Badgley v. Walton
A. James WALTON, Jr. (Retired), Kerry L. Sleeper (Retired), Commissioners of Public Safety and Department of Public Safety.
Supreme Court of Vermont.
William H. Sorrell, Attorney General, and David R. Groff, Assistant Attorney General, Montpelier, for Defendant-Appellee Department of Public Safety.
Present: REIBER, C.J., DOOLEY, JOHNSON, SKOGLUND and BURGESS, JJ.
¶ 1. Plaintiffs George Badgley and Ruth Whitney appeal the Windham Superior Court's dismissal of their claim that the mandatory retirement of public safety officers violates the Common Benefits Clause 471*471 of Chapter I, Article 7 of the Vermont Constitution. We affirm.
¶ 2. Plaintiffs are former state troopers who were forced to retire, under 3 V.S.A. § 459(a)(2), when they reached the age of fifty-five. They brought an action for injunctive relief and damages against the Vermont Department of Public Safety and the Commissioner of Public Safety. Plaintiffs argue that the mandatory retirement of public safety officers violates the Common Benefits Clause of the Vermont Constitution. Vt. Const. ch. I, art. 7. Plaintiffs also argue that the statutory exception to the Vermont Fair Employment Practices Act for law enforcement officers, 21 V.S.A. § 495f, is unconstitutional under the same provision of the Vermont Constitution. Section 495f provides a specific exception to the general prohibition that employers may not discriminate on the basis of age, stating: "Mandatory retirement on account of age, necessitated under a police or firefighter retirement system, is specifically authorized."
¶ 3. With respect to plaintiffs' claims for damages, the trial court granted defendants' motion for judgment on the pleadings, finding that these claims were barred by sovereign immunity. The court then held a bench trial and concluded that the mandatory retirement age did not violate the Common Benefits Clause and entered judgment for defendants on plaintiffs' remaining claims. This appeal followed.
¶ 4. We review a trial court's legal conclusions de novo. Charbonneau v. Gorczyk, 2003 VT 105, ¶ 2, 176 Vt. 140, 838 A.2d 117. We accept a trial court's findings of fact unless the findings are clearly erroneous. Quenneville v. Buttolph, 2003 VT 82, ¶ 11, 175 Vt. 444, 833 A.2d 1263. Plaintiffs have not challenged the trial court's findings of fact.
¶ 5. The trial court made extensive findings with respect to the two plaintiffs who were retired under the mandatory retirement law. We need not repeat these findings except to say that both were in good physical shape with good cognitive skills on the date of their retirement. Both had successful, rewarding and productive careers in the state police. Both have done law enforcement work, but not as state police officers, since their retirement.
¶ 6. As we discuss below, the primary justification for the mandatory retirement policy is maintaining public safety. The trial court, accordingly, made findings in this area. The Department of Public Safety administers annual physical fitness tests to its public safety officers. The tests include sit-ups, push-ups, a "sit-and-reach" stretch test, a timed mile-and-a-half run, and a bench-press test. There is also a body-fat determination. Officers are required to meet established standards in each test for their respective age bracket and gender. The standards are set at the achievement level for the median person in 472*472 the general population for an officer's gender and age bracket. Officers who fail these tests are given the opportunity to train and retake the tests so that they can maintain their job.
¶ 7. Two expert witnesses testified at the bench trial, one for each side. The plaintiffs' expert was Dr. Frank Landy, a professor at Pennsylvania State University. Dr. Landy prepared a major report for Congress on whether mandatory retirement policies should be used for law enforcement officers, concluding that they should be replaced by a physical and cognitive testing regime. He testified that age was not a good indicator of an individual's ability to do the job, at least until one reaches seventy years of age. He concluded that a system that possibly eliminated twenty-five percent of qualified persons from state service was grossly overinclusive.
¶ 8. The findings indicate that the court was skeptical about some of Landy's conclusions. For example, it found that the ability to develop a valid and acceptable testing regime "is more difficult than Landy opines." In later findings, the court noted that the advantages of a mandatory retirement system include administrative simplicity, due to the bright-line nature of such a system. It added that a testing regime would be both costly and susceptible to challenges from individual officers and perhaps the officers' union. Finally, it noted that a testing regime would raise issues about gender differences.
¶ 9. Defendants' expert witness was Dr. Deborah Gebhardt, who has extensive experience designing tests for job evaluation, particularly for public safety departments. Dr. Gebhardt reviewed Vermont trooper performance testing data in relation to a national database. She determined that performance declined with age and that the difference between those who are over fifty-five and those who are between fifty and fifty-five is significant. She found that there was a strong correlation between physical testing results and job performance evaluations. She supported the mandatory-retirement rule set at age fifty-five because of the evidence of decline in physical and cognitive abilities that comes with age. Dr. Gebhardt further stated that she did not believe that tests exist that could adequately and safely replace the age rule. She testified that she understands that the mandatory retirement rule is overinclusive, but believes that the presence of even twenty-five percent of the troopers who could not meet established physical standards is a major safety concern that justifies mandatory retirement. The trial court specifically agreed with this conclusion.
¶ 10. Defendants also offered witnesses, including two former Commissioners of Public Safety, who testified to the administrative and workforce development advantages of a mandatory retirement law. For instance, the court summarized former Commissioner Kerry Sleeper's testimony as follows:
He believes that the mandatory retirement system keeps the force viable and promotes younger troopers remaining with the force as the possibility of moving up the ranks is not blocked by long-timers remaining in the upper positions. He found that the arrangement allowed for better planning of promotions and changes since it could easily be ascertained when an officer would leave and open up a position. Since Vermont trooper pay lags behind many states, Kerry found the age 55 retirement helped in recruiting in that they could anticipate promotions as older troopers left. Obviously, the retirement after twenty years on full benefit was also a 473*473 major factor as was, he felt, the overall reputation of the Vermont State Police.
Other Department witnesses testified similarly.
¶ 11. As discussed in more detail below, the trial court found that the mandatory retirement law does not violate Chapter I, Article 7 of the Vermont Constitution, summarizing that "the overall goal of generally enhanced public safety by a fit force is found to reasonably and necessarily require the mandatory age retirement provision." The court added that the fact that some police forces have employed testing and evaluations as an alternative to mandatory retirement "does not mean such systems are better in achieving the stated goal or make the use of a mandatory age provision unconstitutional."
¶ 12. Before we perform the state constitutional analysis required in this case, it is useful to raise two related matters that give context to that analysis. First, the United States Supreme Court has found a similar mandatory retirement scheme to be constitutional under the Equal Protection Clause of the Fourteenth Amendment to the United States Constitution. See Mass. Bd. of Ret. v. Murgia, 427 U.S. 307, 96 S.Ct. 2562, 49 L.Ed.2d 520 (1976) (per curiam). Murgia involved a challenge to a Massachusetts law under which state police officers were required to retire at age fifty. The Court held that the retirement law did not trigger strict scrutiny under the Equal Protection Clause because "a right of governmental employment" is not fundamental and the class of police officers over age fifty is not suspect. Id. at 313, 96 S.Ct. 2562. Thus, it subjected the retirement law to rational basis review, finding:
[T]he Massachusetts statute clearly meets the requirements of the Equal Protection Clause, for the State's classification rationally furthers the purpose identified by the State: Through mandatory retirement at age 50, the legislature seeks to protect the public by assuring physical preparedness of its uniformed police. Since physical ability generally declines with age, mandatory retirement at 50 serves to remove from police service those whose fitness for uniformed work presumptively has diminished with age. This clearly is rationally related to the State's objective. There is no indication that [the law] has the effect of excluding from service so few officers who are in fact unqualified as to render age 50 a criterion wholly unrelated to the objective of the statute.
Id. at 314-16, 96 S.Ct. 2562 (footnotes omitted). The Court acknowledged that fitness might be determined "more precisely through individualized testing," but held that the state had no responsibility to choose that method. Id. at 316, 96 S.Ct. 2562.
¶ 13. Murgia was a per curiam opinion joined by all justices except Justice Marshall, whose dissent criticized the Court's two-tiered equal protection jurisprudence and suggested a standard involving more active review. See id. at 325, 96 S.Ct. 2562 (Marshall, J., dissenting). Justice Marshall further characterized the mandatory retirement statute as the "height of irrationality" given the use and availability of physical ability testing to determine if a state should permit an officer to stay on 474*474 the job. Id. at 326, 96 S.Ct. 2562. He therefore concluded that the mandatory retirement law was "so overinclusive that it must fall." Id. at 325, 96 S.Ct. 2562.
¶ 14. Shortly after Murgia, the United States Supreme Court upheld a sixty-year-old mandatory retirement age for federal foreign service employees in Vance v. Bradley, 440 U.S. 93, 99 S.Ct. 939, 59 L.Ed.2d 171 (1979). One of the main policies underlying the law at issue was "the rapid advancement of men of ability to positions of responsibility and the elimination of men who have reached their ceilings of performance." Id. at 99, 99 S.Ct. 939 (quotation omitted). Thus, age-based mandatory retirement was accompanied by a "selection out" process under which employees who were not going to be promoted or who did not meet performance standards for their class were also retired. The Supreme Court accepted this purpose as legitimate on the grounds that it assured that "opportunities for promotion would be available despite limits on the number of personnel classes and on the number of positions in the Service." Id. at 101, 99 S.Ct. 939.
