I respectfully dissent.
I must confess, when I first started researching this case, I did not expect to find merit in the contention a cook could be “engaged in active law enforcement.” However, as I dug deeper and as I learned more about what this particular cook did and the conditions under which he had to operate, I became convinced.
Before discussing where and how I differ with the majority opinion, I want to set forth the points of agreement. To begin with, I concede that if Glover is not entitled to the presumption granted those engaged in “active law enforcement” then the majority is correct in finding there is substantial evidence to support the denial of Glover’s service-connected pension.1 Con*1341versely, I believe it is fair to imply from the majority opinion my colleagues agree that if Glover indeed were entitled to the presumption then the evidence in the record would not support a denial of the pension.
To set the latter issue to rest, however, we have only to drop this particular, unusually robust presumption into the pool of evidence which moved the pension board and the trial court to deny Glover his pension. The first effect of the presumption, of course, is to shift the burden of proof from Glover to the board. No longer does Glover carry the burden of proving his heart condition was attributable to his employment. Instead the burden is placed on the pension board to prove Glover’s employment did not contribute substantially to his heart trouble.
But the “active law enforcement” presumption goes far beyond merely shifting the burden of proof from pension applicant to pension board. It creates a conclusive presumption as to the critical issue in dispute in this case—whether Glover’s heart condition derived from his employment or from other diseases unrelated to his employment. Government Code section 31720.52 states in part, “Such heart trouble so developing or manifesting itself in such cases shall in no case be attributed to any disease existing prior to such development or manifestation.” In other words, where this presumption applies the pension board is foreclosed from attributing Glover’s heart trouble to any disease existing prior to his heart attack.
*1342In finding the petitioner to be disabled as a result of nonservice-related causes, both the board and the referee relied exclusively upon the medical reports of Drs. Trostler and Wanamaker who referred to the petitioner’s “major risk factors for hypertension and coronary artery disease.” These risk factors included hypertension which was diagnosed five years prior to the disability, hyperlipoproteinuria, and cigarette smoking. They were deemed to be the major contributing factors to the petitioner’s coronary disease. Indeed the pension board relied solely upon these preexisting diseases to deny Glover his service-connected benefits. This they could not do, assuming Glover was entitled to the “active law enforcement” presumption. For, the presumption instructs the pension board it shall in no case attribute “[s]uch heart trouble ... to any [preexisting] disease. . . .” (§ 31720.5, italics added.)
I am led to conclude it would be impossible for any appellate court that had found Glover eligible for the “active law enforcement” presumption to nevertheless uphold the trial court’s ruling in this case. Once this presumption enters the equation, it is apparent the pension board failed to meet its burden of proving Glover’s heart trouble did not “arise out of and in the course of employment.” Under section 31720.5, the petitioner’s coronary disability cannot be attributed to the petitioner’s preexisting diseases. Yet these preexisting conditions comprised the sole proof offered to establish Glover’s disability was not service-connected. Accordingly, on the basis of evidence before the board in this case, it would have been compelled by the terms of section 31720.5 to have granted a service-connected pension to anyone who qualified as a member “engaged in active law enforcement.”
I do not read the majority opinion as quarreling with this conclusion. Rather we differ only over the issue whether Glover indeed was a member “engaged in active law enforcement.” Having established how and why this issue is critical to the outcome in this case, I now return to explain why I, unlike my colleagues, have become convinced this prison cook was eligible for the “active law enforcement” presumption.
The presumption applies to a member engaged in active law enforcement who is not a safety member. The majority opinion mentions two principal cases interpreting the phrase “active law enforcement,” Ames v. Board of Retirement (1983) 147 Cal.App.3d 906 [195 Cal.Rptr. 453] and Neeley v. Board of Retirement (1974) 36 Cal.App.3d 815 [111 Cal.Rptr. 841]. In Ames, a county correctional officer whose position was not classified as a safety member for retirement purposes requested a writ of administrative mandate to be classified as a safety member. Relying on Neeley v. Board of Retirement, supra, 36 Cal.App.3d 815, the Fifth District reversed the lower court and held that Ames was engaged in active law enforcement because *1343he satisfied three criteria: (1) he had daily contact with prisoners, (2) he was exposed to hazards resulting from prisoner conduct, and, (3) he was exposed to risk of injury from the necessity of being able to cope with potential dangers inherent in the handling of prisoners.
