TAYLOR, C. J.
In this case involving the Worker’s Disability Compensation Act (WDCA), MCL 418.101 et seq., [499]*499the first issue is whether the phrase “the proximate cause” in MCL 418.375(2) means the sole proximate cause, i.e., “the one most immediate, efficient, and direct cause of the injury or damage.” We conclude that it does, as we did in construing the identical phrase in the governmental tort liability act (GTLA), MCL 691.140 et seq., in Robinson v Detroit, 462 Mich 439, 462; 613 NW2d 307 (2000). We therefore overrule Hagerman v Gencorp Automotive, 457 Mich 720; 579 NW2d 347 (1998), which incorrectly construed the phrase to mean “a proximate cause” that is a substantial factor in causing the event. Accordingly, we vacate the decision of the Workers’ Compensation Appellate Commission (WCAC) and remand this case to the WCAC for reconsideration. The second issue is when, in the circumstance of a parent-employee’s death, a child of that person is entitled to a presumption of whole dependency. We conclude that a child is only entitled to the presumption if he or she was under the age of 16 at the time of the parent-employee’s death. Because the WCAC erred in holding to the contrary, on remand, the WCAC must make the necessary factual determinations to apply this holding.
I. FACTS AND PROCEEDINGS BELOW
Randall G. Paige worked as a firefighter for the city of Sterling Heights (hereafter defendant). On October 12, 1991, Paige was sent to the scene of a severe automobile accident. After extracting a three-year-old girl from an automobile and car lying her to an ambulance, Paige began experiencing an ache in his right arm. Approximately 30 minutes later, after he had returned to the fire station, Paige was completing a report of the automobile accident when he again experienced pain in his right arm. This time, the pain in his [500]*500arm was accompanied by chest pains and profuse sweating. Paige was transported to a hospital, where he was diagnosed as having suffered a myocardial infarction. He did not return to work after this incident. In 1993, he was granted an open award of workers’ compensation benefits by magistrate Donald Miller.1
Paige suffered a second myocardial infarction on August 15, 2000. He was diagnosed as having coronary artery disease, and underwent a quadruple coronary artery bypass on August 21, 2000. On January 4, 2001, Paige died in his sleep. An autopsy report prepared by the Oakland County Medical Examiner’s office noted that Paige suffered from occlusions of the left anterior descending coronary artery, right coronary artery, and four coronary bypass grafts. The deputy forensic pathologist who conducted the autopsy opined that Paige “died of arteriosclerotic
Paige’s son, Adam Paige, who was eight years old when Paige suffered his first heart attack and 17 when Paige died, filed a claim for workers’ compensation [501]*501death dependency benefits pursuant to MCL 418.375(2).3 Under this statute, the child of a deceased employee is entitled to death dependency benefits if he or she was dependent on the deceased employee and a work-related injury was the proximate cause of the parent-employee’s death. In making his claim for death dependency benefits, Adam claimed that as a minor he had been dependent on his father for support. Further, he claimed that the work-related heart attack in 1991 had contributed to his father’s death by weakening his heart and, therefore, constituted “the proximate cause” of his father’s death under Hagerman, which held that the phrase does not mean the sole proximate cause of death but, rather, requires only a cause that is a substantial factor in the employee’s death. Hagerman, supra at 728, 736. Defendant opposed the claim for death dependency benefits, arguing that Adam had not introduced evidence establishing that he was, in fact, dependent on his father. Moreover, defendant argued that Hagerman had been impliedly overruled by Robinson, which held that the phrase “the proximate [502]*502cause” means the sole proximate cause or, in other words, “the one most immediate, efficient, and direct cause of the injury or damage. Robinson, supra at 462. Accordingly, defendant asserted that under the Robinson definition Randall Paige’s work-related 1991 heart attack was not “the proximate cause” of his death.
