¶ 1. In this workers’ compensation action, defendant City of Burlington appeals from a judgment based on a jury verdict, finding that plaintiff Paul Crosby suffered a compensable psychological injury arising out of his employment as a firefighter with the City. The City raises two principal claims on appeal: (1) that Chapter II, § 70 of the Vermont Constitution precludes workers’ compensation benefits for psychological injuries unconnected to physical trauma; and (2) that the trial court’s jury instruction on the standard for determining whether such injuries resulted from unusual workplace stress requires reversal *241because it was inconsistent with the standard adopted by the Commissioner of the Department of Labor and Industry and the purpose of the workers’ compensation law. We agree with the second contention, and therefore reverse and remand for further proceedings consistent with the views expressed herein.
¶ 2. Plaintiff began working for the City as a firefighter in 1975 and was promoted to the position of lieutenant in 1989. In the summer of 1994, he stopped working and sought workers’ compensation benefits because he was experiencing stress at a level greater than he could handle. Plaintiff identified the stress causing his injury as anxiety resulting from the collapse of a building during a May 1994 three-alarm fire and his transfer in June 1994 from a shift he had been working since 1989. He alleged that his reaction to the foregoing incidents triggered repressed memories of a gruesome 1991 car fire and caused him to lose confidence in his superiors and his ability to do his job safely.
¶ 3. Between the summer of 1994 and the spring of 1995, plaintiff saw a number of physicians, including two psychiatrists, and a succession of therapists who generally agreed that plaintiff was not fit to return to duty. Several diagnosed his injury as post-traumatic-stress disorder. The City formally terminated plaintiff in March 1995, and later denied his claim for workers’ compensation benefits. The matter was brought before the Commissioner of the Department of Labor and Industry, who found in favor of the City, ruling that plaintiff was not entitled to benefits because he had failed to demonstrate that the stressful stimuli causing his injury were objectively real and unusual.
¶ 4. In so ruling, the Commissioner noted that the diagnoses provided by plaintiff’s medical experts, in contrast to that of the City’s expert, were based on plaintiff’s own subjective beliefs concerning the danger posed by the May 1994 fire, beliefs that were contradicted by other witnesses. In the Commissioner’s view, irrespective of whether plaintiff’s injury was brought on by a sudden stimulus or cumulative stress, plaintiff was required to demonstrate an objectively sound basis for his injury. The Commissioner concluded that he had failed to do so. The Commissioner also concluded that the appropriate control group to consider in determining whether plaintiff had been subjected to unusual stress was firefighters in general rather than all workplace employees. The Commissioner determined that plaintiff was not entitled to workers’ compensation benefits because the evidence demonstrated that the stress he was experiencing stemmed from normal workplace pressures related to fighting fires, being transferred, and engaging in conflicts with his superiors.
*242¶ 5. Plaintiff appealed the decision and sought a de novo jury trial in the superior court. See 21 V.S.A. § 670. Following a two-day trial, the court instructed the jury to determine: (1) whether plaintiff had suffered a psychological injury; (2) if so, whether the injury was caused by factors arising from his employment; and (3) if so, whether the injury resulted from stress that was significantly greater than that experienced by the general population of employees. The jury answered each of the three questions in the affirmative, and the trial court granted judgment in favor of plaintiff. This appeal followed.
I.
¶ 6. The City first contends that the Vermont Constitution bars plaintiffs recovery because it prohibits workers’ compensation benefits for psychological injuries unconnected with physical trauma. We disagree.
¶ 7. In relevant part, Chapter II, § 70 of the Vermont Constitution provides that the “General Assembly may pass laws compelling compensation for injuries received by employees in the course of their employment resulting in death or bodily hurt.” The City argues that the plain meaning of the phrase “bodily hurt,” particularly when considered in its historical context, is that the Legislature may authorize workers’ compensation benefits only for injuries having a physical component. In support of this position, the City notes that at the time § 70 was added to the Vermont Constitution in 1913, the Vermont House of Representatives reported that workers’ compensation would be allowed only for “violence to the physical structure of the body,” Journal of the House of the State of Vermont, Biennial Session, at 1034 (February 20,1913), and compensation for mental injuries unconnected to physical trauma was virtually unknown in Vermont negligence law. See Nichols v. Central Vt. Ry., 94 Vt. 14, 18, 109 A. 905, 907 (1919) (citing contemporary authority for the doctrine that, in absence of statute, no recovery for mental suffering without attendant physical injury is available in ordinary actions for negligence).