¶ 15. As in Murgia, the primary rationale accepted by the Court in Vance was the need to guard against the decline in mental and physical reliability of aging employees. See id. at 97, 99 S.Ct. 939. In upholding the law, the Court noted the extremely difficult burden that the plaintiffs had to overcome to successfully challenge the law. Id. at 111, 99 S.Ct. 939. They were required to demonstrate that "the legislative facts on which the classification [was] apparently based could not reasonably be conceived to be true by the governmental decisionmaker." Id. Therefore, the fact that the plaintiffs admitted the negative effects of the aging process was fatal to their case. Id. at 111-12, 99 S.Ct. 939. Although a forced retirement age of seventy may have been more rational, the Court noted that it is not the role of the courts, when applying rational basis review, to resolve factual disputes as to the best method of furthering the legislative purpose. Id. at 112, 99 S.Ct. 939. Rather, the courts' role is limited to determining whether the legislature had any reasonable basis for acting as it did. Id. at 111, 99 S.Ct. 939. In fact, the Court concluded, the plaintiffs' awareness that the facts were arguable in itself immunized the statute from attack. Id. at 112, 99 S.Ct. 939.
¶ 16. Justice Marshall also dissented in Vance, taking issue primarily with the majority's assertion that the factual record was irrelevant to the proper inquiry. Id. at 124, 99 S.Ct. 939 (Marshall, J., dissenting). He argued that mandatory retirement was unnecessary because of the selection out process used by the employer. Id. at 122, 99 S.Ct. 939. He additionally contested the promotional opportunity rationale on the grounds that it would apply in any employment situation where there are a limited number of high-level positions. Id. at 123, 99 S.Ct. 939.
¶ 17. Plaintiffs here argue that we should follow Justice Marshall's dissents, because the more active standard he espoused is essentially that required by the Common Benefits Clause of the Vermont Constitution. Defendants argue in response that the majority decision in Murgia should control because the standard the majority used is closer to that appropriate under the Common Benefits Clause.
¶ 18. The second matter providing context to our state constitutional analysis involves the treatment of mandatory retirement laws for public safety officers under the federal Age Discrimination in Employment Act (ADEA). In general, mandatory retirement provisions are valid under the ADEA only if age is found to be a bona fide occupational requirement 475*475 for the job. See 29 U.S.C. § 623(f)(1). However, in 1986 Congress adopted a safe harbor that permitted mandatory retirement of state and local firefighters and law enforcement officers if the requirement was in place on March 3, 1983. See id. § 623(j)(1)(A). The safe harbor expired at the end of 1993, but was retroactively reinstated and expanded in 1996. State and local employers are now allowed to enact new mandatory retirement laws for firefighters and law enforcement officers provided that the retirement age imposed is not lower than fifty-five. See id. § 623(j)(1)(B). The safe harbor was, however, made contingent on the development of "valid, nondiscriminatory job performance tests" by the National Institute for Occupational Safety and Health. See id. § 623(j)(1) (employers must comply with § 3(d)(2) of the Age Discrimination in Employment Amendments of 1996); Pub.L. 104-208, § 119(2)(a), 110 Stat. 3009. These tests were to be identified in regulations promulgated by the Secretary of Health and Human Services, and employees who took and passed an identified test were to receive an exemption from the mandatory retirement requirement at issue. Pub.L. 104-208, § 119(2)(a), 110 Stat. at 3009-25. Congress gave the Secretary a four-year deadline to develop advisory guidelines as a beginning of the process of issuing regulations, see id., but there has been no action by the Secretary to date. In the meantime, the courts have enforced the safe harbor provision without the expected testing right. See Correa-Ruiz, 573 F.3d at 11-12.
¶ 19. The legislative history of the 1996 safe harbor provisions reflects a division in Congress over the efficacy of testing as an alternative to mandatory retirement laws. See Drnek v. City of Chicago, 192 F.Supp.2d 835, 841-42 (N.D.Ill.2002) (ultimately affirmed by Minch v. City of Chicago, 486 F.3d 294 (7th Cir.2007)). The court in Drnek quoted floor debate in the House and Senate from sponsors of the 1996 provisions indicating that existing tests were inadequate to ensure a fit and qualified workforce and that the tests discriminated against women and minorities. Id. at 841. The failure of the Secretary to act has left the controversy unresolved at the national level.
476*476 ¶ 20. With this context in mind, we turn to plaintiffs' challenge under the Vermont Constitution. We start by emphasizing that statutes are presumed to be constitutional, Choquette v. Perrault, 153 Vt. 45, 51, 569 A.2d 455, 458 (1989), and are presumed to be reasonable, Colchester Fire Dist. No. 2 v. Sharrow, 145 Vt. 195, 199, 485 A.2d 134, 137 (1984). We have often observed that the proponent of a constitutional challenge has a very weighty burden to overcome. Sharrow, 145 Vt. at 199, 485 A.2d at 137.
¶ 21. The Common Benefits Clause of the Vermont Constitution provides, in pertinent part, "[t]hat government is, or ought to be, instituted for the common benefit, protection, and security of the people, nation, or community, and not for the particular emolument or advantage of any single person, family, or set of persons, who are a part only of that community." Vt. Const. ch. I, art. 7. In Baker, 170 Vt. at 202-11, 744 A.2d at 869-77, we detailed the history of this constitutional provision, its historical context, the evolution of our caselaw interpreting this provision, and the relationship between our inquiry under this clause to that under the Fourteenth Amendment's Equal Protection Clause. We rejected the rigid, multi-tiered analysis of the federal Equal Protection Clause analysis in favor of "a relatively uniform standard, reflective of the inclusionary principle at [the Common Benefits Clause's] core." Id. at 212, 744 A.2d at 878. We explained the relevant inquiry as follows:
When a statute is challenged under Article 7, we first define that "part of the community" disadvantaged by the law. We examine the statutory basis that distinguishes those protected by the law from those excluded from the state's protection. ...
We next look to the government's purpose in drawing a classification that includes some members of the community within the scope of the challenged law but excludes others. Consistent with Article 7's guiding principle of affording the protection and benefit of the law to all members of the Vermont community, we examine the nature of the classification to determine whether it is reasonably necessary to accomplish the State's claimed objectives.
We must ultimately ascertain whether the omission of a part of the community from the benefit, protection and security of the challenged law bears a reasonable and just relation to the governmental purpose. Consistent with the core presumption of inclusion, factors to be considered in this determination may include: (1) the significance of the benefits and protections of the challenged law; (2) whether the omission of members of the community from the benefits and protections of the challenged law promotes the government's stated goals; and (3) whether the classification is significantly underinclusive or overinclusive.
¶ 22. Under the inquiry outlined in Baker, we first identify the part of the community affected by this law. The parties do not dispute that state public safety employees assigned to police and law enforcement duties over age fifty-five who, before their fifty-fifth birthdays, were capable of working as police officers, constitute the affected segment of the community. See 3 V.S.A. § 459(a)(2); see also id. § 455(a)(11)(C), (9)(B) (applicable definitions).
477*477 ¶ 23. We next examine the proffered governmental purpose for drawing a classification that includes some members of the community but excludes others. The purpose asserted here is to maintain a state police force that is mentally and physically capable of performing the task of ensuring public safety. As we noted in Baker, any "statutory exclusions from publicly-conferred benefits and protections must be premised on an appropriate and overriding public interest." 170 Vt. at 206, 744 A.2d at 873 (quotation omitted). It must be conceded that having a mentally and physically capable police force is an appropriate or highly important purpose. See Murgia, 427 U.S. at 325, 96 S.Ct. 2562 (Marshall, J., dissenting) (conceding that "the purpose of the mandatory retirement law is legitimate, and indeed compelling").
¶ 24. Plaintiffs argue that the State also has an interest in guarding against age discrimination, as shown generally in the Vermont Fair Employment Practices Act (FEPA). Cf. Gately v. Massachusetts, 811 F.Supp. 26, 28 (D.Mass. 1992) ("While the public has a compelling interest in the continued efficient provision of law enforcement services, it also has a compelling interest in eradicating age discrimination."), aff'd, 2 F.3d 1221 (1st Cir. 1993). We understand plaintiffs to argue that the second interest weakens the first, or makes it illegitimate, and that we should weigh this effect in examining the interest underlying the law. We disagree. The Legislature adopted both FEPA and the mandatory retirement age legislation for public safety officers, apparently believing that there was no fundamental contradiction between the two laws. In fact, the Legislature included in FEPA a specific exclusion for police and firefighter retirement systems. See 21 V.S.A. § 495f. Our function is not to substitute our view of the appropriate balance for that of the Legislature. In our Common Benefits Clause inquiry, we do not judge whether the policy decision made by the Legislature was wise, but rather whether this decision to exclude a portion of the community from the common protection of the law was reasonable and just in light of its purpose. Cf. Harris v. McRae, 448 U.S. 297, 326, 100 S.Ct. 2671, 65 L.Ed.2d 784 (1980) ("In making an independent appraisal of the competing interests involved here, the District Court went beyond the judicial function. Such decisions are entrusted under the Constitution to Congress, not the courts. It is the role of the courts only to ensure that congressional decisions comport with the Constitution.").
¶ 25. We next turn to the third element of the Common Benefits Clause analysis and consider whether the mandatory retirement provision bears a reasonable and just relationship to legitimate state interests. In making their argument, plaintiffs emphasize the phrase "reasonably necessary" as used in Baker. See 170 Vt. at 213-14, 744 A.2d at 878 ("Consistent with Article 7's guiding principle of affording the protection and benefit of the law to all members of the Vermont community, we examine the nature of the classification to determine whether it is reasonably necessary to accomplish the State's claimed objectives."). They argue that the mandatory retirement policy may be desirable, but it cannot be considered 478*478 necessary, especially in relation to the alternatives. They also cite federal authority that mandatory retirement requirements have not been found reasonably necessary such that age would be a bona fide occupational qualification. See supra, ¶ 18 n. 4.