Magistrate Andrew Sloss resolved both issues in Adam’s favor. First, he determined that the Hagerman definition of “the proximate cause” applied and, therefore, the work-related heart attack that Randall Paige suffered in 1991 did not have to be the sole or most immediate cause of his death but, rather, only needed to be a substantial factor in the events leading to his death. He determined that the 1991 heart attack was a substantial factor in Paige’s death, stating that all three doctors who testified at the hearing on Adam’s claim “agreed that it was a combination of underlying coronary artery disease together with the cumulative damage to the heart that began with his work-related myocardial infarction in 1991” that caused Randall Paige’s death in 2001. The magistrate concluded by determining that Adam was entitled to death dependency benefits as long as he qualified as a dependent. Noting that Adam’s status as a dependent is to be determined as of the date of his father’s 1991 work-related injury, MCL 418.341,4 Magistrate Sloss recognized that Magistrate Miller had listed Adam as Randall Paige’s dependent in his 1993 order granting [503]*503Randall Paige an open award of benefits. He held that this determination of dependency was controlling, and granted Adam’s request for death dependency benefits.
Defendant appealed Magistrate Sloss’s ruling to the WCAC. Again, defendant argued that the magistrate should have applied the Robinson definition of “the proximate cause.” The WCAC, however, rejected defendant’s argument, concluding that Hagerman was controlling because it specifically addressed MCL 418.375(2) while Robinson, on the other hand, involved a provision of the GTLA. Defendant also again challenged Adam’s status as a dependent. Although it did not directly challenge Magistrate Sloss’s reliance on Magistrate Miller’s determination that Adam was, in fact, dependent on his father at the time of the 1991 work-related injury, defendant argued that Magistrate Sloss had erred by failing to address the extent of Adam’s dependency. Specifically, defendant asserted that under this Court’s decision in Runnion v Speidel, 270 Mich 18; 257 NW 926 (1934), the magistrate was required to make a factual determination regarding whether Adam was wholly or partially dependent on his father at the time of the 1991 work-related injury and, because Magistrate Sloss did not do so, and no evidence of whole or partial dependency existed in the record, the correct weekly compensation amount could not be calculated. The WCAC, however, rejected defendant’s assertion that Runnion required such a factual determination of dependency and, instead, relied on Murphy v Ameritech, 221 Mich App 591; 561 NW2d 875 (1997), for the proposition that Adam was entitled to the conclusive presumption set forth in MCL 418.331(b) that he was wholly dependent because he had been under the age of 16 at the time of his father’s work-related heart attack in 1991.
Defendant applied for leave to appeal the WCAC’s ruling in the Court of Appeals, again raising the proxi[504]*504mate causation and dependency issues. The Court of Appeals, however, denied defendant’s application for leave to appeal for lack of merit in the grounds presented. Unpublished order of the Court of Appeals, entered January 10, 2005 (Docket No. 256451). Defendant then applied for leave to appeal in this Court. We scheduled oral argument on whether to grant defendant’s application or take other peremptory action permitted by MCR 7.302(G)(1), and directed the parties to address whether Robinson overruled Hagerman, and whether the WCAC erred by failing to follow Runnion and make a factual determination of the extent of Adam’s dependency on his father at the time of his father’s injury. 474 Mich 862 (2005).
II. STANDARD OF REVIEW
Resolution of the issues in this case involves the interpretation of provisions of the WDCA. Statutory interpretation is a question of law that we review de novo. Reed v Yackell, 473 Mich 520, 528; 703 NW2d 1 (2005). As we stated in Reed, supra at 528-529:
Our fundamental obligation when interpreting statutes is “to ascertain the legislative intent that may reasonably be inferred from the words expressed in the statute.” Koontz v Ameritech Services, Inc, 466 Mich 304, 312; 645 NW2d 34 (2002). If the statute is unambiguous, judicial construction is neither required nor permitted. In other words, “[b]ecause the proper role of the judiciary is to interpret and not write the law, courts simply lack authority to venture beyond the unambiguous text of a statute.” Id.