¶ 8. We find the constitutional argument unpersuasive. The phrase on which the City .relies — “violence to the physical structure of the body”— is taken from an amendment to a House bill that failed to pass the Senate. See Journal of the House of the State of Vermont, Biennial Session, at 1033-34 (Feb. 20,1913); Journal of the Senate of the State of Vermont, Biennial Session, at 972-73 (Feb. 21,1913). Even if we assumed that the language in question barred awards for psychological injuries, but cf. Bailey v. Am. Gen. Ins. Co., 279 S.W.2d 315, 318-19 (Tex. 1955) (court construed phrase “physical structure of the body” to include any harm to *243person), the provisions of a defeated statute can hardly be relied upon to support the interpretation of a completely separate constitutional amendment.
¶ 9. As for the actual text of § 70, “bodily hurt,” we agree with the City that we must consider the language in historical context, but we arrive at a different conclusion from that urged by the City. Section 70 was added to the Vermont Constitution in response to concerns that the employer liability bills being considered at the time were susceptible to constitutional attack. W. Flint, The Progressive Movement in Vermont, “Labor Obtains a Workmen’s Compensation Act,” at 86-87 (Am. Council on Public Affairs, Washington, D.C., 1941). These concerns were fueled by a 1911 decision of the New York Court of Appeals striking down New York’s fledgling workers’ compensation law. Id.; see Ives v. South Buffalo Ry., 94 N.E. 431 (N.Y. 1911). It is thus clear that the purpose of § 70 was to insulate pending workers’ compensation laws from constitutional attack, not to prevent workers from obtaining benefits based on psychological injuries.
¶ 10. The City does not argue that the Legislature intended the term “personal injury” in 21 V.S.A. § 618 to require physical injury or to exclude psychological injury unconnected to physical trauma. Rather, the City would have us hold that such claims are prohibited based on a 1913 constitutional amendment that was added to the Vermont Constitution to assure the survival of the workers’ compensation statute, not to restrict its reach. Absent any more persuasive evidence, we decline to so hold.
II.
¶ll.TheCitynextcontendsthetrialcourterredbyinstructingthe jurors that they should consider the “general population of employees” in determining whether plaintiff was subjected to unusual work-related stress. To understand the claim, a brief review of the legal background is instructive. Our workers’ compensation statute requires employers to compensate “a worker [who] receives a personal injury by accident arising out of and in the course of employment.” 21 V.S.A. § 618(a)(1). In analyzing whether an injury qualifies under workers’ compensation law as an accidental personal injury arising out of and in the course of employment, courts and commentators have divided claims into four general categories: (1) physical injury caused by physical stimulus; (2) physical injury caused by mental stimulus; (3) nervous injury caused by physical stimulus; and (4) nervous injury caused by mental stimulus. See 3 A. Larson, Larson’s Workers’ Compensation Law § 56.01, at 3 (2000). *244This case indisputably falls within the latter group, often described as “mental-mental” claims.
¶ 12. At least twenty-nine states provide workers’ compensation coverage for mental-mental claims, and fifteen do not. Id. § 56.06[3]-[4], at 52. Of the states that accept mental-mental claims, some require no more of a showing than that required of claims for physical injuries, others require a showing that the psychological injury resulted from a sudden stimulus, and still others require a showing that the stress was unusual when compared with one or another control group. Id. § 56.06[2]-[7], at 51-53. Some state legislatures, in the face of court decisions establishing liberal standards for mental-mental claims, have amended their workers’ compensation laws to limit such claims through a variety of means, such as requiring a set amount or type of stress, raising the standard of causation, increasing the burden of proof, imposing specific diagnostic guidelines, limiting the amount of benefits, or even excluding benefits altogether. Id. § 56.06[l][a]-[b], at 47-51.
