[*PG383]BRUM V. TOWN OF DARTMOUTH AND THE PUBLIC DUTY RULE: NAVIGATING AN INTERPRETIVE QUAGMIRE
Abstract: In 1993, a group of youths entered the Dartmouth High School and stabbed sixteen-year old Jason Robinson to death in his social studies classroom. In 1999, in Brum v. Town of Dartmouth, the Supreme Judicial Court of Massachusetts held that the town was immune from suit pursuant to Massachusetts’ statutory “public duty rule,” which insulates public employers from liability where the employer does not “originally cause” the harm. This Article traces the evolution of public tort liability in Massachusetts, suggests a three-part framework for interpreting Massachusetts’ public duty rule and proposes a narrowly-tailored exception to the rule in cases like Brum.
“Nice, huh?” a state police detective said [that] afternoon. “You send your kid to school and he gets killed in homeroom.”1
Dartmouth, Massachusetts is a small town nestled along the rugged coast of southeastern Massachusetts, composed of neat homes with weather-worn shingles and manicured lawns dotting its waterfront.2 On Monday, April 12, 1993, Dartmouth, which had grown well-accustomed to the constant and familiar battery of its shores by the cold waters of the Atlantic, was newly rocked by the unfamiliar and formidable tide of violence.3 At approximately 8:30 a.m., a group of youths entered Dartmouth High School wielding knives, a billy club, a baseball bat and a length of pipe, and stabbed sixteen-year old Jason Robinson to death in his social studies classroom.4
On that day, the people of Dartmouth picked up the newspaper and read the headlines; they turned on the TV and heard the news; they called their neighbors and spoke with their coworkers and [*PG384]awaited the return of their sons and daughters, all of whom tried to explain the day’s events. The people of Dartmouth thought about the young man, Jason Robinson, and of their town. “How could this happen here?”
On that day, the school teachers—who had been teaching at Dartmouth High School for ten and twenty and thirty years, who had witnessed their fair share of fisticuffs among students and who had perhaps briefly glimpsed the attackers parading through the hallways that they knew so well—listened to each other’s stories detailing the terrible commotion that had taken place. The school teachers at Dartmouth High School thought about the young student, Jason Robinson, and about their workplace. “How could this happen here?”
On that day, the students—who saw the young man falter after sustaining what looked like a mere punch to the stomach, who helped the young man to his feet in an attempt to bring him to the nurse’s office and who shook their heads in disbelief upon hearing of the student’s death that afternoon—wrote messages in memoriam on large banners draping the school’s walls, while stepping over the forensic markings covering the school’s floors. The students at Dartmouth High School thought about their peer, Jason Robinson, and about their school. “How could this happen here?”
Following her son’s death, Jason Robinson’s mother sued the town of Dartmouth and several school officials for negligence pursuant to the Massachusetts Tort Claims Act (“MTCA”), which makes public employers liable for the torts of their employees.5 The plaintiff argued that the school officials had breached an affirmative duty to protect her son from harm and that the school officials’ employer, the town of Dartmouth, should be held liable for his death.6 According to the plaintiff, the school officials failed to implement adequate security measures as required by state law, and furthermore, when school officials were explicitly forewarned of the attackers’ arrival at the school, the school officials failed to prevent them from entering the [*PG385]school and proceeding into the social studies classroom.7 If the school “had taken adequate safety measures,” the plaintiff argued, “the deceased’s attackers could not have entered the high school, and the murder would never have taken place in the classroom.”8
In 1995, the Massachusetts Superior Court granted a motion to dismiss on grounds that the school was immune from liability under the MTCA because the school’s failure to adopt security measures was a discretionary function immunized under chapter 258, section 10(b) of the General Laws of the Commonwealth of Massachusetts (“§ 10(b)”), and because the school’s failure to act fell within the statutory public duty rule under chapter 258, section 10(j) of the General Laws of the Commonwealth of Massachusetts (“§ 10(j)”).9 In [*PG386]1998, in Brum v. Town of Dartmouth (“Brum I”), the Massachusetts Appeals Court reversed the superior court’s decision and held the town of Dartmouth liable for the school’s failure to institute security measures and failure to respond to a foreseeable threat of harm.10 In 1999, however, in Brum v. Town of Dartmouth (“Brum II”), the Supreme Judicial Court reversed the appeals court and held that the school and school officials were immune from suit.11
This Article traces the evolution of Massachusetts public duty jurisprudence from its common law origins to its present-day codification in Massachusetts law, with particular focus on the Supreme Judicial Court’s decision in Brum II.12 Because the history of public tort liability is based upon judicial precedent as well as state statutory law, Part I of this Article reviews the evolution of public tort liability in Massachusetts prior to 1994, in light of both the Supreme Judicial Court’s and the Legislature’s attempts to expand and limit public tort liability.13 Examination of this push-and-pull pattern includes the court’s abrogation of the sovereign immunity doctrine, the Legislature’s establishment of public tort liability through the MTCA, the court’s adoption of the common law public duty rule that limited public tort liability and the Legislature’s codification of the public duty rule in 1994, in § 10(j).14
Part II of this Article discusses the Supreme Judicial Court’s application of § 10(j) in recent decisions.15 Due to the relative newness of the provision and the dearth of Supreme Judicial Court case law, Part II also examines a relevant Massachusetts Appeals Court decision to supplement the Supreme Judicial Court’s interpretations of § 10(j).16 Part III of this Article charts the history of Brum v. Town of Dartmouth from its genesis in the Bristol Superior Court of Massachusetts in 1995, to the Massachusetts Appeals Court in 1998, and finally, to the Supreme Judicial Court in 1999.17
[*PG387] Part IV of this Article analyzes the plain language of § 10(j) and suggests a new three-part framework for interpreting the provision in harmony with both its plain language and with existing public duty jurisprudence.18 This Article then uses this framework to explore the weaknesses inherent in the Supreme Judicial Court’s interpretation of § 10(j) in Brum II.19 Finally, this Article explores the fine distinctions drawn by a correct interpretation of § 10(j) by proposing several hypotheticals and their likely resolutions under the provision.20 In light of the harsh results of these distinctions, as demonstrated by Brum II, this Article suggests a new exception to public tort immunity for a public employer’s actual notice of a risk of harm and deliberate indifference toward that risk of harm.21
Massachusetts, like most other states in the United States, imposes no affirmative duty of protection on its public schools through state statutory or constitutional directives.22 In fact, state statutory law and its incorporation of traditional tort principles has made Massachusetts public schools and school districts largely immune to tort actions.23
At common law, an individual could not sue the government in tort.24 This “sovereign immunity” from suit came to the United States from England, where “the King could do no wrong” because there was no court above him and thus, no court could enforce a claim against him.25 When the Declaration of Independence was signed, most American states adopted English common law, with the result that the federal government, the states and all of their agencies were immune from tort actions.26 Although municipalities are not sovereign agencies of the government, but rather are corporations chartered by the sovereign, they are, nevertheless, traditionally accorded immunity from tort liability under common law.27
In 1977, the Supreme Judicial Court joined a growing number of states that had abolished or limited the sovereign immunity of states and municipalities in tort actions.28 In Whitney v. Worcester, the court held that it intended to abolish the sovereign immunity doctrine, and therefore, the city of Worcester and its public school officials were not immune from suit by a student who was permanently blinded after being struck by a defective door at school.29 In Whitney, a school official directed a first-grade student, who was totally blind in one eye and had limited vision in the other eye, to proceed to afternoon re[*PG389]cess in the school yard, which required the student to pass through a defective door without assistance.30 When the door struck and injured the student, school officials directed the student to remain in the classroom, thus preventing the student from obtaining medical help that could have prevented total blindness.31
The court acknowledged that under the state’s sovereign immunity doctrine, both the city and its school officials were protected from liability for their failure to prevent injury by neglecting to accompany the student to recess and by neglecting to seek immediate medical attention after the injury.32 Noting that Massachusetts was one of only five remaining states that retained the sovereign immunity doctrine at both the state and local levels, the court decided that “the time for change [was] long overdue.”33 The court thus stated its intention to abolish sovereign immunity prospectively and erected in its place a new set of criteria under which public employers and employees could be held liable.34 Under a “discretionary-ministerial distinction,” the court reasoned that public employees should not be immune from liability for conduct that rested upon the exercise of judgment and discretion, but that they should be immune from liability for conduct that represented the implementation and execution of governmental planning and policy-making.35 In light of its abrogation of the sovereign immunity doctrine and its proposed “discretionary-ministerial” analysis, the court concluded that the school officials, and consequently the city, were not immune from liability because their failure to take appropriate action constituted ministerial conduct.36
[*PG390] Under pressure from the court, the Massachusetts Legislature enacted the MTCA.37 Modeled after the Federal Tort Claims Act, the MTCA provides that public employers, including the state, cities, towns and other public entities, are liable for the negligent acts and omissions of their employees “in the same manner and to the same extent as a private individual under like circumstances.”38 By establishing standards for bringing suit against public employers for tortious injuries, the Legislature virtually eliminated the traditional distinction between public and private employers.39 In so doing, the Legislature removed public employers’ sovereign immunity defense.40 Moreover, by including the “act or omission” language in the statute, the Legislature also eliminated the common law distinction between misfeasance and nonfeasance, thereby expanding public employers’ liability by enabling suits to be brought against public employers for their employees’ failure to act.41
While opening doors of potential liability through the MTCA, the Massachusetts Legislature simultaneously closed others by attaching several important exceptions to the MTCA that limited public employers’ liability.42 Specifically, § 10(a) of the MTCA excludes claims based upon acts or omissions of public employees where such employees have demonstrated due care in executing state or local laws, regardless of the laws’ validity.43 Closely paralleling a provision in the Federal Tort Claims Act, the § 10(a) exclusion likely stems from a judicial desire to immunize uniquely “public” functions prescribed by law, irrespective of the wisdom of the underlying statute or regulation.44 Moreover, by exempting from liability claims based upon public employees’ acts or omissions involving the exercise of “due care,” the Legislature disposes of frivolous suits that would likely fail under a traditional negligence analysis.45