III. ANALYSIS
If an employee who suffered an injury arising out of and in the course of employment dies before the period within which the employee is entitled to weekly work[505]*505ers’ compensation benefits ends, the employee’s death is considered to have ended the disability and relieves the employer of liability for further weekly benefits to the injured employee. MCL 418.375(1). However, under MCL 418.375(2), in lieu of such weekly payments to the employee, the employer is required to pay death benefits pursuant to MCL 418.3215 if two requirements are met: (1) the work-related injury was “the proximate cause” of the employee’s death, and (2) the deceased employee leaves dependents who were wholly or partially dependent upon the employee for support.6
A. THE PROXIMATE CAUSE
Primarily at issue in this case is the first requirement of MCL 418.375(2) that the work-related injury be “the proximate cause” of the employee’s death. In Hagerman, a majority of this Court relied on Dedes v Asch,7 [506]*506which involved MCL 691.1407(2) (c) of the GTLA, for the proposition that the Legislature’s use of the definite article “the” instead of the indefinite article “a” is inconsequential.8 Under its interpretation of common-law principles of proximate causation, the Hagerman majority rejected the idea that by using the phrase “the proximate cause,” the Legislature meant that the work-related injury had to be the sole proximate cause of the employee’s death in order for the employer to be liable for death benefits under MCL 418.375(2).9 Instead, the majority held that the employer was liable for death benefits even if there was more than one proximate cause of the employee’s death, as long as the work-related injury was a “substantial factor” in the employee’s death.10
In a dissent joined by Justices WEAVER and BRICKLEY, I argued that the Legislature’s use of the phrase “the proximate cause” in MCL 418.375(2) unambiguously indicated its intent that the work-related injury must be the sole proximate cause of the employee’s death in order for the employer to be liable for death benefits. My primary reasons for this conclusion were twofold. First, the term “proximate cause” had a longstanding definition in Michigan’s jurisprudence before the enactment of the WDCA.11 Second, the majority’s analysis had improperly rewritten the statute by failing to recognize the Legislature’s use of the word “the.”12
[507]*507Two years after Hagerman, in Robinson, supra, which involved MCL 691.1407(2) (c) of the GTLA, this Court overruled the part of Dedes, supra, on which the Hagerman majority had based its interpretation of MCL 418.375(2) of the WDCA, and held that the phrase “the proximate cause” as used in MCL 691.1407(2)(c) of the GTLA refers to the sole proximate cause, i.e., “the one most immediate, efficient, and direct cause preceding an injury.”13 The heart of the Robinson majority’s rationale, which relied in part on my dissent in Hagerman that the phrase “the proximate cause” is not synonymous with the phrase “a proximate cause,” was as follows, Robinson, supra at 460-462:
[T]he Legislature has shown an awareness that it actually knows that the two phrases are different. It has done this by utilizing the phrase “a proximate cause” in at least five statutes16 and has used the phrase “the proximate cause” in at least thirteen other statutes.17 Given such a pattern, it is particularly indefensible that the Dedes majority felt free to read “the proximate cause” as if it said “a proximate cause.” The error will not be compounded, as today this Court corrects the flawed analysis of the Dedes majority.
Nevertheless, the fact that the Legislature sometimes uses “a proximate cause” and at other times uses “the proximate cause” does not, of course, answer the question what “the proximate cause” means other than to show that the two phrases should not be interpreted the same way. Our duty is to give meaning to the Legislature’s choice of one word over the other.