¶ 26. In reading Baker as a whole, we think that plaintiffs have placed too much emphasis on the word "necessary." Baker requires that a classification scheme be "reasonable and just" in relation to the governmental purpose. See id. at 214, 217-24, 744 A.2d at 879, 881-86. Our subsequent decisions support this formulation of the Baker test. See, e.g., In re Estate of Murcury, 2004 VT 118, ¶ 15, 177 Vt. 606, 868 A.2d 680 (mem.) (concluding that statutory twenty-one-year limitations period for establishing paternity "furthers reasonable and just governmental objectives"); see also Samaha v. Scott's Constr., Inc., 543 F.Supp.2d 341, 346 (D.Vt.2008) (characterizing the Baker test as evaluating "whether the omission of a part of the community from the benefits of the challenged law bears a reasonable and just relation to the governmental purpose" (quotation omitted)). Indeed, an inquiry into necessity would contravene the deference which must control our inquiry and place us in the position of reviewing the wisdom of legislative choices.
¶ 27. In evaluating whether the mandatory retirement statute bears a reasonable and just relation to the governmental purpose, Baker identifies three factors that may be considered: (1) the significance of the benefits to the excluded group; (2) whether the omission of a part of the community promotes the government's stated goals; and (3) whether the classification is overinclusive or underinclusive. 170 Vt. at 214, 744 A.2d at 879.
¶ 28. With respect to the first factor, the statute at issue deprives plaintiffs and similarly situated persons of the ability to continue working as state police officers at age fifty-five. Additionally, this cut-off deprives anyone who started his or her career with the state police after age thirty-five from attaining full retirement benefits because these benefits are available only after twenty years of service. We acknowledge that forcing a person to retire from his or her chosen profession with his or her chosen employer is a significant burden. The statute does not forbid all employment, however, or even employment as a police officer with a nonstate employer. The right to work as a state-employed police officer is not as significant a governmental interest as the right to the benefits of marriage addressed in Baker or the right to educational opportunities addressed in Brigham v. State, 166 Vt. 246, 692 A.2d 384 (1997). Additionally, although depriving an officer of the ability to earn full retirement benefits imposes a burden, this burden is somewhat mitigated by virtue of it being undertaken knowingly by the officers who entered into their employment after the age of thirty-five.
¶ 29. The nature of the deprivation is demonstrated by the facts of this case. Plaintiffs were productive and effective state police officers who would have maintained their jobs for some period under a regime that solely examined their individual capacities to continue. On the other hand, they were able to find other jobs in law enforcement after their forced retirement, albeit likely with reduced compensation and responsibility.
¶ 30. With respect to the second factor, the mandatory retirement statute does advance the State's goal of having a police force that is mentally and physically capable 479*479 of performing its tasks. The trial court concluded that the risk that an officer stays on the force when he or she cannot capably perform the duties required "clearly grows with age," and that risk was underscored by the evidence from defendants' expert witness. See also Murgia, 427 U.S. at 315, 96 S.Ct. 2562 (noting that "physical ability generally declines with age"). The trial court also found that by age fifty-five the physical skills and abilities of a state police officer executing his or her duties have declined. A mandatory retirement age does not advance the State's goal with respect to younger officers whose capabilities have declined, but there is no evidence that this omission is a significant lapse.
¶ 31. Plaintiffs' main arguments involve the third factor: whether the classification is overinclusive or underinclusive. They argue that: (1) the mandatory retirement law involuntarily retires a high percentage of those who are physically fit to serve— approximately seventy-five percent of those retired—and is therefore too overinclusive to meet the constitutional standard; and (2) physical-ability testing can protect the State's interest without being overinclusive. On this factor, the trial court determined that the age fifty-five rule is overinclusive, as plaintiffs argued, in that a percentage of officers over the age of fifty-five are still capable of meeting the requirements of the job. However, the court also determined that all officers gradually lose their physical skills and abilities as they age. As a group, officers in their fifties are generally less physically skilled than those in their twenties, and even those in their forties. It is also clear that some of the declines, both mental and physical, are modest or minimal and would be hard, if not impossible, to detect. Indeed, they may not even obviously affect the day-to-day work of an officer. At some point in one's seventies to eighties, however, a person would simply not be able to effectively and safely perform the work of a state police officer. Though the chance of a serious incident happening because of an officer's inability to perform his job is low, the consequences could be grave if the inability is discovered during a life-or-death situation. Such an occurrence could jeopardize the safety of the members of the public and fellow officers.
¶ 32. While the court concluded that the mandatory retirement policy was overinclusive, it did not find that seventy-five percent of officers over age fifty-five are physically able to perform the activities of a state police officer. It instead concluded that the overinclusiveness was "limited." In advocating the seventy-five percent figure, plaintiffs rely on the testimony of Dr. Gebhardt, a witness for defendants. Because of her extensive experience in the field, Dr. Gebhardt had a large database of physical-ability tests and peer and supervisor job-performance evaluations for law enforcement workers from other jurisdictions. She analyzed those tests to show that physical ability declines with age and that evaluations of an officer's ability to do the job, as determined by peers and supervisors, also decline with age. The latter finding was shown in part by a chart in her report that displayed the percentage of workers by age who were judged by peers and supervisors to provide "Unacceptable 480*480 Physical Job Performance." The percentage rises with age and equals twenty-five percent for those fifty-five years of age and older. Based on this chart, plaintiffs argue that defendants' witness agrees that seventy-five percent of workers fifty-five years of age and older provided acceptable job performance.
¶ 33. The report is an evidentiary item that the trial court could accept or reject irrespective of its source. There are significant weaknesses in the logical inferences plaintiffs draw from the report that would support rejection by the trial court. For instance, there is no indication, and plaintiffs do not claim, that peer and supervisor evaluation, without more, is an acceptable method of determining the physical ability of employees. Moreover, the line between acceptable and unacceptable job performance was arbitrarily selected by Dr. Gebhardt and not by defendants or, in fact, any law enforcement body. Dr. Gebhardt set the line at "one standard deviation below the mean for the database sample" because "it is a standardized statistical level used by researchers in criterion-related validity studies." Whether defendant Department would use a statistical measure as its demarcation line between acceptable and unacceptable performance is a matter of speculation.
¶ 34. We must also consider the court's conclusion, drawn from defendants' witness, that because of the interdependent nature of police work, even a low percentage of police with unacceptable performance abilities significantly impairs the Department's ability to perform its mission. If one member of a team cannot perform as required, the entire team cannot perform as required.
¶ 35. We cannot conclude, based on the trial court's findings and conclusions and the evidence on which they rest, that the mandatory retirement law is so overinclusive that it violates the Common Benefits Clause as a matter of law if there is no alternative to mandatory retirement to meet the State's legitimate objectives. We emphasize that the trial court found the degree of overinclusiveness to be "limited" and agreed with defendants' expert witness that twenty-five percent of employees not being able to meet performance standards was a "major safety concern" that justified the mandatory retirement policy. Though mandatory retirement is overinclusive in meeting the need to ensure public safety, the degree of overinclusiveness is speculative, especially in relation to the interdependent nature of the work.
¶ 36. This brings us to plaintiffs' major argument, that the degree of overinclusiveness must be viewed in relation to alternatives that will meet the State's legitimate needs without the same degree of overinclusiveness. Specifically, plaintiffs argue that the State's need can be met through performance testing of officers with virtually no overinclusiveness. In response to this argument and the evidence of plaintiffs' expert witness, the trial court determined that there are possible alternatives to the age fifty-five rule, noting that there are physical and mental tests used by law enforcement agencies in other jurisdictions to determine those officers who can perform the necessary duties of their work and those who cannot. It found that these tests can be combined with peer and supervisory reviews to evaluate conduct, personality, and work habits. The trial court responded specifically to the argument that the current state police physical ability testing would meet the State's objective. It found that the current testing involved "no more than very general physical fitness tests and would not as presently designed meet the needs of such a non-mandatory retirement plan."
481*481 ¶ 37. The trial court was skeptical about the feasibility of replacing a mandatory retirement law with performance testing to meet the State's objectives. The two expert witnesses differed on the issue. In response to the strong endorsement of testing by plaintiffs' expert witness, the court observed that coming up with a testing regime "is more difficult than [Dr.] Landy opines." The court cited the cost of establishing testing, although it could not quantify that cost. It found serious unresolved issues relating to how testing could account for gender or age. It found "major policy issues" in establishing a "cut score." It was particularly concerned with the political opposition to testing, the difficulty of overcoming such opposition and the fact "that the field has not progressed as far as it could ... because of political concerns." The court concluded that the fact that some police forces have had to adopt testing and evaluations, as an alternative to age-based mandatory retirement, "does not mean such systems are better in achieving the stated goal." The trial court could not find that a testing and evaluation regime would reduce the overinclusiveness of a mandatory retirement policy and, at the same time, meet the State's public safety goals.
¶ 38. In reaching our conclusion, we must define the relationship between the evidence presented by the parties and our role in determining the constitutionality of the statute involved. This relationship appears to involve the greatest difference between our approach and that of the dissent. We emphasized at the outset that statutes are presumed to be constitutional and we must accord deference to the policy choices made by the Legislature. To implement these considerations, the United States Supreme Court held in Vance that "it is the very admission that the facts are arguable that immunizes from constitutional attack the congressional judgment represented by this statute." 440 U.S. at 112, 99 S.Ct. 939. The holding in Vance was amplified in Minnesota v. Clover Leaf Creamery Co., 449 U.S. 456, 101 S.Ct. 715, 66 L.Ed.2d 659 (1981) in which milk sellers challenged a state statute that banned the sale of milk in plastic, nonreturnable, nonrefillable containers, but not in other nonreturnable, nonrefillable containers:
Although parties challenging legislation under the Equal Protection Clause may introduce evidence supporting their claim that it is irrational, they cannot prevail so long as "it is evident from all the considerations presented to [the legislature,] and those of which we may take judicial notice, that the question is at least debatable." Where there was evidence before the legislature reasonably supporting the classification, litigants may not procure invalidation of the legislation by merely tendering evidence in court that the legislature was mistaken.