In addition, § 10(b) of the MTCA provides that immunity for discretionary functions—traditionally enjoyed by public employers and public employees at common law—should be protected by excluding claims arising from the exercise of or failure to exercise a discretionary function, whether or not the discretion involved was abused.46 The underlying reasoning for this exception, according to the Supreme Judicial Court in Whitney, was the belief that the judiciary’s ability to review the reasonableness of legislative and executive branch policy decisions undermined the separation of powers.47 In addition, the court determined that the discretionary activities of public employees were fundamentally different from conduct regularly performed by privately-employed individuals and therefore deserved distinct treatment under the law.48 Thus, because judicial inquiry into such discretionary activities “might impede governmental operations by subjecting governmental decision-making to after-the-fact judicial tort [*PG392]analysis” and the threat of potential legal liability, the court concluded that these decisions should be insulated from such review.49
Finally, under § 10(c), the MTCA also excludes claims arising out of the intentional torts of public employees.50 Patterned after the Supreme Judicial Court’s consistent holdings favoring immunity for public employers in intentional tort cases, § 10(c) also reflects common law considerations of vicarious liability and proximate cause.51 Indeed, because employees’ intentional torts are, in general, independently motivated and beyond the control of public employers, they necessarily fall outside of the above-mentioned common law doctrines establishing liability.52 Thus, the Legislature likely intended § 10(c) to prevent public employers from bearing the costs of such unforeseeable actions that occur outside of the scope of employment.53
With the passage of the MTCA and the resulting abrogation of sovereign immunity, the Supreme Judicial Court needed to determine the extent to which public employers would be liable for negligence in broad public oversight functions—”public duties” imposed by the Legislature upon public officials “for the common good.”54 Despite the protections against liability already afforded to public employers by the MTCA under §§ 10(a), 10(b) and 10(c), the court was con[*PG393]cerned that governmental inspection and licensing procedures, as well as police and fire protection, would become the target of numerous negligence suits that would cripple public employers’ ability to provide those needed services.55 In 1982, in Dinsky v. Town of Framingham, the Supreme Judicial Court addressed the “public duty” issue for the first time in the context of a public employer’s alleged negligent inspection of a home.56 The Dinsky court held that under the MTCA, plaintiffs must prove the usual elements of a negligence claim to recover damages, and that under the common law “public duty rule,” public employers are not liable for breach of a duty to make inspections because inspection is a duty owed to the general public rather than to specific individuals.57
In Dinsky, after their one-family residence was severely flooded, the plaintiffs sued the town of Framingham for negligently failing to inspect the lot prior to construction and issuing building and occupancy permits in violation of grading and drainage requirements.58 [*PG394]Although the state building code imposed a statutory duty on the State to inspect new lots, the court reasoned that this duty was owed to the public at large and was not intended to furnish individual property owners with a private cause of action.59 Thus, because the public employer owed no “duty” to the homeowners in the first place, the court reasoned that there was necessarily no breach or causation to trigger liability.60 The court determined that absent a special duty owed to the plaintiffs as members of a particular class of persons under the statute—a statutory intent exception to the public duty rule—no cause of action could be maintained.61 Furthermore, the court noted that a contrary decision would deter municipalities from enacting new regulations for the public good out of fear of increased liability.62 Joining other jurisdictions in its adoption of the common law public duty rule, the court concluded that the town was not liable for negligent inspection because inspection was a public duty and not a duty owed to specific individuals.63
Despite the Supreme Judicial Court’s attempt in Dinsky to restrict the application of the public duty rule to negligent governmental inspections, the court since has extended the public duty rule to other cases involving a wide range of public protective services.64 For example, between 1982 and 1993, the court denied liability for failure to enforce the housing code, failure to provide adequate police protection, failure to follow proper gun permitting procedures and failure to revoke an automobile registration.65 There were some exceptions, [*PG395]however, in which the court was willing to impose liability on public employers for the negligent performance of general protective services.66
In addition to the statutory intent exception to the public duty rule articulated in Dinsky, which allows public employers to be sued by plaintiffs who are expressly protected by statute, the Supreme Judicial Court, recognized a second exception based upon the relationship between public employers and plaintiffs.67 In 1984, in Irwin v. Town of Ware, the Supreme Judicial Court held that a town was liable for its police officer’s failure to arrest an intoxicated motorist who subsequently caused an accident resulting in harm to the plaintiffs because a special relationship existed between the police officer and the plaintiffs.68 This special relationship, the court determined, imposed a special duty upon the officer “to take affirmative action to protect the plaintiff[s],” thereby creating an exception to the public duty rule.69 Having stopped a motorist on the road for speeding, the police officer in Irwin subsequently released the motorist who demonstrated clear signs of intoxication.70 Ten minutes later, the motorist collided head-on with the plaintiff’s decedents.71
Noting the heavy burden required to prove that the motorists were members of a special statutory class under Dinsky’s statutory intent exception, the Irwin court articulated another exception to the public duty rule—the special relationship exception—based upon the [*PG396]existence of a special relationship between the actor whose duty was at issue and the potential plaintiff.72 In applying the exception, the Irwin court first considered the applicable statutes governing police officers’ responsibilities toward intoxicated motorists.73 The court noted that although these statutes did not expressly identify the motorists as a protected class under the statutory intent exception, the statutes evinced a legislative intent to protect both intoxicated persons as well as other highway users.74 More importantly, the court also considered the foreseeability, immediacy and severity of the risk of harm; the unique ability of the police officer to prevent the harm; and the public policy concern for providing an effective remedy for persons injured by the negligence of public employers, while limiting the comprehensiveness of the public duty rule.75 In light of statutes “giving police officers the right to deal with intoxicated persons,” as well as the immediacy and foreseeability of the risk of physical injury “created by the negligence of a municipal employee,” the court determined that “a duty of care should reasonably be found.”76 Thus, the Irwin court concluded that because of the existence of a special relationship between the police officer and the motorists, the officer [*PG397]owed the motorists a special duty of protection that constituted an exception to the public duty rule.77
In addition to Irwin, the Supreme Judicial Court has applied the special relationship exception in one other public duty case: A.L. v. Commonwealth.78 In 1988, in A.L., the court held that the commonwealth was liable for a probation officer’s failure to enforce the specific probation conditions imposed upon a sex offender, who subsequently procured a job working with young boys whom he sexually molested, because a special relationship existed between the probation officer and the boys.79 Having received an eighteen-month suspended sentence conditioned upon his refraining from teaching and from associating with young boys, the sex offender obtained employment in a school.80 The probation officer met with the offender regularly, but never attempted to verify that the offender was neither teaching nor working in an environment in close proximity to young boys.81
Despite the lack of any statute specifically identifying the young boys as a protected class, the court reasoned that “the conditions of probation imposed by the sentencing judge . . . were designed to protect young boys” as a class, thus creating a special relationship between the boys and the probation officer.82 In addition to the probation conditions, the court also based its finding of a special relationship on the foreseeability of the risk of molestation created by the officer’s failure to monitor the nature of the offender’s employment, the chronic and [*PG398]immediate nature of the risk of molestation, and the unique ability of the probation officer—as opposed to the young boys or their parents—to prevent the harm, given his knowledge of the offender’s criminal record.83 The court concluded that because of this relationship, the probation officer had a special duty to verify the place of the offender’s employment, thus making the public duty rule inapplicable.84
At the same time that the Supreme Judicial Court was fashioning exceptions to the public duty rule in Irwin and A.L., a separate line of [*PG399]cases arose that further limited the public duty rule.85 Notwithstanding Whitney’s and the MTCA’s rejection of the misfeasance/nonfeasance distinction, this line of cases embraced the distinction and held that where a public employee acts negligently—creating a risk of harm that comes to pass—the public duty rule is not applicable.86 Rather, the employee is liable under a traditional misfeasance analysis, whereby “one who takes action ordinarily owes to everyone else who may be affected thereby a duty to act reasonably.”87
In 1990, in Onofrio v. Department of Mental Health, the Supreme Judicial Court held that the public duty rule was not applicable where a public employee took action that exposed the plaintiff to risk because the employee was bound “as any other person would be [under a traditional misfeasance analysis], to act reasonably.”88 In Onofrio, a Department of Mental Health employee placed a client in a property owner’s home without warning the property owner of the client’s proclivity for starting fires, and the client subsequently set fire to the home.89
[*PG400] The Onofrio court based its application of the public duty rule upon the common law distinction between misfeasance and nonfeasance—between “taking action that expose[s] [another] to risk” and merely failing to prevent a third person’s harmful activity.90 The court reasoned that the public duty rule applied only in situations where “a plaintiff has been directly harmed by the conduct of a third person and only indirectly by a public employee’s dereliction of a duty—a duty imposed on him solely by his contract of employment—to interrupt or prevent the third person’s harmful activity.”91 Simply stated, the court determined that liability would not attach where the public employer failed to prevent or diminish an independent, third party-created risk of harm that came to pass.92 By contrast, under the facts of Onofrio, the court reasoned that the public employee had created the risk of harm that came to pass by actively placing the client in the house, and thus, the employee owed the property owner a duty of reasonable care.93 In light of the misfeasance/nonfeasance distinction, the Onofrio court concluded that although the public duty rule protected public employers from liability for failing to prevent intervening third parties’ wrongful conduct, liability would apply to an employee’s wrongful, risk-creating acts.94