¶ 13. The only case in which this Court has dealt with a mental-mental claim is Bedini v. Frost, 165 Vt. 167, 678 A.2d 893 (1996). There the claimant was a medical receptionist who had worked at her job for less than a year before leaving because of job-related stress. The Commissioner denied her claim for workers’ compensation benefits, finding that she had not been subjected to unusual working conditions, and we affirmed that decision. In so holding, we acknowledged that § 618 does not expressly differentiate between physical and mental injuries, but nonetheless deferred to the Commissioner’s decision to adopt an unusual-stress standard for mental-mental claims based on “reasonable policy concerns.” 165 Vt. at 169, 678 A.2d at 894. These included the Commissioner’s finding that medical authorities often disagree on “the precise etiology of most mental disorders,” that many sources outside of the employment setting — including culture, heredity, social environment, and family relationships — may cause or contribute to psychological injuries, and that medical opinions relating to the cause of such injuries are often based on the claimant’s subjective viewpoint. Id. An unusual-stress standard permits “a more objective inquiry” into the cause of the mental injury, id., protects against fraudulent claims, and “prevents the conversion of workers’ compensation into general health insurance.” Id. at 170,678 A.2d at 894.
¶ 14. Although the Commissioner in Bedini had expressed the unusual-stress standard in terms requiring that the applicant experience “a significantly greater dimension [of stress] than the daily stresses encountered by all employees,” id. at 169,678 A.2d at 894, she did not delineate *245with care the control group of “all employees” whose daily stresses were to be compared with the applicant’s, nor did we address the issue. As Justice Johnson explained in her dissent, however, unusual workplace stress has been measured in at least three different ways. Id. at 173,678 A.2d at 896 (Johnson, J., dissenting). One approach requires claimants to show that they were subjected to unusual pressures compared to other employees in the same workplace with similar responsibilities; another approach measures the pressures experienced by a claimant against those encountered by all employees doing the same job, sometimes referred to as the Wyoming approach; and a third approach requires a showing that a claimant experience pressures of a significantly greater dimension than those generally encountered by all employees in a working environment, often referred to as the Wisconsin approach. Id.; see 2 A. Larson, supra, § 44.05[4][d], at 52-56 (in determining whether claimant was subjected to unusual stress, courts may compare stress encountered by claimant with similar employees’ normal strains, strains of employment life in general, or strains of everyday nonemployment life).2
¶ 15. The second approach — measuring the claimant’s stress against that of all other workers performing the same job — is the standard advocated by the Commissioner in this case, and, while the Department’s decisions have not been a model of consistency, it is the approach the *246Commissioner has generally applied to determine whether an applicant’s work-related pressures meet the unusual-stress standard. For example, in Cross v. Vermont Dep’t of Pub. Safety, Op. No. 27-94WC, at 6 (Aug. 1, 1994), decided prior to Bedini, the Commissioner rejected the applicant’s claim for benefits resulting from stress-related mental injuries, noting that the applicant had “failed to establish that in relation to other similarly situated employees the burdens upon her were greater than they were upon them.” In Bedini, as noted, the Commissioner’s denial of benefits was based in part on a factual finding that other workers in the claimant’s office were subjected to similar stresses.