Id. at 464, 101 S.Ct. 715 (quoting United States v. Carolene Prods. Co., 304 U.S. 144, 154, 58 S.Ct. 778, 82 L.Ed. 1234 (1938)).
¶ 39. To give deference to the Legislature, we must follow at least a modified form of the approach taken in Clover Leaf Creamery for Article 7 purposes. We see no inconsistency between our holding in Baker, which recognized that we must give deference to the Legislature, and applying the Clover Leaf Creamery approach in appropriate cases. We emphasize that in doing so we are not adopting the federal rational basis standard for evaluating most equal protection claims.
¶ 40. Though we have no evidence of the legislative record in this case, the issues are well framed by the national debate. Indeed, the expert evidence in this case reads like a microcosm of the national 482*482 debate. Also, unlike virtually all of our recent Article 7 cases, including Baker, see also Choquette, 153 Vt. at 54-55, 569 A.2d at 460 (because of changed circumstances, fence viewer act can "no longer" withstand constitutional attack), the distinctions among community members represented here are of relatively recent origin, judging by the age of the legislation under review. See 1971, No. 231 (Adj.Sess.), § 4, as amended by 1977, No. 80, § 1 (enacting 3 V.S.A. § 459(a)(2)); 1981, No. 65, § 4 (enacting 21 V.S.A. § 495f). Moreover, the exact issues being debated in this litigation remain under active investigation and consideration in the political process.
¶ 41. The latter point is important for cases like this. A determination of unconstitutionality would end development of the issue in Vermont. The Legislature, by contrast, can experiment with different approaches to protecting public safety, without irrevocably choosing one until the right approach is clear. The underlying facts are in flux, and changes in those facts will affect the nation's and Vermont's responses. Just as Justice Souter observed in Washington v. Glucksberg with respect to a possible constitutional right to assisted suicide, here we "do not decide for all time that [the plaintiffs'] claim should not be recognized." 521 U.S. 702, 789, 117 S.Ct. 2258, 138 L.Ed.2d 772 (1997) (Souter, J., concurring). We decide only that at this time and on this record, we will follow the Clover Leaf Creamery approach and stay our hand. It would be inappropriate for us to intrude under Article 7 into an ongoing political process that has not reached its end point. We conclude that 3 V.S.A. § 459(a)(2) does not violate Article 7 at this time.
¶ 42. In reaching this conclusion, we specifically reject the dissent's argument that when "substantial evidence from qualified experts is adduced against [the law],... the [State] has the burden of meeting it fairly and refuting it." Post, ¶ 60. We agree that the expert evidence in this case was relevant to define the extent and nature of the factual disagreement, but we cannot agree that our decision must rest exclusively on this evidence. The dissent's rule would nullify legislative fact-finding whenever the Court finds that the challenger's expert witness is more persuasive than that put forward by the State. It would mean that no deference would be given to the Legislature's policy choice, nor to the Legislature's own analysis of the factual circumstances that necessarily occurred during the enactment process. As an example of that effect, the dissent would strike down a Vermont statute based primarily on the testimony of an expert witness whose findings and conclusions were rejected by Congress. Further, that rule would place the burden on the State to prove a statute is constitutional, directly rejecting our many holdings that statutes are presumed to be constitutional. Though the dissent pays lip service to deference, its position is the antithesis of deference.
¶ 43. In addition to challenging 3 V.S.A. § 459(a)(2), which contains the age 55 retirement requirement for employees of the Department of Public Safety "assigned to police and law enforcement duties," 3 V.S.A. § 455(9)(B), plaintiffs challenge 21 V.S.A. § 495f, a section of the Fair Employment Practices Act, setting forth exemptions from the requirements of the act and providing that "[m]andatory retirement on account of age, necessitated under a police or firefighter retirement system, is specifically authorized." Plaintiffs' arguments as to why this latter section is unconstitutional are the same as those with respect to 3 V.S.A. § 459(a)(2). For the reasons specified above, we conclude that 21 V.S.A. § 495f authorizing the 483*483 mandatory retirement of police officers does not violate the Common Benefits Clause.
¶ 44. We acknowledge that plaintiffs made many strong policy arguments that the law should abandon a mandatory retirement age for police officers, or that the Legislature should raise the mandatory retirement age to be less overinclusive. These arguments are more appropriately directed to the Legislature than to this Court. We do not decide today that the statute best fulfills the relevant social and economic objectives of the Legislature. We decide only that the mandatory retirement line currently drawn by the Legislature bears a reasonable and just relation to a legitimate state interest, and for that reason, we find no violation of the Common Benefits Clause of Chapter I, Article 7 of the Vermont Constitution.
JOHNSON, J., dissenting.
This is the 21st century. This is the United States of America. This is Vermont. Nobody should lose their job because they have a birthday and because of no other reason.
¶ 45. State Trooper George Badgley's trial testimony reflects both the pain of the loss of a chosen career and the incomprehension of being victimized by a policy allegedly driven by public safety concerns, but actually designed to protect entrenched economic and bureaucratic interests. Now, Trooper Badgley and his co-plaintiff are confronted with a decision of this Court that treats their compelling civil rights claim as if it were no more than a challenge to ordinary economic regulation and that, at the same time, creates confusion and uncertainty in an important area of our constitutional jurisprudence.
¶ 46. The issue in this case is whether the automatic termination of employment of Vermont's state troopers at the age of fifty-five bears a reasonable and just relationship to a significant legislative goal. Public safety is the purported basis for this discriminatory policy, but a review of the trial court's findings and conclusions, as well as the evidence upon which they are based, reveal that the alleged basis is mere pretext. The bulk of the State's evidence pertained to administrative concerns that, as the trial court explicitly stated, do not, in and of themselves, justify the discriminatory policy. These concerns, not a perception of any actual threat to public safety, underlie the law here at issue. And even if we assume that the statute's discriminatory classification is intended to protect the public safety, neither the evidence nor the trial court's findings demonstrate that mandatory retirement is either a reasonable or fair way of meeting that objective. In short, the mandatory retirement law does not meet the constitutional test under the Common Benefits Clause as set forth in Baker v. State, 170 Vt. 194, 744 A.2d 864 (1999), and other decisions of this Court. Therefore, I respectfully dissent from the majority's decision to uphold the statute.
¶ 47. By all accounts, both plaintiffs in this case were exemplary officers who are mentally and physically capable of continuing in their jobs and serving the citizens of Vermont. Their circumstances are particularly compelling because their forced retirement is not the first time that their lives have been affected by misguided discriminatory policies.
¶ 48. At the age of twenty-five, after working in the Rutland County State's Attorney's Office, Ruth Whitney knew she wanted to be a state trooper. At the time, however, she was advised that the Vermont State Police did not accept female candidates. Undeterred from fulfilling her desire to work in law enforcement, Trooper 484*484 Whitney took employment with a state's attorney's office and two municipal police forces. She worked for the Middlebury Police Department for ten years, and was honored as "Officer of the Year" in 1986. In 1994, she applied to the State Police, who were then accepting female candidates. She was the only female candidate, and the oldest candidate at age forty-three, in her class at the Vermont Academy. It is undisputed that Trooper Whitney is physically fit and capable of continuing to perform her job at the same high level as she has done over the past fifteen years. Yet, once again, the State seeks to prevent her from working on the force by raising the same, long-since discredited, stereotypical notions that a person of a particular gender or age would be unable to withstand the physical rigors of police work.
¶ 49. Similarly, Trooper Badgley was unable to apply to be a state trooper in 1980 because he had turned thirty-five, and at that time the Vermont State Police were not accepting candidates who had reached their thirty-fifth birthday. He later joined the force after that policy was ended. Now, like Trooper Whitney, Trooper Badgley is once again, after years of exceptional service, the victim of the same discredited and discriminatory policies that delayed his entry into the Vermont State Police, even though he is concededly mentally, morally and physically fit for the job.
¶ 50. The majority, the trial court, and the parties all acknowledge that the test set forth in Baker governs the determination of whether the mandatory retirement law's discriminatory classification violates the Common Benefits Clause. That provision prohibits governmental classifications that confer benefits and protections to part of the community while arbitrarily or unreasonably denying them to others. We emphasized in Baker that the Common Benefits Clause is distinct from the Fourteenth Amendment's Equal Protection Clause, and that it requires a far more rigorous review than the highly deferential rational-basis standard applied under an equal protection analysis not involving fundamental rights. 170 Vt. at 202, 744 A.2d at 870 (stating that Common Benefits Clause "differs markedly from the federal Equal Protection Clause in its language, historical origins, purpose, and development").
¶ 51. In Baker, we characterized our review under the Common Benefits Clause as a "weighing process" that considers the "nature and importance of the benefits and protections affected by the legislation," thereby imposing "a `more stringent' reasonableness inquiry than was generally associated with rational basis review under the federal constitution." Id. at 203-04, 744 A.2d at 871. We stressed that, though we had not abandoned the traditional deference accorded to legislation, Vermont courts had the responsibility "to evaluate the object and effect" of the legislation so that we could "engage in a meaningful, case-specific analysis" aimed at ensuring that any exclusion of benefits bears "a just and reasonable relation to the legislative goals." Id. at 204, 744 A.2d at 872.