The court did not address, however, the application of liability to a situation where, in the absence of an intervening third party, the employee’s apparent failure to act may have created the risk of harm that came to pass.95 This issue presented itself in 1991, in Mamulski v. Easthampton, where the Supreme Judicial Court held that the public duty rule did not apply to a public employee’s failure to replace a missing stop sign because the employee actively created the risk of harm that came to pass.96 In Mamulski, the plaintiff’s decedent was killed in an automobile accident allegedly resulting from a public employee’s failure to replace a missing stop sign at a major intersec[*PG401]tion, despite numerous warnings by concerned people in the area.97 The court first analogized to Onofrio, citing the lack of any wrongful, third party conduct.98 Noting the absence of an independently-created risk of harm, the court reasoned implicitly that were it not for the employee’s failure to replace the stop sign, the risk of a deadly car accident would not have come to pass.99 The court further reasoned that because the town controlled the public land upon which motorists traveled and actively held out roads to be reasonably safe for public use, the town had “a duty to exercise reasonable care toward persons lawfully on that property.”100 Therefore, the court reasoned implicitly that by failing to replace the stop sign, the town did not merely fail to prevent the risk of harm that came to pass, but rather, the town “itself set[] in motion forces which cause[d] the harm.”101 The court concluded that because the public employee actively created the risk of harm, the public duty rule did not apply and thus, the town was liable for the harm that came to pass.102
In contrast to the majority of cases applying the public duty rule to shield employers from liability, Irwin, A.L., Onofrio and Mamulski reflected the Supreme Judicial Court’s growing dissatisfaction with the public duty rule’s broad application and its interest in shaping [*PG402]cohesive and consistent exceptions to the rule.103 Moreover, these cases also evinced the court’s effort to balance public policy issues of private compensation and public accountability with the protection of government from crippling financial liability and the creation of incentives for the provision of public services.104 In its attempt to narrow the scope of the public duty rule, however, the court exposed cracks in the very justification of the rule.105 Indeed, despite the consensus on granting immunity for public duties under Dinsky, subsequent court cases had not yet drawn “an intellectually defensible line between immune ‘public’ duties and actionable negligence.”106 Far from mere aberrations, these exceptional cases were harbingers of change.
The separate lines of cases narrowing the public duty rule came together in 1992, in Cyran v. Town of Ware.107 In Cyran, the Supreme Judicial Court held that a town’s firefighters were not liable under the public duty rule for negligently fighting a fire because the obligation to provide fire protection was a public duty and plaintiffs were not members of a statutory class required to trigger a special duty exception.108 In Cyran, a fire broke out in a two-family residential building and the plaintiffs alleged that the fire department responded to and fought the fire negligently, causing foreseeable damage to the property.109 The court reasoned that the fire department’s duty was a general duty of fire protection owed to the town and required only that the fire department respond to and deal with the fire as permitted by their resources and training.110 Drawing upon the misfeasance/nonfeasance distinction articulated in the Onofrio line of cases, the court reasoned that because the fire department did not actively create the fire, it was not liable for failing to prevent the damage that the fire caused.111 The court explained that the fire department’s conduct was distinguishable from Onofrio and Mamulski because the [*PG403]fire department failed to act to effectively fight the fire, as opposed to having negligently caused the fire.112
The court also ruled out a possible special relationship exception to the public duty rule because the plaintiffs were not members of a special statutory class.113 Additionally, the court considered the existence of a special relationship exception based upon assurances of special care given by the fire department, but summarily rejected this theory of exception in the absence of such assurances.114 Finally, the court determined that imposing liability for negligent firefighting would violate public policy because cities and towns would be exposed to liability for damages each time a plaintiff considered a fire department’s acts to be unsatisfactory.115 Thus, the Cyran court concluded that in the absence of any statute creating a special relationship, the town was not liable for the firefighter’s negligence under the public duty rule.116
In his concurrence, Justice O’Connor supported the court’s refusal to impose liability, reasoning that the case fell directly under the precepts of the public duty rule: that no State, municipality or county should be liable “for injuries and losses sustained by individuals as a result of a public employee’s failure to act, as required by the terms of his or her employment, to prevent or diminish the harmful consequences of a condition or situation not originally caused by the employee.”117 Consistent with the majority’s holding, Justice O’Connor’s reasoning in support of the public duty rule was implicitly based upon the misfeasance/nonfeasance distinction articulated in Onofrio and Mamulski, insofar as liability would be imposed for originally causing the risk of harm (i.e., by actively creating the risk), but not for “fail[ing] to prevent or mitigate” the risk of harm (by taking no action at all).118 Moreover, Justice O’Connor’s emphasis on “original cause” implicitly suggested an expansion of the public duty rule as defined [*PG404]by prior cases.119 According to Justice O’Connor, the public duty rule would apply not only to situations where an employee failed to prevent an independent, third-party-created harm, but also to any situation “in which a plaintiff has been harmed by a condition or situation which was not originally caused by the public employee, and is attributable to the employee only in the sense that the employee failed to prevent or mitigate it.”120 Under Justice O’Connor’s formulation, this “originally caused by” language would shield public employers from liability for their failure to prevent not only third-party harms, but also any other harms (including contributory wrongdoing on the part of the plaintiff or other extraneous causes of harm, like the fire in Cyran) not “originally caused by” public employees.121
Justice O’Connor also suggested that only two narrow exceptions—neither of which he believed was applicable in Cyran—be made to the public duty rule to bring it into conformity with the “traditional” common law rule: first, where a special duty was owed because the plaintiff was within a specifically identified statutory class; and, second, where a special relationship existed between the plaintiff and the public employee insofar as the employee expressly or implicitly represented to the plaintiff that special care would be taken.122 By limiting the public duty rule to its traditional exceptions, Justice O’Connor implicitly rejected the “new” special relationship exception applied in Irwin and A.L. that established liability on the part of public employers.123 Contrary to the majority, which distinguished Irwin and A.L. from Cyran based upon the existence of special statutory protection for the plaintiffs, Justice O’Connor argued that but for this “new” exception, the three cases were factually indistinguishable and wrongly decided.124 Justice O’Connor urged the majority to overrule [*PG405]Irwin and A.L., thereby bringing public duty case law into harmony with the public duty rule articulated in Cyran.125 By applying the public duty rule, with its narrow exceptions, in this sweeping fashion, Justice O’Connor emphasized that “the public duty rule in Massachusetts [would] be made clear,” predictability of liability would be greatly improved and the twin public policy concerns of private compensation and governmental protection would be reasonably balanced.126
In 1993, in Jean W. v. Commonwealth, the Supreme Judicial Court held that the public duty rule was inconsistent with the MTCA and thereby announced its intention to abolish the rule prospectively after the conclusion of the 1993 legislative session.127 In Jean W., a Massachusetts Parole Board clerk erroneously granted the parole of a convicted murderer.128 Although the parolee regularly reported to his parole officer, the officer failed to discover the error.129 Six months after his release, the parolee raped, beat and threatened the plaintiff, who later sued the commonwealth, the parole board and the Department of Corrections for negligence.130
The Massachusetts Superior Court dismissed the claim on the ground that it was barred by the public duty rule because the plaintiff had not shown that the defendants owed her a special duty different from that owed to the general public.131 On appeal, the plaintiff argued that a special relationship existed based upon the foreseeability of the risk of harm to the plaintiff, and therefore, the public duty rule should not apply.132 Taking the appeal on its own initiative, the Supreme Judicial Court, in a decision featuring four separate concur[*PG406]ring opinions, abrogated the public duty rule and unanimously reversed the lower court’s dismissal.133
According to Chief Justice Liacos, the “inconsistent and irreconcilable parts” of the public duty rule, namely the “public duty-special relationship dichotomy,” had left “both the Justices and the litigants quite incapable of predicting when and why liability will be imposed.”134 As an example, the Chief Justice articulated the confusion surrounding the application of the special relationship exception by comparing Irwin and A.L.—where liability was imposed because of a special relationship exception to the public duty rule—with various other cases where the public duty rule barred the imposition of liability despite the existence of statutes that seemed to confer special relationships.135 In addition, Chief Justice Liacos reasoned that, in light of more recent cases such as Cyran, the framework for applying the public duty rule as determined by the court in Onofrio and Mamulski had also led to inconsistent results.136 Narrowly interpreting the former cases to require the presence of an independent, third-party-created harm in order to trigger the public duty rule, the Chief Justice held that the public duty rule should not to have applied to Cyran because there was no intervening perpetrator.137 Moreover, the Chief Justice explicitly rejected Justice O’Connor’s broad interpretation of the Onofrio line of cases in Cyran, where Justice O’Connor reasoned that the public duty rule’s application in Onofrio and Mamulski rested not on the presence or absence of a third-party-created harm, but on whether the employee “originally caused” the risk of harm.138 According to the Chief Justice, this misfeasance/nonfeasance distinction was too flexible to provide “certainty and predictability essential to the [*PG407]law,” because the defendants’ conduct in Mamulski and Cyran could be characterized as either acting negligently (misfeasance) or failing to act at all (nonfeasance), thus leading to contradictory outcomes.139
