asserting a claim against the state to show that the conduct complained of constitutes a tort which would be actionable if it were done by a private person in a private setting." Id. at 595 P.2d 539 (emphasis added). The Edgar court’s focus clearly was on the conduct, not the underlying function or rule of the government.
. Since Varig follows in time the passage of the Idaho Act, it is less persuasive as to the intent of the legislature in 1971. We must look to all of the pre-1971 federal case law (including but not exclusively the Dalehite decision). Nevertheless, Varig is consistent with the prior case law and with Sterling’s claim.
. It is interesting to note that the dissent argues that Indian Towing, supra, is the only Supreme Court case "out of step with all of the remainder of the United States Supreme Court decisions." Presumably then, Rayonier, supra, is a decision that is "in step" with that which the dissent argues. Thus, the dissent should be agreeing with that which Rayonier says on this matter— that the Federal Act has waived immunity for certain government conduct. Inexplicably, however, the dissent later criticizes the Rayonier decision on this point.
. The duty recognized by the majority opinion, as embodied in § 319, is, of course, not necessarily one owed to the world at large. The element of foreseeability remains. Megeff v. Doland, 123 Cal.App.3d 251, 176 Cal.Rptr. 467, 470 (1981).
. ‘ Does the dissent really mean to say that the legislature decided that Bloom should drive his uninsured motor vehicle for recreational purposes while totally drunk — all in violation of a court order?
. As counsel to this case will know, the issue of whether the probation officer was acting within the course and scope of his employment was never raised by either party, was never briefed by either party, and was never discussed by either party at oral argument. The dissent’s decision to raise it now does a disservice to the parties, and to the law of this case, because this Court should not have to authoritatively decide the issue until it has been fully briefed and argued.
. The proper resolution of the retroactivity or prospectivity of today’s opinion is better left for another case, with the issue properly raised and argued.