¶ 16. The Commissioner’s decisions since Bedini have not always clarified whether the control group of similarly situated employees includes all workers in similar jobs regardless of employer, or all workers employed in similar jobs by the same employer, but none appears to have applied a “working world at large” standard. In Estate of Fatovich v. Burlington Free Press, Op. No. 19-97WC, at 8 (July 29, 1997), for example, the Commissioner rejected the applicant’s claim for injuries from work-related stress, observing that the employee’s psychological distress was not caused by “stressfiil work events which were greater than the stress experienced by similarly situated employees.” Similarly, in Bell v. EHV Weidman, Op. No. 03-01WC, at 11 (Feb. 5, 2001), the Commissioner — citing Bedini — observed that an applicant claiming mental injury from workplace stress must demonstrate “that the stress is of significantly greater dimension than the daily stresses encountered by similarly situated employees.” Again, in DuBuque v. Grand Union Co., Op. No. 34-02WC, at 11 (Aug. 20, 2002), the Commissioner found that a claimant seeking benefits for mental injury resulting from work-related stress had failed to prove that “the stress is of significantly greater dimension than the daily stresses encountered by similarly situated employees.” And in a case involving a firefighter claiming mental injury from a stressful work environment, the Commissioner specifically found that the claimant had not demonstrated work-related stresses “that were of a significantly greater dimension than the daily stresses encountered by other firefighters.” Gallipo v. City of Rutland, Op. No. 22-00WC, at 7 (July 12, 2000).
¶ 17. Thus, in this — as in most recent decisions — the Commissioner has applied the unusual-stress standard to require a comparison of claimant’s stress to that of other similarly situated employees performing the same or similar work. The Commissioner’s approach is hardly unique. Although new to this Court, the question of the proper control group to be used for purposes of assessing whether an applicant’s work-related *247stress is unusual has been extensively explored by other courts and commentators. Justice Johnson cogently summarized the three basic approaches in her Bedini dissent; one requires a showing that the claimant suffered greater than normal stress as compared to all workers in general; a second compares the claimant’s stress to other workers in the same workplace with similar responsibilities; and a third measures the claimant’s stress as against other similarly situated employees, regardless of employer. Bedini, 165 Vt. at 173, 678 A.2d at 896-97 (Johnson, J., dissenting); see generally 2 A. Larson, supra, § 44.05[4][d], at 52-53; Graves v. Utah Power & Light Co., 713 P.2d 187, 192 (Wyo. 1986) (discussing basic approaches for applying unusual-stress standard).3
¶ 18. Although a few courts have adopted the so-called Wisconsin standard that compares a claimant’s stress to the daily strains which all employees must experience, see, e.g., Townsend v. Me. Bureau of Pub. Safety, 404 A.2d 1014, 1019 (Me. 1979); School Dist. # 1, Vill. of Brown Deer v. Dep’t of Indus., Labor & Human Relations, 215 N.W.2d 373, 377-78 (Wis. 1974), others have specifically rejected it in favor of the Wyoming standard that looks to the “day-to-day mental stresses experienced by other workers employed in the same or similar jobs.” Graves, 713 P.2d at 193; accord Dunlavey v. Econ. Fire & Cas. Co., 526 N.W.2d 845, 858 (Iowa 1995) (employee must establish that mental injury was caused by workplace stress of greater magnitude “than the day-to-day mental stresses experienced by other workers employed in the same or similar jobs”); Williams v. DePaul Health Ctr., 996 S.W.2d 619, 628 (Mo. Ct. App. 1999) (“[w]e are persuaded that the proper comparison ... is to compare Employee’s work-related stress with the stress encountered by employees having similar positions, regardless of employer”); see also McClain v. Texaco, Inc., 780 S.W.2d 34, 37 (Ark. Ct. App. 1989) (holding that “ultimate test is whether the stress constitutes an abnormal working condition for that type of employment,” rather than claimant’s coworkers); Southwire Co. v. George, 470 S.E.2d 865, 870-71 (Ga. 1996) (Sears, J., specially concurring) (plurality in dictum urges adoption of Wyoming over Wisconsin standard for reasons set forth in Dunlavey, 526 N.W.2d at 857-58); Davis v. Workmen’s Comp. Appeal Bd., 751 A.2d 168, 177 (Pa. 2000) (for work conditions to be considered abnormal, “ ‘they must be considered in the context of the specific employment’ ”) (quoting Wilson v. Workmen’s Comp. Appeal Bd., 669 A.2d 338, 343 (Pa. 1996)).