In addition to the confusion engendered by the inconsistent application of the public duty rule, the Chief Justice likened the rule’s comprehensiveness to the “antiquated and outmoded concepts of sovereign immunity” that the court and the Legislature had shed, respectively, in Whitney and in the MTCA.140 Calling attention to the fundamental inconsistency between the abolition of sovereign immunity and the preservation of the public duty rule, the Chief Justice criticized the unfairness inherent in a rule that resulted in “a duty to none when there is a duty to all” on the one hand, and “tortured analyses” resulting from desperate attempts by courts “to avoid harsh results without squarely facing the problem” on the other.141 The Chief Justice suggested that abrogation of the public duty rule would remedy such discrepancies by eliminating the distinction between private persons and public employers as envisioned by the MTCA and thus would regularize the imposition of liability.142 Acknowledging fears of the potential financial burden that would be imposed on public employers by the rule’s abrogation, the Chief Justice indicated that the plaintiff’s burden of proof of negligence (duty-breach-causation-damages) and the statutory cap on damages, in addition to preexisting immunity for discretionary acts under § 10(b) and intentional torts under § 10(c), would protect the commonwealth from spiraling costs.143 Finally, turning to the facts of Jean W., the Chief Justice held [*PG408]that the case be remanded to allow the plaintiffs to establish the existence of a special relationship between either themselves and the defendants or between the defendants and the third party perpetrator (a second type of special relationship), which would trigger a duty of protection.144
Justice O’Connor, joined by Justices Nolan and Lynch, supported the reversal of the lower court’s decision based upon the misfeasance/nonfeasance distinction.145 Applying the reasoning of Onofrio, Justice O’Connor determined that Jean W. was “not a case in which the plaintiffs were injured as a result of a public employee’s failure to act to rectify a situation not created by the employee,” but rather, the employee actively exposed the plaintiff to a new risk by releasing the convicted murderer.146 Despite his refusal to apply the public duty rule in Jean W., however, Justice O’Connor did not support abrogation of the public duty rule.147 Rather, Justice O’Connor championed the rule’s application in previous cases as being consistent with traditional tort principles of misfeasance and nonfeasance, as well as with current social values and public policy concerns of private compensation, governmental protection and predictability.148
Furthermore, in opposition to Chief Justice Liacos’ assertion that the public duty rule had no basis in principles of misfeasance and nonfeasance, Justice O’Connor vehemently argued that “there [is] no question about it. . . . The traditional public duty rule does make such a distinction,” and that “for nearly ninety years the court has been distinguishing between a defendant’s acts and omissions in determining whether a duty of care exists . . . .”149 This distinction, Justice O’Connor reasoned, was well-established in prior case law where the court consistently held that public employees, “required by the terms of their employment to prevent or diminish harmful consequences to the public of conditions or situations not originally caused by the employees, did not owe to individual members of the public a duty of [*PG409]protection commensurate with their duty to the public as a whole.”150 Justice O’Connor argued that by abrogating the public duty rule the court not only paved the way for broad-based public tort liability, but also destroyed the critical distinction between acts and omissions—thus imposing a duty of care upon public employees regardless of whether they created the risk of harm.151
Moreover, according to Justice O’Connor, this “new” duty of care ran counter to the purpose and intent of the MTCA.152 Indeed, whereas the MTCA sought to make public employers liable in the same circumstances that would result in private employer liability, Justice O’Connor determined that the court’s abrogation of the public duty rule and its attendant undermining of the misfeasance/nonfeasance distinction would force public employers to shoulder an additional burden of liability not borne by private individuals.153 This new duty, Justice O’Connor reasoned implicitly, would make public employers liable for their failure to act in response to a situation that they did not create, i.e. nonfeasance, but would have no application to private individuals’ failure to prevent or diminish harm.154 Justice O’Connor thus concluded that the public duty rule properly gauged the imposition of governmental liability under the MTCA and cautioned against its proposed abrogation, which he determined would lead to catastrophic governmental liability by creating entirely new duties for public employers.155
Prompted by the court’s announcement in Jean W. that it would abrogate the public duty rule and by fears of unlimited liability for cities and towns, the MTCA was amended to protect public employers from liability in situations previously covered by the common law public duty rule.156 The MTCA amendments, added by Statute 1993, chapter 495, were enacted as part of the supplemental budget and took effect when the governor signed the budget on January 14, 1994.157 The amendments added several new exceptions to liability for public employers under § 10 of the MTCA.158 Closely paralleling the holdings of prior public duty cases, the first four subsections, §§ 10(e)–10(h), immunize public employers from any claim based upon the execution or failure to execute specific public functions.159 Specifically, § 10(e) shields public employers from liability for negligent licensing;160 § 10(f) bars liability for negligent inspection by public employers;161 § 10(g) bars liability for negligent fire prevention [*PG411](not including claims based upon the negligent operation of motor vehicles);162 and § 10(h) bars liability for negligent police protection (not including claims based upon the negligent operation of motor vehicles, or the negligent protection, supervision or care of persons in custody).163 In addition, § 10(i) shields public employers from liability for claims based upon the release, parole or escape of persons in public custody, unless gross negligence is shown.164 Together, these amendments codify the Supreme Judicial Court’s decision in Dinsky and its progeny—that public employers should not be held liable for negligence in broad public oversight functions.165
In contrast to the above-mentioned sections that provide public employers with immunity for specific public functions, § 10(j) applies to all public functions—shielding public employers from liability for “any claim based on an act or failure to act to prevent or diminish the harmful consequences of a condition or situation, including the violent or tortious conduct of a third person, which is not originally caused by the public employer or any other person acting on behalf of the public employer.”166 In the wake of the court’s abrogation of the common law public duty rule in Jean W., § 10(j) was intended to replace the rule by providing a broad, catch-all exception to liability for public employers, not limited to any particular public function.167
[*PG412] Consistent with the Onofrio line of cases decided by the Supreme Judicial Court, § 10(j) is founded upon a misfeasance/nonfeasance distinction: a public employer will be held liable for actively creating a risk of harm that comes to pass, but will not be held liable for failing to prevent a risk of harm “not originally caused by the public employee,” even where the employee has a statutory or contractual employment duty to act.168 Section 10(j)’s “originally caused by” phrase thus sharpens Onofrio’s misfeasance/nonfeasance inquiry by emphasizing the employee’s material involvement in the creation of the risk of harm as grounds for imposing or refusing to impose § 10(j) liability.169 Furthermore, as alluded to above, the emphasis on original causation also expands the scope of the public duty rule by shielding public employers from liability for failure to prevent not only third-party-created risks of harm, but any risks of harm not originally caused by public employees.170
Reluctant to provide unqualified immunity to public employers who failed to prevent harm, the drafters of the MTCA amendments crafted several exceptions to § 10(j) that are consistent with the holdings of prior public duty cases.171 Section 10(j)(1) embodies a re[*PG413]stricted form of the traditional special relationship exception, providing that § 10(j) will not apply where public employees give specific assurances of safety or assistance—beyond general representations that investigation or assistance will be or has been undertaken—to an individual or to a member of the individual’s family and where the individual or family member relies on those assurances to their detriment.172 Section 10(j)(1) reflects the common law view that even where a public employee does not create the risk of harm, an employee’s assurances to a plaintiff that special care will be taken “justify[] the plaintiff’s reliance on the employee’s carrying out his responsibility.”173 Additionally, § 10(j)(2) bars the application of § 10(j) to any claim based upon the actual intervention of a public employee that causes injury to the victim or places the victim in a worse position than he was in before the intervention.174 This exception reflects traditional common law tort principles insofar as it allows the imposition of liability against a rescuer who negligently assists a victim or who actively creates the risk of harm that comes to pass.175 Consistent with Doherty v. Belmont, where, in 1985, the Supreme Judicial Court held that negligent maintenance of property was risk-creating conduct and was therefore actionable under the MTCA, § 10(j)(3) excludes negligent maintenance of public property from protection under § 10(j).176 Finally, § 10(j)(4) prevents the immunization of public employers for negligent medical or other therapeutic treatment and [*PG414]is consistent with prior case law that held medical malpractice actionable under the MTCA.177
In the wake of the MTCA amendments, the Supreme Judicial Court wasted no time in applying § 10(j)’s statutory public duty immunity.178 In applying § 10(j), the court did not attempt to define the plain language of the provision, but rather imposed liability on public employers for creating the risk, consistent with Onofrio’s misfeasance/nonfeasance distinction.179 Thus, in 1994, in Pallazola v. Town of Foxborough, the Supreme Judicial Court held that § 10(j) applied retroactively to bar a plaintiff’s claim against a town for failure to provide sufficient police protection and prevent the unlawful removal of a [*PG415]portion of an aluminum goal post from a football stadium.180 In Pallazola, the plaintiff was seriously injured when a portion of a goal post, removed and carried by other spectators, came into contact with an overhead power line and fell on him.181 Relying on the plain language of § 10(j), the court held that § 10(j) barred the plaintiff’s claim because the failure to prevent harm fell within the limits of the public duty rule.182