We agree with the following analysis found in the dissent in Hagerman v Gencorp Automotive, 457 Mich 720, 753-754; 579 NW2d 347 (1998):
“Traditionally in our law, to say nothing of our classrooms, we have recognized the difference between ‘the’ and [508]*508‘a.’ ‘The’ is defined as ‘definite article. 1. (used, esp. before a noun, with a specifying or particularizing effect, as opposed to the indefinite or generalizing force of the indefinite article a or an). ...’ Random House Webster’s College Dictionary, p 1382. Further, we must follow these distinctions between ‘a’ and ‘the’ as the Legislature has directed that ‘all words and phrases shall be construed and understood according to the common and approved usage of the language ....[’] MCL 8.3a; MSA 2.212(1). Moreover, there is no indication that the words ‘the’ and ‘a’ in common usage meant something different at the time this statute was enacted ....”
Further, recognizing that “the” is a definite article, and “cause” is a singular noun, it is clear that the phrase “the proximate cause” contemplates one cause. Yet, meaning must also be given to the adjective “proximate” when juxtaposed between “the” and “cause” as it is here. We are helped by the fact that this Court long ago defined “the proximate cause” as “the immediate efficient, direct cause preceding the injury.” Stoll v Laubengayer, 174 Mich 701, 706; 140 NW 532 (1913). The Legislature has nowhere abrogated this, and thus we conclude that in MCL 691.1407(2)(c) the Legislature provided tort immunity for employees of governmental agencies unless the employee’s conduct amounts to gross negligence that is the one most immediate, efficient, and direct cause of the injury or damage, i.e., the proximate cause.
[509]*509Despite the fact that MCL 418.375(2) of the WDCA, at issue in this case, and MCL 691.1407(2) of the GTLA, which was at issue in Robinson, both use the phrase “the proximate cause,” Adam argues that the definition of “the proximate cause” from Robinson should not be applied to MCL 418.375(2). Adam’s primary argument in support of this assertion is that the GTLA, as a statute in derogation of the common law, is generally said to be strictly construed in favor of governmental immunity,14 while the WDCA, being a remedial statute, is generally said to be liberally construed to grant, rather than deny, benefits.15 Although we have stated and utilized these preferential rules of construction in the past, their application is unnecessary in this case because the proper definition of the phrase “the proximate cause” can be ascertained solely by reference to the common meaning of the term “the” and the peculiar meaning that the phrase “proximate cause” has acquired in the law. These preferential rules of construction do not nullify the general rule that statutes should be reasonably interpreted consistent with their plain and unambiguous meaning. See Northern Concrete Pipe, Inc v Sinacola Cos-Midwest, Inc, 461 Mich 316, 320-321; 603 NW2d 257 (1999). More importantly, they do not override the Legislature’s clear directive in MCL 8.3a that common words, such as “the,” are to be construed according to their common meaning and that words that have acquired a peculiar and appropriate [510]*510meaning in the law, such as “proximate cause,” are to be accorded such peculiar and appropriate meaning.
Accordingly, we overrule Hagerman and hold that the phrase “the proximate cause,” as used in MCL 418.375(2) of the WDCA, refers to the sole proximate cause. In deciding to overrule Hagerman, we have not only considered the fact that it was wrongly decided but also whether less injury will result from overruling it than from following it.16 In making this determination we have considered whether Hagerman defies “practical workability,” whether reliance interests would work an undue hardship, and whether changes in the law and facts no longer justify the Hagerman decision.17
Hagerman defies practical workability because a person reading the statute surely would not know that he or she cannot rely on what the statute plainly says. That is, a reader and follower of the statute would, because of Hagerman’s rewrite, not be behaving in accord with the law. Such a regime is unworkable in a rational polity. This all gets back to the unrebutted truth that “it is to the words of the statute itself that a citizen first looks for guidance in directing his actions.”18 Furthermore, Hagerman is not only inconsistent with the plain language of the statute, it is also inconsistent with this Court’s decision in Robinson. How are the people of this state to know what “the proximate cause” means when there is one case from this Court that states that it means one thing and another case that states that it means something else? When identical words in the law, lying within a similar statutory context, mean something