¶ 52. We also stressed, in examining past cases, that we would not be content to accept prejudices from a bygone era or policies that "failed to establish a reasonable relation to the public purpose in the light of contemporary circumstances." Id. at 205, 744 A.2d at 873; see MacCallum v. Seymour's Adm'r, 165 Vt. 452, 461, 686 A.2d 935, 940 (1996) ("acknowledg[ing] the vast cultural and social changes that have 485*485 occurred and their effect on adoption practice and the public attitudes about adoption" in holding that a statute denying an adopted person's right of inheritance violates the Common Benefits Clause); Choquette v. Perrault, 153 Vt. 45, 53-54, 569 A.2d 455, 460 (1989) (determining that a statute requiring a landowner to share the cost of a fence constructed along a common border by, and solely for the benefit of, the adjacent property owner, may have been reasonable when most Vermont land was open and farmed and when most rural landowners owned livestock, but was arbitrary under modern circumstances when applied against a party that owned no livestock).
¶ 53. Thus, we require that statutory exclusions from publicly conferred benefits be premised on an appropriate and overriding public interest, and that the classifications revealed by such exclusions "bear a reasonable and just relation to the governmental objective in light of contemporary conditions." Baker, 170 Vt. at 206, 744 A.2d at 873. To ensure this rigorous review, Baker instructs us to consider: (1) the significance of the benefits and protections at stake; (2) whether omission of members of the community from those benefits and protections promote the government's stated goals; and (3) whether the challenged classification is significantly underinclusive or overinclusive. Id. at 214, 744 A.2d at 879.
¶ 54. Here, while purportedly accepting the rigorous Baker test, the majority ultimately applies a minimal rational-basis standard of review. In the end, after perfunctorily going through the Baker criteria, the majority adopts as "valid here" the following comment from Vance v. Bradley, 440 U.S. 93, 112, 99 S.Ct. 939, 59 L.Ed.2d 171 (1979), a case that expressly employed the federal rational-basis test more than thirty years ago to uphold the mandatory retirement of foreign service workers at age sixty: "it is the very admission that the facts are arguable that immunizes from constitutional attack the congressional judgment represented by this statute." Ante, ¶ 38. Thus, despite the more rigorous standard set forth in Baker, the majority ultimately reverts to a rational-basis test that immunizes from constitutional attack any law that has any conceivable legitimate purpose. See Vance, 440 U.S. at 111, 99 S.Ct. 939 (framing question under rational-basis standard as whether legislative facts justifying classification could "reasonably be conceived to be true by the governmental decisionmaker"). That is the majority's error here.
¶ 55. Our first task under Baker is to define the part of the community disadvantaged by the law through the loss of benefits made available to another part of the community. Here, the disadvantaged members of the community are persons who have joined or will join the Vermont State Police, who wish to continue their service past the age of fifty-five, and who are fully capable of doing so, both mentally and physically. A subset of this class consists of those who join the force after the age of thirty-five and thus will never be able to qualify for full benefits upon retirement. The statutory basis distinguishing this class, chronological age, is an unalterable physiological characteristic that raises concerns beyond mere economic classifications. As Justice Thurgood Marshall stated in his dissent in Massachusetts Board of Retirement v. Murgia, 427 U.S. 307, 321, 96 S.Ct. 2562, 49 L.Ed.2d 520 (1976) (Marshall, J., dissenting), wherein he called for application of a standard of review similar to our Baker standard and more rigorous than the rational-basis standard applied by the majority:
There is simply no reason why a statute that tells able-bodied police officers, ready and willing to work, that they no longer have the right to earn a living in their chosen profession merely because they are 50 years old should be judged by the same minimal standards of rationality that we use to test economic legislation that discriminates against business interests.
¶ 56. "Consistent with the core presumption of inclusion," Baker requires some consideration of the "relative `weights' or dignities of the contending interests." 170 Vt. at 214, 744 A.2d at 879 (citation omitted). Whatever the State's justifications for discriminating against persons over fifty-five years of age, those justifications must be weighed against plaintiffs' significant interests in earning a living and in the context of the historic discrimination against older workers that is now universally recognized as inconsistent with public policy through the passage of the federal Age Discrimination in Employment Act (ADEA), 29 U.S.C. §§ 621-634, and subsequent state fair employment practices acts throughout the country. E.g., C. Yates, Annotation, Application of State Law to Age Discrimination in Employment, 51 A.L.R.5th 1, 10-38 (1997) (collecting and analyzing state and federal law demonstrating that "[d]iscrimination based upon age has been subjected to close legislative scrutiny, both state and federal").
487*487 ¶ 57. Baker next requires us to identify the government's purpose in drawing the discriminatory classification and to determine whether it bears a reasonable and just relation to the accomplishment of legitimate state objectives. 170 Vt. at 214, 744 A.2d at 879. According to the majority, public safety is the principal purpose underlying the mandatory retirement law. Yet, apart from a few general statements by witnesses opining that the law benefits public safety, the only evidence offered by the State to support a public safety rationale was a report prepared in response to this litigation and arriving principally at the unremarkable conclusion that physical performance of individuals generally declines with age.
¶ 58. Limited as it is in what it offers in support of the State's public safety rationale, the report is essentially the only empirical evidence that the State has to promote that rationale. And even Dr. Gebhardt, the State's principal expert witness and the report's chief author, conceded during her trial testimony that her report indicated fully seventy-five percent of officers fifty-five and older were physically capable of performing their job duties, thus making the classification significantly overinclusive. See Baker, 170 Vt. at 214, 744 A.2d at 879 (citing as factor for consideration in determining constitutionality of law whether law is significantly underinclusive or overinclusive). The trial court acknowledged this fact, but concluded that the twenty-five percent failure rate justified the mandatory retirement law.
¶ 59. The majority contends that it is improper to infer from the Gebhardt report that seventy-five percent of state troopers over the age of fifty-five have the physical capacities required for the job. Its contention rests on the assertion that "the line between acceptable and unacceptable job performance was arbitrarily selected by Dr. Gebhardt and not by defendants or, in fact, any law enforcement body" and that whether the state police would use that demarcation line "is a matter of speculation." Ante, ¶ 33. But this criticism applies equally to the trial court's assumption, upon which it based its conclusion that mandatory retirement promotes public safety, that twenty-five percent of troopers above the age of fifty-five could not meet the physical demands of the job. And if the 75/25 capability ratio is of no evidentiary value, then there is no evidence whatsoever that mandatory retirement promotes public safety, beyond the fact that, to some unquantified degree, physical capabilities tend to diminish with age—an undisputed proposition that begs the question of the case, which is whether an arbitrary age limitation bears a reasonable and just relation to protecting public safety.
¶ 60. Moreover, the implication of the majority's reasoning here is that no amount of evidence regarding the feasibility, utility, or reliability of any scheme of performance measurement could ever show that a state police mandatory retirement policy was an unjust or unreasonable way of achieving a presumed public safety objective. All the State would have to do is show that the performance-measurement scheme was not the one it might employ. I believe that a state agency has a higher duty than that, and I believe that our judicial system requires much more than such a toothless standard. When a serious challenge is made to a discriminatory employment scheme, and substantial evidence from qualified experts is adduced against that scheme—evidence such as that presented by plaintiffs, which I discuss below—the defendant has the burden of meeting it fairly and refuting it. The State plainly failed to do so here, especially as the majority has downgraded the 488*488 only empirical evidence offered by the State to justify mandatory retirement.
¶ 61. In marked contrast, plaintiffs presented substantial evidence, most of it accepted by the trial court, demonstrating that mandatory retirement based solely on chronological age does little if anything to promote public safety. Plaintiffs' primary witness was Dr. Frank Landy. In 1992, as director of the Pennsylvania State University Center for Applied Behavioral Sciences, Dr. Landy headed a team of experts, which included Dr. Gebhardt, who conducted a congressionally commissioned study to determine if the elimination of chronological age as a basis for hiring or retiring police officers would have any impact on public safety. In addition to Dr. Gebhardt, Dr. Landy's committee was composed of nineteen leading experts in the relevant fields of inquiry. The resulting five volume report was reviewed independently by three separate sources— an internal group of experts who were not involved in the background research, a group of congressional committee staffers, and an independent group of scientists named by the Department of Labor and the EEOC. The report was admitted into evidence in this case.
¶ 62. Based on two years of study, the committee concluded that chronological age is not an accurate predictor of either job performance or ability to perform the duties of a public safety officer and that the public would be better served if available testing regimens, rather than a chronological age-cutoff, were used to determine ability to perform the job. The committee emphasized that individuals well into their sixties were capable of performing the physical tasks involved in law enforcement and that physiological declines often attributed to the aging process are more indicative of alterable, poor lifestyle patterns. Accordingly, the members of the committee, including Dr. Gebhardt, unanimously recommended eliminating the exemption for public safety 489*489 officers from the ADEA's ban on mandatory retirement.