In December of 1994, in Bonnie W. v. Commonwealth, the Supreme Judicial Court held that the commonwealth was not liable for a parole officer’s negligent failure to supervise a parolee and prevent the parolee from gaining access to the plaintiff’s mobile home that resulted in the sexual assault of the plaintiff.183 Conversely, the Bonnie W. court held that the commonwealth was liable for the parole officer’s negligent recommendation of the parolee’s continued employment to park management as well as the officer’s misrepresentation of the parolee’s criminal history.184 In Bonnie W., a man convicted of rape and robbery was released on parole and then hired as a maintenance man at a trailer park in which the plaintiff lived.185 In the ensuing months, the parole officer failed to meet with the parolee as the parole board rules required and, in response to questions by park management pertaining to the parolee’s employment, the parole officer negligently recommended the parolee’s continued employment at the park and made several misrepresentations about the parolee’s criminal record.186 Possessing keys to all of the park units, the parolee gained access to the plaintiff’s mobile home and sexually assaulted her.187
Because the plaintiff’s claim of negligent supervision was based upon the negligent failure of the parole officer to prevent harm, as specifically excluded by § 10(j), the court reasoned that § 10(j) barred that claim.188 Relying upon Onofrio’s misfeasance/nonfeasance distinction, however, the court reasoned that the commonwealth was liable for the parole officer’s negligent recommendation of the parolee’s continued employment at the trailer park and misrepresenta[*PG416]tion of the parolee’s criminal history.189 Insofar as the parole officer “[took] action that exposed the plaintiff to risk,” the court reasoned that the parole officer “was bound, as any other person would be, to act reasonably.”190 Thus, the court concluded that § 10(j) barred the plaintiff’s claim based on the parole officer’s negligent supervision because such conduct constituted a failure to prevent harm.191 The court also held that § 10(j) did not bar the plaintiff’s claim based on the parole officer’s negligent recommendation and misrepresentation because such conduct constituted an active creation of the risk of harm.192
In addition, in 1996, in Lawrence v. City of Cambridge, the Supreme Judicial Court of Massachusetts stated, in dictum, that insofar as the police did not make explicit and specific assurances of safety or assistance to the plaintiff, § 10(j) barred a storeowner’s claim against the city based upon the failure of the police to protect him from attack.193 In Lawrence, third-party assailants assaulted a storeowner outside of his store, after which he identified the suspects for police who subsequently arrested the assailants.194 In light of one assailant’s extensive criminal record, the police promised to protect the plaintiff in the days leading up to the assailant’s arraignment and did so for three consecutive days.195 On the fourth day, the police were not present and the plaintiff was shot after closing the store.196 Briefly reiterating the plain language of § 10(j), the court stated that absent explicit and specific assurances, § 10(j) barred the plaintiff’s claim based on the police officers’ failure to fulfill their promise of protection because such conduct constituted a failure to prevent harm.197
Consistent with the reasoning of the Supreme Judicial Court, the Massachusetts Appeals Court likewise based its application of § 10(j) on an implicit misfeasance/nonfeasance analysis, while refusing to define the plain language of the provision.198 Thus, in 1998, in Allen v. [*PG417]City of Boston, the Massachusetts Appeals Court denied a city’s motion for summary judgment, holding that § 10(j), in the absence of further evidence, did not bar the plaintiff’s claim against the city based on the school’s negligent “handling” of a student (Mr. Reynolds), who fatally stabbed a fellow student.199 In Allen, Mr. Reynolds had been suspended from two other schools for possessing a knife, before being enrolled at the school in which the attack occurred.200 The school provided no evidence in support of their motion for summary judgment—other than the text of the “Boston Public Schools Code of Discipline”—that they considered questions of whether to enroll Mr. Reynolds at the school; whether to integrate Mr. Reynolds with the general school population; and how to oversee Mr. Reynold’s attendance.201 In addition, the school supplied no evidence detailing its consideration of security questions in light of the threat of harm that Mr. Reynolds posed.202 The court reasoned implicitly that because the school knew of the threat posed by Mr. Reynolds and nevertheless actively enrolled him without concern for his placement at the school or the safety measures in existence, the city failed to adequately support the contention that it was immune from liability under § 10(j) for Mr. Reynolds’ violent act.203 Thus, the court concluded that § 10(j) did not bar the plaintiff’s claim because based on the evidence, the city the did not establish that the school merely failed to prevent harm.204
Despite its several interpretations of § 10(j), the Supreme Judicial Court did not directly address § 10(j)’s plain language until 1999, in Brum II, in which the court held that a town was not liable for the death of a student in the school because the school officials’ failure to implement proper security measures and to apprehend the attackers did not “originally cause” the condition or situation that led to the harm.205 Because Brum II represents the Supreme Judicial Court’s most recent and most comprehensive analysis of § 10(j), its holding provides the benchmark against which all future public tort liability cases will be measured.206 Moreover, Brum II reversed the Massachusetts Appeals Court in Brum I, which departed from prior public duty jurisprudence (favoring public immunity) and held that a school was liable for failure to properly secure the school and to deter attackers.207 The appeals court’s decision in Brum I, although subsequently reversed, is important because of the Supreme Judicial Court’s adoption of portions of its reasoning and because of its uniqueness in the scheme of Massachusetts public duty jurisprudence.208
On April 12, 1993, shortly before the beginning of classes at Dartmouth High School, two groups of youths, including several Dartmouth High School students, were involved in a violent altercation on the school premises.209 Tensions between the two groups had escalated during the prior week and had erupted in a physical confrontation on the previous evening.210 School officials detained two of Jason Robinson’s friends for their role in the altercation on the morning of April 12.211 The other group of youths immediately fled the school following the incident.212 One of the detained students involved in the ongoing dispute warned the principal that at least three youths from the opposing group had threatened to return to the high [*PG419]school to retaliate against him and his friends, including Robinson.213 Shortly after 8:00 A.M., the principal and other school officials witnessed the trio enter the front door of the high school.214 The intruders openly brandished weapons, including two knives, a billy club, a baseball bat and a length of pipe.215 The school officials did nothing to confront or obstruct the three youths, who proceeded unimpeded to a second-floor classroom.216 At least two of the three students entered a social studies classroom believing that they would find a particular individual, and when they failed to find him, they instead attacked and stabbed Robinson, who died in the classroom.217
In 1995, Robinson’s mother, Elaine Brum, commenced action against the town of Dartmouth, school officials and other municipal officials.218 The plaintiff sought damages for the violation of Robinson’s federal and state civil rights, for Robinson’s wrongful death and for negligence under the MTCA, including (1) the failure of the responsible municipal officials and school officials to institute any security measures to protect the high school students, and (2) the failure of the principal to respond to the foreseeable (and explicitly forewarned) threat of harm that the youths who murdered Robinson presented.219 The defendants filed a motion to dismiss.220
On October 20, 1995, a judge of the Bristol Superior Court of Massachusetts granted the motion to dismiss, holding that the town was immune from liability because (1) the adoption of security measures was a discretionary function under § 10(b), and (2) the school officials’ failure to act—while not a discretionary function—fell within the “public duty rule” of § 10(j).221 The plaintiff appealed, and in February, 1998, in a 2–1 vote, the Massachusetts Appeals Court reversed the judgment with respect to the negligence claims.222 The court held that the school had no discretion to adopt and implement [*PG420]a school safety policy because of a state statute requiring schools to publish policies pertaining to student and teacher conduct, and to enact safety standards and procedures at the school.223 Therefore, the court determined that the officials’ failure to take such action was not immune under § 10(b).224 The court also held that the school officials’ failure to prevent harm was not immune under § 10(j) because the failure to implement a safety policy, along with the failure to deter the trio from proceeding into the school, “caused” the “condition” that led to the murder.225
Turning to § 10(j), the appeals court framed the issue as whether the school’s failure to adopt or implement any security measures, as well as the school officials’ failure to deter the “known and imminent threat” posed by the attackers’ entry into the school, originally caused the condition or situation that led to the death of Robinson.226 Finding no case definitively interpreting the meaning of the words “not originally caused by,” the court decided that it was able to apply the words in accordance with their plain meaning: to “create[]” a condition or situation, or a “risk,” that leads to harm.227 The court thus reasoned that by failing to adopt and implement security measures, the school officials originally caused a condition or situation “of total insecurity” at the school which led to “foreseeable harmful consequences” (i.e., the “invasion and fatal attack on Robinson by the returning trio of violent students”).228 Moreover, the court determined that by failing to deter the “known and imminent threat” that the attackers posed once they entered the school, the school officials exacerbated the condition of insecurity that they had created, “thereby materially contributing to the circumstances” that led to the to the harmful consequences.229 According to the court, the failure of the [*PG421]school to act in either instance was an original cause of the harm insofar as it “created” the risk of harm that came to pass.230
In a footnote to the Brum I opinion, the court explicitly rejected the town’s argument that because the school merely failed to prevent or mitigate the harm, as opposed to actively causing the harm, the public duty rule had no application.231 Quoting Whitney, the 1977 case in which the court vacated a decision denying recovery to a student injured at school and announced its intention to abolish the sovereign immunity doctrine, the court stated that municipal liability “should not be influenced by the finite distinctions drawn in [the] cases [involving misfeasance-nonfeasance analysis], distinctions which have no real connection with sound reasoning or policy.”232 Therefore, the court reasoned that the applicability of § 10(j) in Brum I did not hinge upon the misfeasance/nonfeasance analysis involving passive versus active involvement in the harmful consequences that came to pass.233 Rather, the court determined that § 10(j)’s applicability [*PG422]rested on whether the public employees’ action or inaction created the risk; in other words, whether the public employees’ failure to institute security measures in violation of § 37H “made possible” the harmful consequences that came to pass and whether the public employees’ action or failure to act, after the attackers had entered the building, “brought about or contributed” to those consequences.234 The court thus concluded that § 10(j) did not bar the plaintiff’s claim based on the school’s failure to institute adequate security measures and the school officials’ failure to deter the youths from entering the school because those failures constituted risk-creating conduct.235