altogether different, [511]*511we do believe that there is a “practical workability” problem, not in the sense that a court of law cannot render some decision — no opinion of this Court is “unworkable” in that sense — but in the sense that the law is made a mockery, meaning one thing in one paragraph and something else in the next. The law is thereby made less workable in the sense that it is made more confusing and less decipherable to the ordinary citizen. As we noted this very term in Joliet v Pitoniak, 475 Mich 30, 40; 715 NW2d 60 (2006), when two decisions from this Court contain conflicting analysis, this Court is “obligated to resolve this conflict and decide which decision best reflects the legislative intent expressed in the words of the statute ....” This is true even where, as here, the conflicting decisions address the same or similar language, but not the same statutes.19
Regarding reliance interests, Hagerman, having been decided just eight years ago, has not become “so embedded, so accepted, so fundamental, to everyone’s expectations that to change it would produce not just readjustments, but practical real-world dislocations.”20 Such reliance is only present where the prior decision has caused a large number of persons to attempt to conform their conduct to a certain norm. For example, where an entire class of individuals or businesses purchase insurance and another entire class does not in reliance on a decision by this Court, this may be viewed as the sort of reliance that could cause “practical [512]*512red-world dislocations.” Cf. Pohutski v City of Allen Park, 465 Mich 675; 641 NW2d 219 (2002). There is a significant distinction between merely complying with precedent and affirmatively altering one’s behavior in reliance on precedent. Where there is mere compliance with precedent, the overruling of that precedent will not cause “practical real-world dislocations,” but where a great number of people affirmatively alter their behavior in reliance on precedent, the overruling of a precedent may cause “practical real-world dislocations.”21 This Court’s decision in Hagerman cannot be said to have caused a great number of persons to affirmatively alter their conduct in any way, except in [513]*513the sense that any law requires general compliance with its terms. It cannot be seriously argued that Randall Paige positioned himself in reliance on Hagerman. He, as indeed any injured employee we might see, did not script his unfortunate injuries and illnesses with reference to Hagerman or any other case of this Court. Nor did his lawyers proceed any differently because of Hagerman. Furthermore, for most of the duration of this litigation Hagerman’s status was precarious, and known to be such, because Robinson, which made Hagerman untenable, was decided only two years after Hagerman.
Finally, we need not consider whether changes in the law and facts no longer justify Hagerman because Hagerman itself was never justified as it was a change in the law that this Court had the power, but not the authority, to make. It was not justified from its inception.
Thus, with Hagerman no longer controlling, we return to the language of the statute. It is the case that in order for an employer to be hable for death benefits under MCL 418.375(2), the deceased employee’s work-related injury must have been “the one most immediate, efficient, and direct cause preceding [the death].”22 We therefore remand this case to the WCAC for a determination whether Randall Paige’s work-related injury was “the proximate cause” of his death under this standard.
B. RESPONSE TO JUSTICE CAVANAGH
The dissent of Justice CAVANAGH stridently criticizes the positions the majority has taken. His theme is that our positions are tedious in that we have argued them [514]*514in the past, as well as that we are irresponsible. It is true that we have argued them previously, but in the law consistency is not normally seen as a defect; if it is, the dissent’s arguments against our rather simple thesis, which holds that one who says “the proximate cause” has said something different than one who says “a proximate cause,” are equally shopworn. In attempting to provide buoyancy for his argument that we are irresponsible, Justice CAVANAGH restates the simply incorrect claim that we have overturned cases at an unprecedented rate. Yet, as we pointed out with statistics in Sington v Chrysler Corp, 467 Mich 144, 166-170; 648 NW2d 624 (2002), and Mack v Detroit, 467 Mich 186, 211; 649 NW2d 47 (2002), and as Victor E. Schwartz has also discussed in his article A critical look at the jurisprudence of the Michigan Supreme Court,23 we have not done that. Unwilling to rebut either the statistics or the Schwartz analysis, Justice CAVANAGH continues making the claim, hoping, one surmises, that readers will not know any better. We think they will.