The majority applies this three-pronged test to exclude Glover from the category of members “engaged in active law enforcement.” I apply the same test but reach the opposite conclusion.
The majority takes the analysis one level deeper, however. They point to Ames along with its kindred case, Kimball v. County of Santa Clara (1972) 24 Cal.App.3d 780 [101 Cal.Rptr. 353], and emphasize both granted “active law enforcement” status to corrections officers while Neeley denied such status to a criminal identification technician. The majority gleans from these cases a single overriding criterion—“the extent to which the [job] category exposes its holders to potentially hazardous activity.” The majority finds a prison cook to be like an identification technician and not like a corrections officer. I, on the other hand, applying the majority’s own measure of “potential hazard” find this prison cook’s duties, while obviously not identical to those of a corrections officer, to be much closer to those of a corrections officer than an identification technician who, after all, spends his days in comparative safety, far removed from criminals.
Glover was at various times a supervising cook and a head cook whose primary official duties were culinary and administrative. But whether an employee’s service places him in active law enforcement is determined by the “service which is actually performed by an employee and not necessarily only that which is set out in his job description.” (Kimball v. County of Santa Clara, supra, 24 Cal.App.3d 780, 785.) The record demonstrates Glover’s position in reality required him to have daily contact with prisoners and to supervise prisoners working in his kitchen. These duties, in turn, exposed him to potential hazard on a daily basis.
The statement of facts in the majority opinion is replete with evidence all three Neeley factors were present. (Indeed it is difficult to square that statement of facts with the legal conclusions the majority reaches later in the opinion.) But it may be useful to review the full record and what it says about Glover’s actual duties in the light of those factors and then in the light of the majority’s own single-factor test.
The first prong of the Neeley test requires that the job entail “daily contact” with criminals or prisoners. The county’s own job analysis specified Glover was “in daily contact with inmates, ...” Elsewhere this job analysis indicates: “The Head Cook supervises food preparation and serving *1344of over 5000 meals a day. He is continuously training staff and inmates in this field.” (Italics added.) Moreover, the record demonstrates Glover’s kitchen crew consisted entirely or primarily of inmates at the various jails and camps in which he served. Thus, there is no disputing this first prong is satisfied.
The second prong of the Neeley test requires the employee to be exposed to hazards resulting from prisoner misconduct. Once again, the county’s own job analysis points to “[pjroblems with inmate help.” Furthermore, Glover testified to the occurrence of frequent fights among inmates. Additionally, the reports filed by all three doctors are replete with discussions of fighting incidents which Glover had to break up. One reason for the frequency of these fights was the nature of the Hall of Justice jail, a minimum-maximum security facility. Although the appellant’s crews were minimum security inmates, the proximity of minimum and maximum security prisoners presented potential problems. Glover testified this combination was more troublesome since the inmates “did not have enough outlet.”
Moreover, this prison cook’s risk from prisoner misbehavior was doubled by the fact these were not unarmed prisoners. They had sharp kitchen knives up to 12 inches in length; they had meat cleavers; and they had heavy 5-foot wood and metal pallets. The majority emphasizes these potential weapons were kept under lock and key and were issued to the inmates by a deputy. Still, there is no denying the prisoners had these dangerous objects in their possession while on duty preparing food. I suspect it would be small consolation to most people that the inmates working beside and behind them had to account for their sharp 6- and 12-inch knives, their meat cleavers and heavy pallets at the end of each shift. This precaution may have protected the rest of the prison from an influx of these lethal weapons. It did little to protect Glover and his fellow cooks from exposure to the danger these weapons posed in the hands of the prison labor assigned to his kitchen.
The third prong of the Neeley test is that the employee be exposed to a risk of injury resulting from the necessity of handling prisoners. While he was head cook at the Hall of Justice jail Glover was responsible for any emergency arising at the culinary unit. At the other jail facilities in which he served, Glover had similar responsibilities for any fights or other emergencies which might break out while he was on duty supervising the kitchen. According to Glover’s testimony, although he himself was never assaulted by an inmate, he had disarmed prisoners who had seized kitchen knives.