[*PG423] In a dissenting opinion, Justice Kass attacked the majority for basing its decision not on the language of § 10(j), but on public policy considerations favoring private compensation for harms inflicted by third parties “or other extraneous causes . . . that competent public action might have prevented.”236 According to Justice Kass, it was the task of the Legislature—not the court—to resolve such public policy questions and that in this instance, where a student was killed by third-party attackers, “the Legislature ha[d] done so unmistakably.”237 The Justice reasoned implicitly that under § 10(j), liability should be imposed only insofar as the public employer originally caused the harmful consequences.238 Therefore, because “[n]o school official stabbed Robinson to death,” but rather, the school officials merely failed to prevent the killing, the Justice concluded that the school officials’ conduct fell squarely within the exclusionary language of § 10(j).239
In addition to citing the absence of any affirmative action on the part of the school officials in support of immunity for the school, Justice Kass also argued that the Legislature’s inclusion of the phrase “violent or tortious conduct of a third person”—conspicuously absent from the original draft of the MTCA amendments—indicated an intent to exclude the type of claim brought by the plaintiff in Brum I.240 The Justice asserted that because Robinson’s death was originally caused by the violent acts of third parties, the school was explicitly excluded under § 10(j).241 Finding the language of § 10(j) unambiguous, Justice Kass concluded that the majority’s analysis was nothing short of “judicial nullification of a legislative act.”242
In 1999, in Brum II, the Supreme Judicial Court reversed the Massachusetts Appeals Court.243 The court stated that the issue in Brum II, whether the school officials’ failure to implement proper security measures and to otherwise deter the intruders originally caused the condition or situation that led to Robinson’s death, presented an “interpretive quagmire,” and cited for support § 10(j)’s inconsistent application by superior court judges.244 The court also acknowledged that the Supreme Judicial Court had never directly addressed the meaning of “originally caused by,” specifically at issue in Brum II.245
Turning to the “convoluted and ambiguous” language of § 10(j), the court reasoned that the principle purpose of § 10(j) was to provide public employers with a substantial measure of immunity from liability arising out of a public employee’s failure to prevent or diminish “certain harmful consequences.”246 The only exception to this broad immunity, the court determined, was where the condition or situation giving rise to such consequences was “originally caused by the public employer.”247 In addition, by interpreting the “including the violent or tortious conduct of a third person” clause—in harmony with proper grammar and common sense—to modify the noun “consequences,” and not the words “condition or situation,” the court reasoned that violent or tortious third-party conduct was not a condition or situation but was rather a harmful consequence.248 The court thus determined that § 10(j) immunized public employers from all harmful consequences befalling a plaintiff, including the consequence of violent or tortious third party conduct, except where the employee originally caused the condition or situation leading to those harmful circumstances.249
[*PG425] The court next applied this interpretation of § 10(j) to the facts of Brum II.250 The court, echoing the arguments of the plaintiff, determined that “the killers’ acts or the death of their victim” were “the harmful consequences,” and, implicitly adopting the appeals court’s reasoning, further determined that the “injury-causing condition of physical insecurity” was the condition or situation that led to such consequences.251 Reasoning that the failure to secure the school and to deter the intruders constituted a “neglect of duty” of protection and that this neglect of duty was in fact “a failure to act to prevent or diminish” harm, the court held that the school officials’ conduct did not originally cause the condition or situation that led to harm.252 Citing Kass’s dissent in Brum, the court reasoned that although the school officials might have prevented the killing, their failure to prevent it was in the “excluded category” of § 10(j).253 Indeed, to recast the school’s failure to prevent the killing as an original cause of the harmful condition or situation, the court added, would undermine the principal purpose of the provision, which was to immunize “act[s] or failure[s] to act to prevent,” and would allow the exception to immunity to swallow the rule.254 The court further stated that it would be hard put to find any example of a condition leading to a harmful consequence, where the condition was originally caused by the public employer versus being brought about by the public employer’s failure to prevent the condition.255
In contrast to the policy interests that the Massachusetts Appeals Court invoked to support the broad application of public tort liability under § 10(j), the Brum II court relied upon Justice O’Connor’s concurrence in Cyran as the most likely interpretation of § 10(j), thus supporting public immunity where the school merely “failed to prevent or mitigate” the harmful condition or situation.256 The court also relied on the circumstances giving rise to § 10(j)’s enactment for support, concluding that, in the wake of Jean W., § 10(j) was a statutory codification of the public duty rule, “intended to provide some substantial measure of immunity from tort liability to government [*PG426]employers.”257 In light of this legislative intent, the court reasoned that it should avoid interpreting the provision “so broadly as to encompass the remotest causation and to preclude immunity in nearly all circumstances.”258 As further support for its holding, the Brum II court likened the failure of school officials to secure the school and to apprehend the intruders with the parole officer’s failure to properly supervise the third-party parolee in Bonnie W. and to the police officers’ failure to protect the storeowner in Lawrence.259 The court observed that in each case, the actions of third parties originally caused the plaintiffs’ injuries; the defendants, on the other hand, never acted at all, and thus were not liable for failing to prevent the harm.260 By contrast, the court determined that the school’s failure to act was unlike the parole officer’s negligent recommendation of the parolee in Bonnie W. because the parole officer’s conduct constituted an “affirmative act” that gave the parolee access to the trailer and thus caused the harm.261
In the final footnote to its opinion, the majority stated that it was “in complete sympathy with the concurrence’s observations that it is unfortunate that school officials should escape all legal accountability for their failure to protect the children under their supervision.”262 Despite the unsettling results of immunizing the town, the majority believed that it would “distort the general regime of § 10(j)” for the court to interpret the provision to give rise to liability in Brum II.263 Indeed, because § 10(j) applied to all public employers and in a wide range of circumstances, the court left the task of changing the law explicitly to the Legislature.264 Thus, unwilling to undermine the principal purpose of § 10(j) by imposing liability for inaction, the court concluded that § 10(j) barred that plaintiff’s claim based on the school’s failure to act to prevent harm.265
[*PG427] In a powerfully-worded concurrence, Justice Ireland, joined by Justices Abrams and Marshall, agreed with the decision, but suggested that the Legislature respond to the “unfortunate” result by changing the law.266 Justice Ireland argued that the decision compelled by § 10(j) was wrong from a public policy perspective because regardless of who or what caused harm to a student on school property, “parents reasonably should be able to expect that the schools to which they entrust their children will take reasonable steps to protect their children from harm when, as here, the school officials are put on notice that the children are or may well be in jeopardy.”267 In addition to parents’ reasonable assumptions about the safety of their children, the Justice suggested that the Legislature was of the view that schools were responsible to protect students, as demonstrated by its passage of § 37H requiring standards and procedures assuring school building security and the safety of students and school personnel.268 The security measures mandated by the statute, the Justice added, could reasonably be interpreted to protect students against any harm, regardless of who caused it.269
Justice Ireland argued that a duty of protection should apply where a school official has advance notice of the risk of harm and where the officials show deliberate indifference toward that risk by failing to take any action to prevent it.270 Repeating the facts of the case, Justice Ireland emphasized that the school officials took no action to prevent the risk of harm despite notice of a prior altercation between several students and the assailants, and notice of threats of retaliation.271 In addition, when several school officials witnessed the assailants enter the school, the officials took no action to prevent the risk of harm.272 In fact, the only measures taken to prevent harm, the Justice continued, consisted of a “No Trespassing” sign that directed visitors to the school office.273 The Justice added that although school [*PG428]officials are not absolute guarantors of students’ safety, they should “be expected to take reasonable measures to protect children when they have advance notice of danger” and that it is up to the Legislature to impose such a duty on school districts.274
Brum II marks the first time that the Supreme Judicial Court has attempted to address the ambiguity plaguing § 10(j), and its decision leaves much to be desired.275 In fact, although it reversed the appeals court decision holding the town liable, the Supreme Judicial Court’s reasoning in Brum II is as profoundly dissatisfying, unsettling and ultimately ambiguous as the appeals court’s reasoning in Brum I.276 Indeed, in its attempt to clarify the language and intent of § 10(j) once and for all, the Brum II court reduced § 10(j) to a mere act/failure-to-act inquiry, failing to appreciate the fine distinctions made by the provision with regard to creation of risk and perpetuating an oversimplified statutory interpretation for lower courts to follow.277 As a [*PG429]result, § 10(j) remains an “interpretive quagmire,” engulfing public duty jurisprudence in a morass of ambiguity.278 In the wake of Brum II, important public policy questions continue to plague § 10(j), demanding that a clear judicial interpretation be made on behalf of “the [c]ommonwealth and other ‘public employers’ who must evaluate, settle or otherwise litigate [MTCA] claims” and whose perception of the ambiguity in the law may have a “chilling effect” upon their ability to provide important public services; “the citizens of the [c]ommonwealth whose tax dollars ultimately pay judgments under [the MTCA]”; “injured litigants who must determine whether compensation for their injuries is barred as a matter of law”; and public employees whose decision to act or not act is informed by their understanding of the duty owed to students.279
Painting with broad brush strokes, the Brum II court’s decision ignores or misunderstands critical elements set out in § 10(j)’s plain language, namely, the creation of the risk.280 Because the true meaning of § 10(j) remains unclear and will continue to elude courts in the future, § 10(j)’s plain language warrants closer examination. What follows is a suggested interpretation of the plain language of § 10(j), including a discussion of the phrases: (1) “act or failure to act [*PG430]to prevent or diminish”; (2) “originally caused by”; (3) “condition or situation”; and (4) “harmful consequences.”