With regard to Justice CAVANAGH’s claim that history’s judgment of us will be unkind, this also is not a new claim.24 We think the concern should be his. Our core argument is that texts should be approached using the same doctrines every time. This could be described as a “truth in reading” approach. His is the less easily defended notion that sometimes you read statutes using textual and grammatical rules that all users of the language normally employ, but on other entirely unpredictable occasions you do not. Accordingly, while Justice CAVANAGH in some cases does use the textual rules that [515]*515courts have traditionally employed,25 in others he jumps the textualist rails and employs interpretive approaches that disregard what the instrument actually says and instead rely on extratextual sources such as legislative testimony,26 the perceived intent of the Legislature,27 overarching policy considerations,28 or even what has been described as the theory of “legislative befuddlement,” which holds that the Legislature can, if we desire, be held to not know what it is doing and thus we need not do what it directs.29 It bears emphasizing that he has in the past provided no rationale regarding which technique he will use in any given case so that litigants, or even citizens attempting to structure their conduct to accord with the law, have no idea which Justice CAVANAGH, the traditionalist or the deconstructionist, will decide the case. In response to this assertion, he now argues that he only departs from the traditional approach when a statute is unclear or ambiguous, post at 540, yet even a casual review of the [516]*516cases cited herein reveals that this defense will not bear scrutiny and that in fact he will find a way, no matter how tendentious (see in particular Mayor of Lansing v Public Service Comm, 470 Mich 154; 680 NW2d 840 [2004]), to declare that which he wishes to be ambiguous or unclear to be exactly that. It is an approach of ambiguity by fiat.
Supplementing all these extratextual tools Justice CAVANAGH uses to reach a desired outcome is his utilization of the notion of legislative acquiescence, which he deploys when an effort is made to overrule a past case where the law was not followed. On such occasions, he argues, as he does in this case, that this Court should retain the previous interpretation of a statute that is clearly wrong simply because the Legislature has not amended the statute to correct our error.30 However, as this Court explained in Donajkowski v Alpena Power Co, 460 Mich 243, 261; 596 NW2d 574 (1999), the doctrine of legislative acquiescence is not recognized in this state for the sensible reason that “sound principles of statutory construction require that Michigan courts determine the Legislature’s intent from its words, not from its silence.” (Emphasis in original.)31 Not content to merely ignore Donajkowski, he advances a new argument for legislative acquiescence, which is the startling notion that once this Court “ ‘interprets a statute, then the statute becomes what this Court has [517]*517said it is’ ” and that it is “ ‘neither more nor less than an amendment,’ ”32 therefore making it impermissible for this Court to ever revisit its interpretation of the statute. This is an odd argument for Justice CAVANAGH to make, and undeniably inconsistent with his own practices, given that he has in other cases in the last several years supported this Court’s decisions to correct erroneous interpretations given to statutes in the past.33 Moreover, his authority for this audacious statement is an unenthusiastic reference to United States Supreme Court Justice Hugo Black’s lone dissenting statement in the 1970 case of Boys Markets, Inc v Retail Clerks Union, Local 770.34 This is an unconvincing authority to cite, as even he seems to acknowledge, because the majority did not share Justice Black’s view35 and, in that very case, overruled an earlier case [518]*518that had improperly construed the statute at issue.36 The consequential point, however, is that this dubious view of judicial power, even if it could be construed as defensible under the United States Constitution, is not defensible under the Michigan Constitution. Our Constitution strictly forbids a court from exercising legislative power by providing that “[n]o person exercising the powers of one branch [of government] shall exercise powers properly belonging to another branch... .”37 In short, we cannot “amend” statutes and Justice CAVANAGH’s view is directly at odds with our own Constitution.