*248¶ 19. Sound policy reasons support such an approach. As noted, Bedini itself was based on a recognition that purely psychological injuries may result from so many “diverse factors” that “a high degree of uncertainty exists in the diagnosis of cause.” 165 Vt. at 169, 678 A.2d at 894. It is precisely this uncertainty of origin that has persuaded many courts and commentators to conclude that the optimum control group for determining unusual stress is that of other similarly situated employees in the same or similar jobs. In contrast to the broad “all employees” standard, which the court in Graves aptly noted is “too amorphous to be practical,” 713 P.2d at 193, or the narrow class of workers with the same employer, which may be too limiting where the business has few employees, see Dunlavey, 526 N.W.2d at 857, a control group comprised of similarly situated workers in the same general field provides a relatively precise, fair, and empirically workable standard.
¶ 20. A control group comprised of the “working world” would offer little in the way of assuring the validity of claims. It is difficult to imagine that a claimant or employer under an “all employees” rubric would not be able to produce some witness from the workplace whose “work-related stress is either significantly less or significantly greater than the stress experienced by the claimant.” N. Riley, Mental-Mental Claims — Placing Limitations on Recovery Under Workers’ Compensation for Day-to-Day Frustration, 65 Mo. L. Rev. 1023, 1043 (2000). Such a moveable standard could defeat the very purpose of the unusual-stress requirement, which is to ensure compensation for deserving claimants while simultaneously protecting against “fraudulent claims and prevent[ing] the conversion of workers’ compensation into general health insurance.” Bedini, 165 Vt. at 170, 678 A.2d at 894. As the court in Dunlavey explained: “By comparing the stresses endured by similarly situated employees, the Wyoming standard provides the employees with compensation for legitimate work related injuries while at the same time limits the employers’ liability to injuries caused by its industry.” 526 N.W.2d at 857.
¶ 21. The similarly-situated standard also offers the practical advantage of allowing both parties to focus on producing evidence of actual employment conditions in a specific field, “rather than trying to take into account the level of stress placed on the work force as a whole.” Riley, supra, 65 Mo. L. Rev. at 1043. This has the added benefit of promoting consistency among similar cases; an all-employees test, in contrast, could result in dissimilar outcomes depending upon the stress levels in the at-large employment context on which the parties choose to focus. See G. Dawes, Eligibilityfor Workers’Compensation^ Cases ofNontraumatic *249Mental Injury: The Development of the Unusual Stress Test in Wisconsin, 1987 Wis. L. Rev. 363,372 (different results could occur where “each court focuses on the stresses present in a different context of daily life”).
¶ 22. These various advantages have led commentators to conclude that the similarly-situated standard provides a “realistic and balanced test of legal causation,” M. Duckworth & T. Eick, Recent Developments in Mental/Mental Cases Under the Iowa Workers ’ Compensation Law, 45 Drake L. Rev. 809, 837 (1997), that is “superior to the other comparison standards.” Riley, supra, 65 Mo. L. Rev. at 1043. The result, to be sure, may be that some employees in high-stress jobs will fail to establish claims where other workers, confronted with similar strains, might succeed. By its nature, however, the unusual-stress test “is necessarily underinelusive.” Dawes, supra, 1987 Wis. L. Rev. at 373. Bedini established that not every employee who suffers mental inj ury from workplace stress will recover; to establish the requisite causal nexus between psychological injury and workplace stress a claimant must meet a heightened standard of proof premised upon a showing of unusual stress. 165 Vt. at 169-70, 678 A.2d at 894. By focusing on the conditions and pressures endemic in the claimant’s specific field of employment, the Commissioner does not undermine the purpose of the workers’ compensation law — as plaintiff asserts — but rather serves that purpose by ensuring that only genuine claims for workplace stress are compensated.