First, contrary to Whitney’s and the MTCA’s across-the-board imposition of liability for negligent acts and omissions, § 10(j)’s initial proposition is that public tort liability will not apply to any claim based upon the public employer’s (or employee’s) act to prevent or diminish harmful consequences, or failure to act to prevent or diminish harmful consequences, unless exceptional criteria apply.281 The “act” and “failure to act” language is derived from the common law public duty rule that the Supreme Judicial Court first employed in Dinsky to relieve public employers from civil liability newly imposed by the MTCA’s abrogation of the sovereign immunity doctrine.282 Under that rule, a public employer was not liable for failing to perform the public duty (here, the duty to prevent or diminish harmful consequences), or for inadequately or erroneously performing the public duty, unless exceptional criteria applied.283 Therefore, the inclusion of this language presumes that immunity will apply to public employers, unless the presumption is rebutted by reference to an exception (i.e., where the public employer or employee originally causes a condition or situation that leads to harm) contained within the provision.284
Second, under § 10(j), a claimed act or failure to act is not actionable unless the harmful condition or situation is “originally caused by” the public employer or employee.285 The phrase “originally caused by” is the most elusive and controversial language contained in § 10(j).286 Until Brum I and Brum II, both the Supreme Judicial Court and the Massachusetts Appeals Court had expressly declined to consider the meaning of “originally caused by.”287 In light of the enduring ambiguity surrounding the phrase and the inconsistency with which the words have recently been interpreted, it is useful to chart the intended meaning of the phrase.288
Under traditional negligence analysis, a plaintiff proves causation by showing that the defendant’s breach of a duty of a reasonable, [*PG431]prudent person in fact caused the harm that came to pass, and more specifically, that the defendant’s breach of a duty was the proximate cause of that harm.289 By contrast, § 10(j) does not look to whether a public employer breached the duty of a reasonable, prudent person because under § 10(j), the public employer’s alleged act or failure to act to prevent or diminish harm implicates a distinctly public duty.290 In addition, § 10(j) looks not to whether the public employer caused the harmful consequence, but to whether the employer caused the condition that led to that harm.291 Thus, § 10(j) necessarily demands an entirely different inquiry into causation than does traditional negligence because the former does not implicate the actual merits of a negligence suit, but rather involves a threshold inquiry into whether a suit may in fact be brought.292 For these reasons, the meaning of “originally caused by” does not rest in common law tort analysis.293 Rather, the “originally caused by” phrase derives its meaning from Justice O’Connor’s concurrence in Cyran (in which the phrase first appeared), the decision by the committee formulating § 10(j) to use this language, the evolution of public duty case law that gave rise to the phrase and the writings of Professor Joseph W. Glannon, who participated in the drafting of § 10(j) and whose writings have greatly informed courts’ interpretation of the provision.294
Section 10(j)’s “originally caused by” language is adapted from Justice O’Connor’s concurrence in Cyran.295 In that opinion, Justice O’Connor stated that the “traditional [common law] public duty rule” distinguished between a public employee’s originally causing a condition or situation that led to the plaintiff’s harm and a plaintiff’s being “harmed by a condition or situation which was not originally caused by the public employee, and is attributable to the employee only in the sense that the employee failed to prevent or mitigate it.”296 The traditional public duty rule, Justice O’Connor further stated, applied only in the latter case.297
The meaning of Justice O’Connor’s phrase “originally caused by,” then, relies upon a negative inference: original cause is not the failure [*PG432]to prevent or mitigate an injury-causing condition or situation.298 Thus, original cause is the act of creating the injury-causing condition or situation.299 To illustrate the meaning of this phrase, Justice O’Connor suggested an example used in Onofrio, involving a police officer’s negligent driving of a cruiser that caused injury.300 Justice O’Connor reasoned implicitly that in such a case, the public duty rule would not be applicable because the police officer did not merely fail to prevent or mitigate the injury-causing condition or situation, but was rather an active participant—and in fact the sole participant—in creating that condition or situation.301 In his concurrence in Jean W., Justice O’Connor further defined the meaning of the phrase by repeating the public duty rule’s distinction between “originally caus[ing]” the harm and “fail[ing] to act to rectify a situation not created by the [public] employee,” and by explicitly comparing the parolee’s originally causing harm in Jean W. to the employee’s “taking action that exposed [another] to risk” in Onofrio.302 Moreover, Justice O’Connor’s concurrence in Jean W. explicitly premised the application of the traditional public duty rule—and thus, the “originally caused by” phrase—upon the distinction between misfeasance and nonfeasance, and between acts and omissions.303 Based upon this time-honored distinction of “long duration,” Justice O’Connor stated that there was to be “no question about it” that the phrase “originally caused by” presumed the misfeasance of a public employee and only misfeasance could overcome the application of the public duty rule.304
Significantly, by invoking the Supreme Judicial Court’s prior public duty jurisprudence in defining the meaning of “originally caused by,” Justice O’Connor implicitly recognized that the misfeasance required to trigger public tort liability necessarily involved a public employee’s creation of the risk.305 Indeed, despite the confusion engendered by the Onofrio line of cases, Justice O’Connor understood those cases to rely upon an inquiry into the creation of risk: where the public employer had actively created the risk of harm that came to pass [*PG433](misfeasance), the public duty rule did not apply, and where the public employer merely failed to prevent an independent, third-party risk of harm from coming to pass (nonfeasance), the public duty rule did apply.306 Justice O’Connor’s articulation of the “originally caused by” phrase in Cyran was therefore an attempt to bring the public duty rule into conformance with this inquiry.307 Furthermore, by adopting both Justice O’Connor’s words as well as his reasoning in drafting § 10(j), the committee formulating the provision necessarily intended to preserve this inquiry into the creation of the risk.308 Thus, from its common law origins to its codification by statute, the phrase “originally caused by” implies the creation of risk.309
In addition to Justice O’Connor’s explication of the “originally caused by” phrase and its codification by statute, the writings of Joseph Glannon, a legal scholar and member of the committee that formulated § 10(j), have been widely cited by courts for their elucidation of § 10(j)’s language and are thus instructive in determining the meaning of the “originally caused by” phrase.310 Consistent with Justice O’Connor’s and the committee’s understanding of the phrase, Glannon notes that the provision’s inclusion of the words “originally caused by” requires some creation of the risk—”something more than the pure failure to alleviate private harm. . . . some involvement of a public employee in creating the initial injury-causing scenario, not simply a failure to respond adequately after it arises.”311 Glannon further states that § 10(j), by adopting O’Connor’s exact language, “embodies the distinction . . . between publicly-created risk”—that is, a public employer’s “creating a risk which causes harm to another” and [*PG434]“the fail[ure] to prevent harm from a private [independent source of] risk.”312 In other words, Glannon adds, § 10(j) should not bar recovery if the public employee “materially contributed” to the creation of the risk that led to injury.313
Third, insofar as “originally caused by” refers to the employer or employee’s creation of the risk of harm, it follows that the “condition or situation” is the actual risk of harm that comes to pass and thus leads to harmful consequences.314 The Legislature’s insertion of the clause, “including the violent or tortious conduct of a third person,” insofar as it modifies the phrase “a condition or situation,” bolsters the view that a condition or situation is the risk of harm that comes to pass, such as a person’s wrongful or otherwise “risky” conduct and from which harmful consequences necessarily arise.315 Furthermore, this definition of the phrase is supported by a memorandum that Senator Jacques of the Senate Judiciary Committee circulated prior to the passage of § 10(j), in which the words “condition or situation” and “risk” were used interchangeably.316
Finally, the American Heritage Dictionary defines “consequence” as “[s]omething that logically or naturally follows from an action or condition.”317 Because the plain meaning of “consequence” presumes a precipitating action or condition, a harmful consequence is that which follows certain risky actions or conditions; it does not, however, “include the precipitating action or condition itself.”318 Under § 10(j), “harmful consequences” should thus be understood as the logical results arising from a preceding risky “condition or situation” or as the natural product of the risk of harm that came to pass.319
Assuming that a plaintiff’s claim against a public employer alleges a negligent act or failure to act to prevent or diminish harmful conse[*PG435]quences, thus bringing it within the scope of § 10(j), the court must next determine whether public duty immunity will apply or whether the public employer is excepted from immunity for creating the risk that came to pass.320 As demonstrated by the prior attempts of both the Supreme Judicial Court and the Massachusetts Appeals Court to avoid defining § 10(j) altogether and more recently by the Massachusetts Appeals Court’s completely erroneous interpretation of § 10(j) in Brum I and the Supreme Judicial Court’s insufficient and dissatisfying explanation of § 10(j) in Brum II, § 10(j)’s plain language presents a formidable obstacle to making this reasoned determination.321 Correct and consistent results depend upon a proper and thorough reading of § 10(j) by the courts, for whom the following three-step analysis should serve as a guide: (1) determine the “harmful consequences” (i.e., death, physical injury, emotional distress, property damage); (2) determine the risk of harm (“condition or situation”) that produced the harmful consequences; and (3) determine whether the public employer or employee’s alleged wrongful acts or inaction created (“originally caused”) that risk of harm that came to pass.322 If a public employee originally causes a condition or situation that leads to a harmful consequence, the public employer should be held liable under § 10(j); if not, the public employer should not be held liable.323 This three-step analysis, although not explicit in either public duty jurisprudence or in current legal scholarship, is consistent with both prior public duty cases interpreting the common law public duty rule as well as recent cases decided under § 10(j).324 Because these cases necessarily inform the future application of § 10(j), an analysis demonstrating § 10(j)’s consistent application to public duty case law will serve as a useful guide for determining the future liability of public employers under § 10(j).325