The county’s own job analysis of Glover’s position emphasizes Glover “is locked in with inmates that he instructs and works with while preparing *1345food for 1600 inmates.” (Italics added.) The situation the job analysis describes hardly reflects those “unusual incidents or conditions which are typically of relatively short, intense duration” which have been found insufficient to qualify a member as one engaged in “active law enforcement.”3 Instead, a prison facility is fertile ground for tensions and resentment towards those similarly confined as well as others. Although at times there were deputy sheriffs available for assistance they were unarmed and thus of not much help in case an inmate suddenly attacked Glover with a 12-inch kitchen knife or wood and metal pallet. In fact, at the time of his heart attack, there was no deputy sheriff on duty in the kitchen area. Glover had to be escorted outside to a safe zone by a prison trusty.
It is true Glover’s principal duties were not precisely those of a correctional officer. Yet he had similar responsibilities to “handle” the 10 to 40 inmates assigned to him each shift. He was similarly exposed to potential injury in carrying out these supervisorial responsibilities. He worked in close proximity with inmates on a daily basis and was responsible for containing any confrontation or dispute that might arise. Moreover, these inmates were armed with deadly knives and blunt instruments unlike the unarmed inmates with whom correctional officers typically deal. As the majority’s own statement of facts concedes, although Glover himself was never assaulted by an inmate, he had broken up many fights, had disarmed prisoners, and had to “play both sides” to protect his own safety. There is nothing in the Neeley test to suggest a corrections officer or other employee must actually have been personally assaulted to be entitled to the “active law enforcement” presumption. It is exposure to the risk of injury at the hands of suspects or prisoners, not the actual suffering of such injuries, which generates the stress and entitles one to the benefits of this presumption.
It is not critical to my analysis, but I find it instructive to note the Fifth District has granted “active law enforcement” status to a category of employee whose duties are nearly identical to those of a prison cook—the prison work program supervisor. (Barrett v. Stanislaus County Employees Retirement Assn. (1987) 189 Cal.App.3d 1593, 1599 [234 Cal.Rptr. 900]) These work program supervisors happened to be employed at an “honor” farm. Even though this is a minimum security facility, the court had no trouble ruling they were entitled to be reclassified as “safety members” of the pension system, if they were willing to pay in any accrued pension contributions. Unlike traditional corrections officers, these employees do not guard prisoners, as such. Instead, like Glover and other prison cooks, they supervise prisoners in the performance of certain work functions. The prisoners might be turning big rocks into little stones, raw steel into license *1346plates, or performing similar productive activities. Those supervising them must “handle” these prisoner work crews and risk harm at their hands for the same reasons and to the same degree as Glover did while supervising the inmate crews assigned to produce food in his kitchen. I see no reason to deny prison cooks who are asked to perform their cooking responsibilities with prisoner crews the same “active law enforcement” status accorded other prison work program supervisors.
I conclude this opinion by applying the “single prong” test the majority distilled from Ames, Neeley, and the underlying statutes. That test, as will be recalled, asks “the extent to which the [job] category exposes its holders to potentially hazardous activity.” I agree with my colleagues this is the proper standard. It is the potential danger from working with criminals or crime that is the touchstone, not the particular way in which the danger arises. Nor is it required the potential danger have become actual. Rather it is the seriousness of the hazard and the constancy of the risk.
To my mind, the majority misapplies its own test when it emphasizes Glover was never actually assaulted and only occasionally had to break up fights between inmates. The risk of assault was always there and indeed assault by a very deadly weapon, not merely fists. Similarly, fights between inmates requiring Glover’s intervention may have been only occasional happenings. But the risk was constant. So was the risk a fight between two prisoners might escalate into a general melee or that one of the combatants might turn on the intermeddler, Glover. Therefore, section 31470.3 disallowing “active law enforcement” status to those, like deputies serving as clerks and stenographers, who only occasionally are asked to expose themselves to the risk of dangerous activities has no application to this case. Here, as is almost always the case with those engaged in active law enforcement, the risk was constant but the dangerous incidents rather infrequent. The clerk or stenographer, on the other hand, works in a nearly risk-free environment over 99 percent of the time. Only when an emergency arises is there a possibility he may be drafted into a temporary assignment which exposes him for that brief period of time to a risk of harm from criminal activity.