¶ 63. The Landy report's findings and conclusions, recapitulated during Dr. Landy's full day of testimony, are consistent with all other related empirical studies, every one of which has confirmed that fitness resulting from lifestyle choices has far more to do with ability to perform police work than chronological age. See Schiff, supra, at 47-51. The findings and conclusions put to rest the claim that mandatory retirement of law enforcement officers promotes public safety. That notion must now be seen for what it is: stereotypical, discriminatory, outdated, and flatly erroneous. As Dr. Landy testified, "from the science part of it, you can turn out the lights and go home. We agree ... that age is: (A), a lousy predictor of human performance; and (B), there are alternatives available." See Gately v. Massachusetts, 811 F.Supp. 26, 31 (D.Mass.1992) (Gately I), (noting that "the Landy report has overturned comfortable old assumptions about the reliability of age as an indicator of ability" to perform police functions), aff'd, 2 F.3d 1221 (1993), cert. denied, 511 U.S. 1082, 114 S.Ct. 1832, 128 L.Ed.2d 461 (1994). In short, law enforcement agencies cannot point to any "convincing, empirically-based data supporting age restrictions in hiring and retirement policies, and the age standard remains an arbitrary one." Schiff, supra, at 52.
¶ 64. In light of these facts, the trial court made numerous findings crediting Dr. Landy's testimony and the extensive data supporting it, including that: (1) "age is a poor predictor of ability to continue in certain work, including law enforcement"; (2) "[t]his would be true until at least 490*490 closing in on 70 years of age"; (3) "[t]here are large individual differences between people for such duties and abilities and these [individual differences] grow as you go up in the age brackets"; (4) therefore, "age is less reliable [as a predictor of fitness and job performance] as you go up in years until around 70 or 80 years of age"; (5) "[a]lthough some decline as one ages is inevitable, [the expert's] studies showed that the rate could not be predicted by simple age brackets and could also be slowed by proper exercise, diet, and such efforts"; (6) it is "also important to note that a `decline' in some areas [of physical fitness] could not necessarily be found to be significant"; (7) "[p]hysically, such concerns as stamina, muscular strength, flexibility are all very individualistic and cannot be assumed by general age brackets"; and (8) "as much as 85% of [police] work" is related to cognitive abilities and personality traits rather than physical abilities or fitness.
¶ 65. These findings further demonstrate the lack of empirical evidence in the record indicating that age-based mandatory retirement is an effective means of protecting the public. Cf. Gately v. Massachusetts, No. Civ. A. 92-13018-MA, 1998 WL 518179, at *6 (D.Mass. June 8, 1998) (Gately II) (in litigation involving an ADEA challenge to mandatory retirement for Massachusetts State Police, the trial court stated that "there is no evidence that [in the six years since the consolidation of the four police forces in Massachusetts] any officer age 55 or older has endangered himself or herself, another officer, or the public while in the line of duty"). Indeed, as noted, Dr. Gebhardt's tests, which were introduced by the State, demonstrated that a certain percentage of officers of all ages would be unable to pass physical performance tests and, thus, that mandatory retirement, by itself, could not ensure that public safety officers would be able to do their job. According to Dr. Gebhardt's report, the percentage of officers unable to pass physical performance tests would be nearly as high for those fifty-to-fifty-four years of age as those fifty-five and older, and a significant number of officers in their forties would fail such tests. As Dr. Landy testified, if you took Dr. Gebhardt seriously, "you would probably not hire anybody who was over 30-years-old."
¶ 66. Taking into account all of the foregoing, the State's public safety rationale amounts to no more than the following: (1) human beings decline physically as they age, albeit at a different rate with each individual; (2) police officers require a certain level of physical fitness to perform certain aspects of their job; (3) police work concerns public safety; and (4) the percentage of officers who are incapable of passing physical performance tests increases with the age of the officers.
¶ 67. Even assuming these facts raise significant public safety concerns, we are not thereby automatically bound to accept the Legislature's method of addressing them. To the contrary, where, as here, that response discriminates against one segment of the citizenry by excluding it from benefits made available to others, our jurisprudence requires us to undertake "a meaningful, case-specific analysis to ensure that [the] exclusion ... bear[s] a just and reasonable relation to the legislative goals," Baker, 170 Vt. at 204, 744 A.2d at 872.
¶ 68. Faced with the discriminatory nature of the challenged statute, with the significant evidence showing mandatory retirement laws to be an ineffective method for addressing public safety concerns, and with the obviously more direct alternative of administering performance tests to 491*491 address any such concerns, the State argued at trial that: (1) physical strength tests currently employed by the Department of Public Safety are not valid job-performance tests but rather minimal fitness tests; and (2) it would be impossible to implement viable performance tests.
¶ 69. The trial court ultimately agreed with the State, opining that the fitness tests currently in use in Vermont are not as good as plaintiffs' expert believes them to be because they require only that officers meet the level of the fiftieth percentile of a person in his or her age bracket. The court further found that there would have to be "an administrative decision" on the cutoff score for the tests if they were to be used as performance tests, and that this would run into gender issues or "political concerns" from "police unions not wanting to have such determinative testing." The court acknowledged that "[m]any state police forces do not have a mandatory age policy and [instead] use fitness and performance tests with evaluations," but speculated that a testing system could lead to legal challenges. The court also expressed a concern that developing tests "would have an expense," but conceded that the administrative costs of Vermont's current fitness system "are minimal" and that it could not "put a figure to this expense," given the testimony of a State's witness that "[h]e did not worry about the costs of such a program" because "it did not appear that the tests they developed would be particularly expensive for the department to run."
¶ 70. Several of the trial court's conclusions are based on speculation rather than evidence, and others are inconsistent with its own subsidiary findings indicating that Vermont tests could be modified into performance tests with little cost and that many other jurisdictions have successfully implemented performance tests for law enforcement agencies. Under our standard of review, "when subsidiary factual findings are inconsistent with ultimate factual findings, the ultimate factual finding may not stand." Borden v. Hofmann, 2009 VT 30, ¶ 11, 185 Vt. 486, 974 A.2d 1249.
¶ 71. Regarding the State's contention that Vermont's fitness tests are not performance tests, the Department of Public Safety's own written rules: (1) explicitly state that one of the tests' purposes is to "implement the Department's philosophy that physical fitness is vital to satisfactory job performance by ensuring each new member's effectiveness to carry out the demands of the job"; and (2) allow the Department to dismiss an officer who cannot pass the tests. Moreover, both the State's and plaintiffs' experts recognized that the Vermont fitness tests could be turned into performance tests simply by establishing a cutoff score, a fact that the court also acknowledged while expressing concerns about "political" repercussions from doing so.
¶ 72. As for the State's claim that implementing viable performance tests is not possible, the evidence unequivocally demonstrated that performance tests for law enforcement personnel are available and that implementing such tests would be neither difficult nor costly. One of the ironies of this case is that the State purports to rely on physical performance testing to demonstrate that some percentage of the population cannot perform police work at fifty-five years of age, but at the same time, argues that physical performance testing is incapable of measuring job performance. If performance tests can be used to determine that a certain percentage of fifty-five-year-olds cannot do police work, they can be used, and indeed have been used for years, to test individuals for their ability to perform police work at any age.
492*492 ¶ 73. Dr. Gebhardt herself testified that a focus of her career has been to develop physical testing to correlate with job performance, and that she has been successful in developing tests that could be used to review the performance of police officers. She also acknowledged that she, along with other members of the Landy committee, had concluded that testing is both more effective and fairer than mandatory retirement. She also agreed that mental, cognitive, and emotional factors make up eighty percent of police work, and that her study did not address these factors. Finally, she acknowledged that Vermont already had in place tests that were adequate for measuring physical performance.
¶ 74. I find no testimony to support the trial court's "Krupp" finding, cited by the majority, that Dr. Gebhardt "does not believe that there are performance tests that could adequately and safely replace the age rule." Ante, ¶ 9. Rather, Dr. Gebhardt concluded, and the trial court found, that the Vermont tests as presently designed could not be used as performance tests without a cutoff score. As the district court concluded in Gately II, "the defendants cannot argue persuasively that there is no acceptable alternative to mandatory retirement" because "the Gebhardt-Landy test itself [which was designed by Dr. Gebhardt and endorsed by Dr. Landy] precludes the defendants from proving that it is impossible or highly impractical to test each officer individually." No. Civ. A. 92-13018-MA, 1998 WL 518179, at *7 (quotations omitted). Further, in granting the original preliminary injunction on the mandatory retirement law, the district court noted in Gately I that "the most thorough and authoritative evidence presented states unequivocally that currently available tests are more effective than age in identifying officers who may be unable to perform the law enforcement and public safety tasks required of them." 811 F.Supp. at 31.
¶ 75. Nevertheless, seizing upon the trial court's stated belief that obtaining agreement on particular performance tests would be difficult because of union resistance and gender issues, the majority accepts the trial court's conclusion that performance testing would not be "at this time ... an adequate replacement for a mandatory retirement policy to meet the State's public safety objectives." Ante, ¶¶ 36-37. But, as indicated above, the majority's conclusion is plainly inconsistent with the evidence and the trial court's findings of fact. The experts for both sides agreed that the Vermont tests could be used to determine fitness for job performance as long as a cutoff score were determined. Dr. Landy testified that the current Vermont fitness tests measure virtually the same physical skills that Dr. Gebhardt measured and reported in her own study. Even assuming that the Vermont tests would not be adequate to measure job performance, both experts agreed that adequate physical performance tests have been available for years. Dr. Gebhardt, herself, has developed such tests, and the trial court found that "[m]any state police forces ... use fitness and performance tests with evaluations" as a basis for retiring officers who are physically unfit for duty. According to Dr. Landy, eleven or twelve such tests were available or in use by 1993, and that number has now risen to twenty-five or thirty.