¶ 23. Although the similarly-situated standard thus enjoys the support of the Commissioner, commentators in the field, and other states, plaintiff asserts that it is predicated on the discredited doctrine of “assumption of the risk.” Plaintiff is mistaken. It is fair to measure a claimant’s stress by the conditions normal to his or her field or profession not because the claimant implicitly “assumes” the risks of employment, but rather because it is reasonable to assume that the claimant is prepared to deal with the normal strains of his or her occupation through training, temperament, and experience. Thus, the standard does not prejudice workers in high-stress fields, or benefit workers in low-stress occupations, but “allows for a uniform application of a legal standard across the wide spectrum of all jobs.” Duckworth & Eick, supra, 45 Drake L. Rev. at 837.
¶ 24. While other approaches are certainly possible, the Commissioner’s reliance on a control group comprised of “similarly situated” employees is well supported by established authority and sound policy. Therefore, under our traditionally deferential standard of review the Commissioner’s approach is entitled to control. See In re Duncan, 155 Vt. 402, 408, 584 A.2d 1140, 1144 (1990) (absent compelling indication of *250error, interpretation of statute by administrative body responsible for its execution will be sustained on appeal). We hold, therefore, that the trial court erred in instructing the jury to determine whether plaintiffs stress was unusual as compared with the general population of employees, rather than with all other employees performing similar work. Accordingly, the judgment must be reversed.
¶ 25. The parties raise several additional claims that require little discussion. The City urges an unusual-stress rule that would bar any claim for injuries sustained in the performance of duties within the claimant’s job description. The City cites no persuasive authority for such a rule, which we reject as inflexible and unnecessary to serve the purposes of the unusual-stress standard. The City also contends the “working world” standard contained in the trial court instruction violates the Common Benefits Clause, Vt. Const., ch. I, art. 7, by treating dissimilar employees the same. Our holding renders this argument moot. Finally, plaintiff asserts on cross-appeal that regardless of the applicable control group, there was no need to give an unusual-stress instruction because his injuries were the result of a “sudden stimulus.” See 3 Larson, supra, § 56.04[2]-[7] (discussing distinction some courts have drawn between injuries caused by sudden stimulus and those caused by cumulative stress). We need not address the question, however, as the record shows that plaintiffs psychological injury claim had several sources, including personnel problems unrelated to the 1991 and 1994 fires, and we therefore find no error in the instruction.
The judgment is reversed, and the matter is remanded for further proceedings consistent with the views expressed herein.
Although plaintiff here claims that Bedini adopted the “all-employees-in-the-workforce” standard, our opinion was concerned solely with whether the workers’ compensation statute supported the Commissioner’s decision to “differentiate between physical and mental injuries” by adopting an unusual-stress standard. 165 Vt. at 170,678 A.2d at 894. To assert that Bedini resolved the complex control-group issue without comment or discussion strains credulity. If anything, however, the Commissioner’s broad reference in Bedini to “all employees” masked a decision to utilize a much narrower control group composed of similarly situated employees of the same employer. See Graves v. Utah Power & Light Co., 713 P.2d 187, 192 (Wyo. 1986) (the ‘“all employees’ standard could be based upon three different groups,” a worker’s “fellow employees” doing the same or similar job, workers generally in the same job, or the “working world at large”). As Justice Johnson observed in her dissent, while the Commissioner had “articulated” a standard based on “all employees,” the approach that came “closest to what the Commissioner ha[d] actually done” utilized a control group of “other employees in the same workplace with similar responsibilities.” Bedini, 165 Vt. at 173, 678 A.2d at 896 (Johnson, J., dissenting). Thus, by adopting the Commissioner’s approach, it is at least arguable that Bedini endorsed the most narrow control group of the three. This is the conclusion, in fact, of the preeminent workers’ compensation authority, who has categorized Bedini as a case requiring the claimant to “show that his or her stresses at work were significantly greater than the stress levels affecting co-employees.” 3 A. Larson, Larson’s Workers’ CompensationLaw § 56.06D[6], at 143 (2000). It is clear to us, however, that Bedini did not actually address or decide the control-group issue.
Still another approach is to compare the claimant’s stress with the strains of everyday non-employment life in general. See 2 A. Larson, supra, § 44.05[4][d], at 53.