In applying the majority’s test, I ask whether a reasonable person would feel “exposed to potentially hazardous activity” were he or she locked up most of the day giving orders to 5 or 10 or as many as 40 criminals who were armed with sharp knives, meat cleavers and pallets, and now and again erupted in violent fights. The answer I arrive at is that anyone in Glover’s job would have to be ««reasonably oblivious or foolhardy to feel otherwise. The trial judge himself conceded, “I’ve been in the jail, and I got to tell you just walking in there is stressful.” If the judge feels that way *1347when steel bars separate him from unarmed inmates how much greater is the exposure to potentially hazardous activity when one is on the other side of the bars, locked in with the prisoners and, worse yet, when those prisoners are armed with lethal weapons.
I do not wish to be misunderstood as arguing for the proposition every cook is “engaged in active law enforcement.” Accurately stated, Glover’s “job category” was not merely “a cook” or “a head cook.” The term “prison cook” captures more of the reality of Glover’s employment, but still is not sufficient in itself to place him in “active law enforcement.” What exposed Glover to hazardous duty was the fact we have chosen to operate our jail kitchens with inmate labor rather than paid civilian staff. A fully descriptive title for his position would be something along the lines of “cook and prison inmate supervisor.”
If and when we were to shift to full-time paid kitchen staffs in our jails, the cooks and head cooks in those jails would no longer be “engaged in active law enforcement,” at least as I would construe the term. But in the meantime, I would regard it as fundamentally unfair for the public to save money by thrusting these cooks into an environment which exposes them to potentially hazardous activity at the hands of prisoners, yet deny them the benefit of the heart attack presumption which this environment justifies. In my view, for reasons discussed in this opinion, the public—as represented by our Legislature—intends no such result. Properly construed, section 31720.5 was designed to extend the heart attack presumption to all those, including prison cooks, who as part of their public employment are exposed on a regular basis to potential danger from criminal suspects and imprisoned criminals.
In furtherance of section 31720.5 and the public policy it implements, I would reverse the trial court and order issuance of a service-connected pension to the late Mr. Glover.
Appellant’s petition for review by the Supreme Court was denied January 18, 1990. Kennard, J., did not participate therein.
I find this a closer question than does the majority, however. The evidence supporting the pension board’s finding that Glover’s employment did not contribute substantially to his heart problems was marginal, at best. It consisted essentially of two (out of three) physicians *1341who opined Glover’s job as a prison cook did not cause him significant stress. The doctors had before them information Glover’s working days were filled with situations calculated to produce stress in any ordinary person—locked in a kitchen with inmate workers who were armed with sharp kitchen knives and meat cleavers and who occasionally started fights he had to break up. Yet these physicians accepted at face value Glover’s statements he was happy in his job and not particularly bothered by the fact he had to supervise a crew of armed inmates every day. One does not have to be an expert in psychology to recognize some people are unwilling to admit even to themselves the stress they are experiencing. Glover’s acknowledged chain-smoking may have been a more accurate indicator of what his job was doing to his psyche—and to his heart—than any verbal reassurances he gave to the doctors who examined him.
Had I been the trial judge in this case I would have applied the “independent judgment” test to reverse the pension board’s denial of Glover’s claim even without the benefit of the “active law enforcement” presumption. For reasons suggested above, I would have given more credence to the expert opinion linking Glover’s job to his heart condition than I would have given to the contrary opinion. However, our scope of review on appeal is quite different. It is not for us to apply our “independent judgment” to the record before the trial court. As the majority opinion emphasizes, our only role is to determine whether there was “substantial evidence” to support the trial court’s exercise of that judgment. The evidence here was weak, but nevertheless reached the minimal threshold of being “substantial.” On that basis— and that basis only—I subscribe to the majority’s finding that in the absence of the “active law enforcement” presumption substantial evidence supports the denial of Glover’s service-connected pension.
All further statutory references are to the Government Code unless specified otherwise.
Section 31470.3 and discussion at pages 1346-1347,post.