¶ 76. Given the paucity of evidence indicating that mandatory retirement is an effective means of addressing public safety concerns, we are left with the administrative concerns that made up the bulk of the State's evidence and the trial court's findings. But neither these administrative 493*493 concerns—even if proved, which they were not—nor the trial court's speculative fears that some parties might object to implementing a system that does not discriminate based on age can possibly justify the conferring of emoluments or advantages on one set of persons while denying them to others equally qualified. These are not the kind of justifications that could possibly support a constitutional challenge to a discriminatory classification based on age. Cf. Frontiero v. Richardson, 411 U.S. 677, 690, 93 S.Ct. 1764, 36 L.Ed.2d 583 (1973) (stating that "any statutory scheme which draws a sharp line between the sexes, solely for the purpose of achieving administrative convenience, necessarily commands `dissimilar treatment for men and women who are ... similarly situated,' and therefore involves the `very kind of arbitrary legislative choice forbidden by the [Constitution]'" (citation omitted)).
¶ 77. The trial court noted that mandatory retirement is supported by the state police union and is seen as having a number of benefits, including encouraging "younger officers ... to stay on the force with expectations of openings through promotions being reasonably easy to foresee" and "keeping new ideas coming forward with new personnel." According to the court, "the 55 year figure offers young recruits the advantage of not having to work well past that time to get good retirement benefits. Many can retire even earlier if they started in their twenties or early thirties." In the same vein, the court cited both the belief of former commissioners that mandatory retirement keeps the force "viable" and "energized" and the belief of the president of the state police union that mandatory retirement coupled with allowing full retirement after twenty years was "good for recruiting."
¶ 78. Even a cursory examination of the transcripts and the trial court's decision reveals that these secondary administrative concerns, not public safety, are the 494*494 real reasons underlying the mandatory retirement law, or at least why, in the face of overwhelming evidence that such laws fail to promote public safety, they continue to exist. I would vote to uphold the law if there was any evidence that it actually promoted public safety, but neither the record in this case, nor the history of mandatory retirement for law enforcement officers generally, supports the alleged public safety rationale. Indeed, it is difficult to imagine what could be less effective at protecting the public than a system that consistently removes the most experienced 495*495 police officers, while doing nothing to ensure the fitness of officers below the retirement age.
¶ 79. Not only is the challenged law significantly overinclusive, in that seventy-five percent of the officers who reach the mandatory retirement age would be able to pass physical performance tests, but, at the same time, it is underinclusive in that it permits officers below the mandatory retirement age who cannot pass minimum fitness tests to continue to work. In view of the fact that performance tests exist that could address public safety concerns, the State's use of mandatory retirement cannot be sustained as a valid substitute for individual determinations based on such tests. I am not advocating that the Court adopt a better or different approach than the Legislature itself has taken to implement a state policy goal, but I am arguing that this Court should strike a discriminatory classification that cannot logically be sustained as implementing a valid governmental purpose.
¶ 80. The majority here has taken the Baker case and made it unrecognizable. It purports to apply the Baker analysis, but in fact has undercut Baker's attempt to provide a substitute for traditional equal protection analysis in cases involving important personal and civil rights. Baker may have eliminated the multi-tiered mode of analysis, but it surely did not intend that such rights could be denied upon the mere showing that there is some conceivable fact relating the denial to a governmental purpose.
¶ 81. The majority does not say so explicitly, but it has applied the rational-basis test in this case—a test that heretofore has been employed only where commercial or other lesser rights have been implicated. More specifically, while acknowledging that Baker requires an inquiry into whether a challenged discriminatory governmental classification promotes a legitimate public goal in a manner that is neither significantly underinclusive nor overinclusive, the majority applies a test requiring no judicial inquiry beyond determining whether the reasons underlying the State's action are wholly fictitious. Relying primarily on a thirty-year-old case that imposed a rational-basis analysis in upholding a 1946 law intended to "`insure the rapid advancement of men of ability to positions of responsibility and the elimination of men who have reached their ceilings of performance,'" Vance, 440 U.S. at 99, 99 S.Ct. 939 (quoting Congressional Record), the majority's decision is in a time warp unaffected by our Baker decision or contemporary societal values that underlie the ADEA and other federal and state laws.
¶ 82. The mandatory retirement of fit, experienced, and capable state police officers, solely because of their age, is plainly unconstitutional under settled precedent of this Court. The law violates the guiding principle of the Common Benefits Clause to safeguard the rights and liberties of all Vermonters, and, in the end, it actually undermines public safety—the purported purpose of the law—because it prevents the most experienced officers from serving the public, while doing nothing to ensure that unfit officers are removed from the force. Equally unfortunate, from the point of view of this Court as an institution, today's decision makes our Common Benefits jurisprudence seem, not rigorous and principled, as it ought to be, but merely idiosyncratic.
 Section 459(a)(2) provides that any member of Group C of the Vermont Employees Retirement System "shall be retired on a normal retirement allowance on the first day of the calendar month next following attainment of age 55." Group C of the Retirement System consists of "any regular officer or employee of the department of public safety assigned to police and law enforcement duties." 3 V.S.A. § 455(a)(11)(C), (9)(B).
 Plaintiffs also appeal the trial court's grant of defendants' motion for judgment on the pleadings, arguing that if there is a constitutional violation, they are entitled to damages, either under the Vermont Fair Employment Practices Act—after the exception for law enforcement officers in 21 V.S.A. § 495f is eliminated as unconstitutional—or because the Court should recognize a direct cause of action for damages. As we hold herein that there is no violation of the Common Benefits Clause, we need not address these arguments.
 The United States District Court for the District of Vermont has ruled that Murgia governs an equal protection challenge to the Vermont statute and ruled, therefore, that the mandatory retirement requirement does not offend the Equal Protection Clause. See Galvin v. Vermont, 598 F.Supp. 144, 151 (D.Vt. 1984). Plaintiffs do not attempt to challenge the conclusion of that court, as they do not raise a federal equal protection challenge here.
 There are a number of decisions addressing whether age is a bona fide occupational requirement for law enforcement officers, and most have found that it is not. Compare Binker v. Pennsylvania, 977 F.2d 738, 742 (3d Cir. 1992) (noting that in previous decision, court "concluded that the record did not establish age as a bona fide occupational qualification" for state police officers in Pennsylvania); EEOC v. Ky. State Police Dep't, 860 F.2d 665, 667 (6th Cir. 1988) (age is not a bona fide occupational qualification for state police officers); with EEOC v. Mo. State Highway Patrol, 748 F.2d 447, 457 (8th Cir. 1984) (age is a bona fide occupational qualification for state highway patrol officers); EEOC v. New Jersey, 631 F.Supp. 1506, 1515 (D.N.J. 1986) (age is bona fide occupational qualification for state police). Because a bona fide occupational qualification must be "reasonably necessary" to the normal operation of the employer, see 29 U.S.C. § 623(f)(1), and because those words appear in Baker v. State, 170 Vt. 194, 214, 744 A.2d 864, 878 (1999), the Common Benefits Clause case on which they primarily rely, plaintiffs urge us to adopt the ADEA decisions favorable to them and rely upon them as precedent.
 The chronology of events with respect to the ADEA and law enforcement officers is provided in an appendix to Correa-Ruiz v. Fortuno, 573 F.3d 1, 16-17 (1st Cir.2009).
 As the dissent points out, post ¶ 61, plaintiffs' expert witness was the principal author of a congressionally-sponsored study on the alternatives to mandatory retirement, but the conclusions and recommendations of the study were not accepted by Congress.
 The trial court partially based its decision on the governmental purpose of encouraging recruitment because younger officers can better anticipate possible advancement when it is known that superiors will retire at age fifty-five, which is a particular concern because of the few supervisory positions in the state's small police force. This purpose was supported by the testimony of defendants' witnesses. We have not considered this purpose in reaching our decision.
 The dissent's description of the State's interest as a "pretext," see post, ¶ 46, is hyperbolic and an indication that the dissent has gone beyond an appellate role under a limited standard of review and into that of a trial judge. The trial court focused on the public safety purpose of the law in making its findings and conclusions. The dissent's conclusion that the State's purpose does not justify the age discrimination it finds is present does not make the justification a "pretext."
 In Baker, Justice Dooley wrote: "It is ironic that in a civil rights case we overrule our precedent requiring the State to meet a higher burden in civil rights cases, but still conclude, under the lower standard, that the State has not met its burden." Baker, 170 Vt. at 236, 744 A.2d at 894 (Dooley, J., concurring). Here, the irony is that, in a civil rights case, the majority purports to apply the "more stringent" reasonableness inquiry called for by Baker and its antecedents, yet in fact applies the less stringent "rational basis" analysis traditionally associated with the review of economic regulation.
Indeed, in response to the dissent, the majority explicitly embraces the federal "rational basis" standard of review, stating that "we will follow the Clover Leaf Creamery approach and stay our hand." Ante, ¶ 41. As the majority points out, in Minnesota v. Clover Leaf Creamery Co., 449 U.S. 456, 101 S.Ct. 715, 66 L.Ed.2d 659 (1981), milk sellers challenged a statute that banned the sale of milk in nonreturnable, nonrefillable plastic containers but not in nonreturnable, nonrefillable paperboard containers. The Court and the parties agreed that the applicable standard of review concerning the economic legislation was "the familiar `rational basis' test" and that the legislature's purposes in enacting the law— promoting resource conservation, easing solid waste disposal problems, and conserving energy—were legitimate state purposes. Id. at 461-62, 101 S.Ct. 715. The narrow issue was whether the state legislature's economic classification distinguishing between plastic and nonplastic nonreturnable milk containers was rationally related to achievement of the statutory purposes. Id. at 463, 101 S.Ct. 715. Citing the district court's candid admission "that the evidence was in `sharp conflict,'" id. at 464, 101 S.Ct. 715, the Court concluded that the federal Equal Protection Clause did not require the state legislature to "strike at all evils at the same time or in the same way," id. at 466, 101 S.Ct. 715 (quotation omitted), and thus held that the challenged statute bore a rational relation to the state's environmental protection objectives, id. at 470, 101 S.Ct. 715.