With the claimed federal authorities exposed as no authority at all, we return to the fact that Justice CAVANAGH chooses to ignore the holding of this Court in Donajkowski, just as he has ignored this Court’s holdings rejecting his unprincipled approach to declaring statutes ambiguous.38 In doing so, Justice CAVANAGH reveals how little fidelity he has to precedent when he does not like the precedent. His argument on stare decisis then is, and should be seen as, entirely inconsis[519]*519tent. His test on when to leave the text and search for meaning elsewhere is really no more sophisticated than doing so when the desired outcome is one the text alone will not allow. This is, of course, an indefensible theory of jurisprudence even superficially. Further, it is dangerous because with it comes the death of predictability in the law. If institutionalized as a practice, our citizens could never tell in advance which judge, and thus what preferences, will control. If fully implemented in the law, our courts would be seen as only a scramble for jackpots. Much more can be said negatively of this “judicial supremacist” approach, and we have,39 but at root it gives to judges, not to the people through the Legislature, control of public policy.40 Our constitutions have never authorized such a usurpation,41 and the cultivation and seizure of such power, we believe, itself invites history’s reproach.
This response has also prompted Justice CAVANAGH to claim that we are attacking him personally and being insufficiently respectful of our predecessors on this Court. This is not only inaccurate but peculiar coming from a justice who himself has this term accused the majority of writing an opinion to advance the majority members’ interests,42 and has, in the past, accused the justices in the majority of making “unforgivable” fabri[520]*520cations,43 basing decisions on the view of what is “socially acceptable behavior,”44 and having a “complete lack of respect” for civil rights.45
All we are doing is pointing out the problems with his methodology of deciding cases. That is not a personal attack. His claim should be seen as the latest volley in a years-long effort by the remnants of the pre-1999 Court and its supporters to do what they can to bring back the less disciplined approach of that Court.
In that era, Justice CAVANAGH was much more influential because he had more colleagues who shared his approach. His influence has waned and with it the influence of those who benefit from the legal regime of which he was an unquestioned leader — a regime where the decisions were highly unpredictable, inconsistent, and virtually any claim was a possible winner. He and they are very unhappy with the changes and have not accommodated well to the current situation. The fact that we point out that Justice CAVANAGH has articulated no consistent legal principles or methodology for deciding cases is neither a personal attack nor an occasion for martyrdom. However, for Justice CAVANAGH, it is an inconvenient fact.
We close by returning to this case and what should not be lost sight of here. That is that in Justice CAVANAGH’s world it is perfectly normal, indeed correct, that sometimes absolutely identical phrases in our statutes, here “the proximate cause,” have different meanings in different statutes. To express the notion is to expose its flaw. To the extent that Justice CAVANAGH [521]*521continues to espouse it and its justifying nostrums, we will continue to do our best to write of their shortcomings and to expose how compromising to the development of a principled jurisprudence they are.
C. DEPENDENCY
If the work-related injury qualifies as “the proximate cause” of the employee’s death under the definition we have set forth above, the next inquiry under MCL 418.375(2) is whether the employee left dependents and, if so, whether they were “wholly or partially dependent on him or her for support. . . .” The answers to these questions are provided in MCL 418.341, which provides, in relevant part:
Questions as to who constitutes dependents and the extent of their dependency shall be determined as of the date of the injury to the employee, and their right to any death benefit shall become fixed as of such time, irrespective of any subsequent change in conditions except as otherwise specifically provided in sections 321, 331 and 335.
Accordingly, under this statute, the workers’ compensation magistrate must determine whether there were persons dependent on the deceased employee, and the extent of such dependency, by looking at the circumstances at the time of the work-related injury — not at the time of death. In the present case, Magistrate Miller listed Adam Paige as a dependent of Randall Paige when he issued his 1993 order granting Randall Paige an open award of benefits. Defendant did not appeal Magistrate Miller’s 1993 order. Therefore, the issue whether Adam was dependent on his father at the time of the work-related injury is res judicata,46 and defen[522]*522dant may not challenge it now. But, as defendant correctly argues, Magistrate Miller did not determine the extent of Adam’s dependency on his father at the time of the work-related injury, i.e., whether Adam was wholly or partially dependent upon Randall Paige. Without such a determination being made, the rate of any weekly death benefits to which Adam may be entitled cannot be calculated.
The WCAC rejected defendant’s argument and held that Adam is conclusively presumed to be wholly dependent under MCL 418.331, which provides, in pertinent part:
The following persons shall be conclusively presumed to be wholly dependent for support upon a deceased employee:
(b) A child under the age of 16 years ... upon the parent with whom he or she is living at the time of the death of that parent.... In all other cases questions of dependency, in whole or in part, shall be determined in accordance with the fact, as the fact may be at the time of the injury.
The WCAC’s conclusion that Adam, who was under the age of 16 at the time of the injury but over the age of 16 at the time of the death, is entitled to the conclusive presumption of whole dependency was erroneous. In Runnion, supra, we interpreted the predecessor of MCL 418.331(b), which was substantively similar,47 consistently with its plain terms, i.e., that the presumption of whole dependency applies only if the [523]*523child was under the age of 16 at the time of the employee’s death. If the child was, like Adam in this case, over the age of 16 at the time of the employee’s death, the fact that the child was under the age of 16 at the time of the injury does not entitle the child to the conclusive presumption of whole dependency. Instead, “[w]hether there was actual dependency, total or in part, at the time of the injury is a question of fact.”48
In the present case, the WCAC noted our decision in Runnion but essentially ignored it, relying instead on statements made by the Court of Appeals in Murphy, supra, to conclude that a child is entitled to the presumption as long as the child was under the age of 16 at the time of the work-related injury. There are two problems with the WCAC’s having disregarded Runnion and relied on Murphy. First, Runnion directly addressed the proper interpretation of MCL 418.331(b) with regard to the issue presented here, while Murphy involved an altogether different issue implicating MCL 418.335.49 Second, and more important, even if Murphy had directly addressed the statute and issue presented [524]*524in this case, the WCAC would not be justified in choosing to follow Murphy instead of Runnion. The obvious reason for this is the fundamental principle that only this Court has the authority to overrule one of its prior decisions. Until this Court does so, all lower courts and tribunals are bound by that prior decision and must follow it even if they believe that it was wrongly decided or has become obsolete. Boyd v W G Wade Shows, 443 Mich 515, 523; 505 NW2d 544 (1993). In short, the WCAC may not, as it has attempted to do here, presume to overrule this Court by disregarding Runnion and seeking to impose its own construction of MCL 418.331(b).
Accordingly, should the WCAC determine on remand that Randall Paige’s work-related injury was the proximate cause of his death, we direct it to further determine the extent of Adam Paige’s dependency on Randall Paige at the time Randall Paige suffered the work-related injury.
IV CONCLUSION
We hold that the definition of the phrase “the proximate cause” set forth in Robinson, supra, applies to MCL 418.375(2) of the WDCA. In so holding, we overrule Hagerman, supra. Accordingly, we vacate the decision of the WCAC and remand this case to the WCAC for a determination of whether Randall Paige’s work-related injury was “the proximate cause” of his death under the Robinson definition. Furthermore, the WCAC erred in determining that Adam Paige is entitled to a conclusive presumption of whole dependency under MCL 418.331(b). If, on remand, the WCAC determines that Randall Paige’s work-related injury was “the proximate cause” of his death, we direct the WCAC to determine the extent of Adam Paige’s dependency upon [525]*525Randall Paige at the time Randall Paige suffered the work-related injury in accordance with Runnion, supra.50
Corrigan, Young, and Markman, JJ., concurred with Taylor, C.J.
2 Arteriosclerosis is a hardening of the arteries. Stedman’s Online Medical Dictionary, <http://www.stedmans.com/section.cfm/45> (accessed April 14, 2006).