This is the "approach" that the majority announces it will apply in the instant case— even though the Court in Clover Leaf Creamery explicitly acknowledged that "[a] state court may, of course, apply a more stringent standard of review as a matter of state law under the State's equivalent to the Equal Protection or Due Process Clauses," id. at 461 n. 6, 101 S.Ct. 715, and this Court did exactly that less than a decade ago in Baker. Apparently, it matters neither that we explicitly rejected the less rigorous federal rational-basis test in Baker nor that the instant case involves claims of age discrimination rather than mere economic classification. It is disconcerting to see how quickly and completely this Court has backtracked from the standard of review announced in Baker.
 Congress commissioned the study in connection with its temporary exemption for law enforcement agencies from the ADEA's ban on state and local mandatory retirement laws. Congress enacted the ADEA in 1967 and amended it in 1974 to cover federal, state, and local governments. M. Schiff, The Age Discrimination In Employment Act: Whither the Bona Fide Occupational Qualification and Law Enforcement Exemptions?, 67 St. John's L.Rev. 13, 14 (1993). The law was supported by undisputed empirical evidence that "the process of psychological and physiological degeneration caused by aging varies with each individual," that chronological age alone is a poor indicator of ability to perform a job, that mandatory retirement does not take into consideration the differing abilities of people to do a job, and that, despite these "well-established medical facts," there continued to exist persistent and widespread use of age limits in employment decisions attributable only to arbitrary discrimination. W. Air Lines, Inc. v. Criswell, 472 U.S. 400, 409-11, 105 S.Ct. 2743, 86 L.Ed.2d 321 (1985). Aware of the inconsistent assumptions underlying the ADEA and its various mandatory retirement laws, Congress amended the ADEA in 1986 to prohibit mandatory retirement for employees based on chronological age, but granted state and local governments a temporary exemption for public safety officers in response to claims that a law enforcement exemption was needed to protect the public. Schiff, supra, at 14. Congress granted the temporary exemption for seven years, from January 1987 through December 1993, and tasked the Secretary of Labor and the EEOC to study and determine whether elimination of chronological age as a basis for hiring or retiring police officers would have any impact on public safety. Id. at 15. Congress also directed the EEOC to determine whether testing could measure the ability of law enforcement officers to do their job and, if so, to make recommendations for specific standards for testing. Id. at 15-16.
 F. Landy, Alternatives to Chronological Age in Determining Standards of Suitability for Public Safety Jobs (January 31, 1992).
 Despite the Landy report's conclusions and recommendation that law enforcement agencies not be exempted from the ADEA's ban on mandatory retirement, the EEOC did not propose testing guidelines. A political stalemate ensued, and the exemption from the ADEA's ban on mandatory retirement was extended indefinitely for state and local governments like Vermont which had enacted mandatory retirement for public safety officers before March 1983. Over the years law enforcement agencies have "lobbied Congress with position papers that relied largely on non-empirical, anecdotal data to support the law enforcement exemption" from the ban on mandatory retirement in the ADEA. Schiff, supra, at 45. The law enforcement position has historically been based on nothing more than "a stereotypical feeling that older police officers will experience more physical and medical problems, more injuries, and more lost time to the police department and, consequently, will be less able than younger police officers to protect the public." Id.
Contrary to the majority's assertion, there is no ongoing "national debate" as to the science of whether performance tests are more effective and appropriate in protecting the public than the mandatory retirement of public safety officers in their fifties. Ante, ¶ 40. To the extent that any such debate ever took place, mandatory retirement has been thoroughly discredited as a public safety strategy. Nonetheless, the majority suggests that there is an "active investigation" of underlying facts "in flux," and that this Court's interference with the "ongoing political process" will irrevocably hamstring the Legislature into making choices not yet thoroughly explored. Ante, ¶¶ 40-41. The majority likens the situation to a court's interference with policies regarding assisted suicide. Ante, ¶ 41. The situation here is completely inapposite. We are not dealing with an issue, such as assisted suicide, that has moral repercussions for many people. That may have been true in Baker, but even still we adopted a more rigorous standard that should have been applied in this case. Nor does this case involve an active investigation of underlying facts in flux. Rather than an active political debate over the underlying science, there is merely a political quagmire stemming from interests unrelated to public safety. The only thing "in flux" is our discarded Baker standard. By abandoning that standard and approving, without rigorous review, a discriminatory policy based on unsupported claims of protecting public safety, we are abdicating our constitutional judicial role in this case.
 These findings mirror the bulk of the State's evidence, which was filled with the kind of stereotypical justifications for age limitations that do not meet any test of constitutional sufficiency. For example, the State's first witness, former commissioner of the Department of Public Safety, Kerry Sleeper, testified that requiring individuals to retire at a set age "allows management and certainly the Commissioner of Public Safety to strategically manage the organization so that it can more effectively meet the needs of the public." He stressed that managers responsible for budgeting their resources need to know how long a promoted officer will serve so that that information can be built into the budget. He stated that, like the rest of "corporate America," the Department needed to be "mean and lean" and reduce "the likelihood of people being entrenched in a job that they really don't want" to do but continue to do because they are waiting for retirement. Similarly, another former Commissioner, James Walton, testified that mandatory retirement for state troopers is good because the job "requires energy, not only physical energy, but emotional energy, intellectual energy." He elaborated that "the one thing you don't want in a police agency is stagnation. You don't want a silting-in effect at the top of the organization, for that stops movement all the way up through the organization." As he explained, you want "an organization that brings energetic people in at the bottom and maintains that, as much of that energy as possible through their careers, and in a timely way, so to speak, as harsh as it may sound, move people out so that other people can move into the positions of leadership and service in the department." The director of recruiting for state police and the president of the Vermont state police union testified, among other things, that the principal draw for potential recruits was to be able to obtain full retirement after twenty years of service, and that the union was opposed to abolishing mandatory retirement because of its desire to maintain and increase state police benefits and of its belief that "our benefits were based on the fact that this was a job or a profession that was intended for younger people." The State offered no empirical evidence, however, to support even these secondary justifications for the mandatory retirement law.
 The lack of a connection between public safety and the chronological age of law enforcement officers is revealed through an examination of the history of mandatory retirement laws, which date back to 1947, when Congress passed legislation "permitting investigatory personnel of the Federal Bureau of Investigation to retire at age 50 at an enhanced annuity." Johnson v. Mayor & City Council of Baltimore, 472 U.S. 353, 364, 105 S.Ct. 2717, 86 L.Ed.2d 286 (1985) (citing Act of July 11, 1947, ch. 219, 61 Stat. 307). Designed to stimulate the morale of FBI officers and to provide an incentive for them to remain in the service for a reasonable period of time, the law "was intended only to give certain employees the option to retire early." Id. The Attorney General at the time explained that "the Department of Justice sought to maintain the FBI `as a young man's service'" and to assure that "`men in their 60's and 70's ... faced with the rigors of arduous service demanded of special agents'" would not be "`forced to carry on for lack of an adequate retirement plan.'" Id. (citations omitted). Soon thereafter, Congress extended this program to other employees who lobbied for the benefits of an early retirement system. Id.
It was not until 1974 that Congress changed the law to require public safety personnel to retire at age fifty-five if they had completed twenty years of service. Id.; Schiff, supra, at 20 (noting that, by 1978, the optional retirement program legislated by Congress in 1947 to reward thirty-six FBI agents who were subject to hazardous duty had evolved into a mandatory retirement program and age-based system covering 52,000 employees). In so doing, "Congress undoubtedly sought in significant part to maintain a youthful work force and took steps through the civil service retirement provisions to make early retirement both attractive and financially rewarding." Johnson, 472 U.S. at 365, 105 S.Ct. 2717; Schiff, supra, at 20 (noting legislative history that emphasized maintaining "a young and vigorous work force"). But nothing in either "the language of the 1974 amendment nor its legislative history offered any indication why Congress wanted to maintain the image of a `young man's service,' or why Congress thought that 55 was the proper cutoff age, or whether Congress believed that older employees in fact could not meet the demands of these occupations." Johnson, 472 U.S. at 365, 105 S.Ct. 2717. Thus, the history of the civil service provision mandating early retirement for specified public safety officers "makes clear" that it was not based on a bona fide occupational qualification, to use the phraseology of the ADEA. Id. at 363, 105 S.Ct. 2717. As the United States Supreme Court stated, the "absence of any indication that Congress established the age limit based on the demands of the occupation" suggests that the federal mandatory retirement law was the result of "age stereotyping" rather than establishing legitimate actual occupational qualifications. Id. at 366, 105 S.Ct. 2717 (quotation omitted).
Vermont's early retirement law for state troopers followed a similar course. The law initially provided that certain personnel "may be retired" with full benefits at the age of fifty-five. 1971, No. 231 (Adj.Sess.), § 4. In 1977, however, the law was amended to provide that those personnel "shall be retired" at age fifty-five. 1977, No. 80, § 1. Legislative history is sparse from that time period, but we may presume that the forces driving the Vermont law were the same as those driving the federal law. Indeed, the state police union was the principal force pushing for a mandatory retirement exemption to Vermont's employment discrimination law in 1981, and in the instant case the president of the Vermont state police union testified that the union opposed plaintiffs' efforts to challenge mandatory retirement because of its belief that doing so would undermine the union's goal of "maintaining and increasing the benefits that we had." This history confirms that our mandatory retirement law, like the federal law, emerged out of concerns unrelated to public safety.