It is instructive to apply the proposed three-step analysis to prior public duty cases because § 10(j) is the legislative embodiment of the common law public duty rule and should yield results that are consistent with that rule.326 Under Onofrio, where an employee of the Department of Mental Health placed a client in the plaintiff’s home [*PG436]without first warning the plaintiff of the client’s proclivity for starting fires and where the court refused to apply the public duty rule, § 10(j) would probably not apply insofar as the public employee “[took] action that exposed [the homeowner] to risk.”327 Applying § 10(j)’s three-step analysis to Onofrio demonstrates that: (1) the plaintiff’s home was damaged by a fire set by the client; (2) the risk of harm that came to pass was the tenant’s starting a fire in the home; and (3) the employee created the risk of harm by actually placing the client—who had a propensity for starting fires—within the home without any warning to the homeowner.328 In Onofrio, the employee did not merely fail to prevent the risk of harm which came to pass, but rather, the employee’s conduct materially contributed to the risk of harm that came to pass; but for the employee’s placement of the client, the plaintiff would not have been exposed to the risk of the fire that came to pass.329
In Mamulski, where the court refused to apply the public duty rule, § 10(j) would probably not apply to bar liability, insofar as: (1) a motorist was killed in a car accident; (2) the risk of harm that came to pass was a motorist’s failure to stop at an intersection and collision with an oncoming vehicle; and (3) the town created the risk by failing to replace the stop sign that it knew to be missing.330 Although the town technically “failed” to replace the stop sign, this does not mean that such conduct amounted to “a failure to prevent or mitigate” the risk of harm that came to pass.331 Indeed, the risk of harm would not have existed but for the town’s conduct and thus the town would be liable for exposing the motorist to that risk.332
On the other hand, in the case of A.L., where a probationer abused young children and where the court refused to apply the public duty rule because a “special relationship” existed between the town and the children, § 10(j) would likely bar liability.333 In A.L., (1) the plaintiffs were injured as a result of abuse, (2) the risk of harm that came to pass was the probationer’s taking advantage of his proximity to the plaintiffs and perpetrating such abuse, and (3) the town did not create this risk, but rather, the town merely failed to prevent the [*PG437]risk by improperly supervising the probationer’s place of employment.334 Indeed, the nature and extent of the town’s involvement with the probationer’s abuse of the plaintiffs was insufficient to constitute a source or root of that risk of harm.335 Regrettably, but for the employees’ conduct, the plaintiffs would still have been exposed to the risk of harm.336 By contrast, if the public employees had placed the probationer in proximity to the children, such involvement would have constituted a creation of the risk, exposing the plaintiffs to a risk of assault that did not previously exist.
Despite the fact that before Brum I and Brum II neither the Massachusetts Appeals Court nor the Supreme Judicial Court had explicitly interpreted the language of § 10(j), their decisions to apply and to not apply public duty immunity are nonetheless consistent with the three-step analysis under § 10(j).337 In 1994, in Pallazola, the Supreme Judicial Court held that where a football fan was injured by a goal post carried by other fans, § 10(j) barred the fan’s claims against the town based on the town’s failure to provide sufficient police protection and the to prevent the unlawful removal from the stadium of a portion of the aluminum goal post.338 Although the Pallazola court did not definitively interpret the language of § 10(j) before applying public duty immunity (stating merely that the “sort” of claim asserted against the town was foreclosed by § 10(j)), the court’s holding comports with the three-step analysis.339 In Pallazola, (1) the plaintiff was injured by an electrified goal post, (2) the risk of harm that came to pass was the crowd’s removing and carrying a portion of the unwieldy goal post that came into contact with a high voltage power line, and (3) the failure of the town to provide sufficient police protection or to prevent the removal of the goal post in no way “materially contributed” to the risk of harm that came to pass.340 Indeed, the police officer’s conduct was not a source or root of the crowd’s risky behavior, but rather, it was a failure to prevent that risk of harm from coming to fruition.341
[*PG438] Likewise, in Bonnie W., the Supreme Judicial Court held that where the plaintiff was assaulted by a parolee, § 10(j) barred the plaintiff’s claim of negligent supervision, but did not bar the plaintiff’s claim of negligent recommendation and misrepresentation.342 Despite its lack of elaboration on the language of § 10(j), the court’s reasoning is nonetheless consistent with the three-step analysis.343 Where the court held that § 10(j) barred the parole officer’s negligent supervision of the parolee: (1) the plaintiff suffered injury as a result of an assault perpetrated against her in her mobile home; (2) the risk of harm that came to pass was the parolee’s taking advantage of his employee status, gaining access to and assaulting the plaintiff; and (3) the failure of the parole officer to supervise properly the parolee did not materially contribute to the risk of assault of the plaintiff.344 Indeed, if the parole officer never met with the parolee at all, the plaintiff would still have been exposed to the risk of assault; the parolee’s own actions and intentions created the risk of harm that came to pass and the parole officer merely failed to prevent it.345 On the other hand, consistent with the court’s analysis, § 10(j) did not bar the parole officer’s affirmative efforts in support of the parolee’s continued employment, insofar as those acts materially contributed to the risk of assault by actively maintaining the parolee’s employment status.346 This latter point is an extremely important one, for by disallowing the application of § 10(j) immunity to the parole officer’s negligent recommendation and misrepresentation in Bonnie W., the court enabled a public employer to be held liable in cases where the plaintiff is injured by the conduct of a third party.347 The test, consequently, is whether the defendant’s conduct was separate in time and place from the third party’s tortious conduct, or whether the defendant’s conduct was so inextricably linked to the third party’s conduct as to have materially contributed to (originally caused) the risk of harm that came to pass.348
In Lawrence, the Supreme Judicial Court stated in dicta that where a storeowner was assaulted by a third party outside of his store, [*PG439]§ 10(j) barred the storeowners’ claim against the city based upon the failure of the police to protect him from attack.349 Although the court again did not engage in any significant statutory interpretation of § 10(j), the court’s holding is consistent with the three-step analysis insofar as: (1) the plaintiff was shot multiple times; (2) the risk of harm that came to pass was the brutal attack perpetrated by a third-party; and (3) the police officer’s failure to protect the plaintiff, despite knowledge of the risk of assault (the plaintiff had warned the police of such an attack), did not create the risk of attack that came to pass.350 Indeed, the third party’s murderous intent to keep the plaintiff from testifying against him at a grand jury hearing created the risk of attack that came to pass; the officer’s failure to act was independent of that risk.351 If, however, the police had purposefully “arranged” with the third party to be absent from the crime scene on the night of the attack, this conduct would have materially contributed to the risk of assault by affirmatively “placing” the third party in a position to effect that risk of harm.
Furthermore, in a recent Massachusetts Appeals Court decision decided under § 10(j), the appeals court’s reasoning is supported by the proposed three-step analysis.352 In Allen, where a student was fatally stabbed by a second student during an altercation in school, the court denied the city’s motion for summary judgement, holding that § 10(j), in the absence of further evidence, did not bar the plaintiff’s claim against the city based on the school employee’s failure to properly “handle” the student.353 Although the court did not explicitly interpret the language of § 10(j), the court’s holding squares with § 10(j)’s three-step analysis: (1) the plaintiff’s decedent was stabbed and killed; (2) the risk of harm that came to pass was the stabbing of the decedent by a second student as the culmination of an altercation; and (3) because the town did not provide sufficient evidence to the contrary, the town’s failure to take special precautions in enrolling the student and in providing for the safety of others within the school—in view of the student’s known history of weapons violations at other schools—materially contributed to the risk of harm that came [*PG440]to pass.354 This is a close case, but insofar as the school knew of the threat posed by the student and yet did not consider this threat when making the decision to enroll the student and to plan for safety, the school’s “failure” to handle the student properly was analogous to affirmatively “placing” the student in the position to effect the risk of harm.355 Indeed, by actively bringing the student into the school without consideration of his background—much like the parole officer’s negligent recommendation of the parolee’s continued employment in Bonnie W.—the school exposed the plaintiff to a risk of attack that did not previously exist.356
In contrast to its decisions in recent public duty cases, the Supreme Judicial Court in Brum II misconstrues the language of § 10(j), and in so doing, renders much of § 10(j)’s language a nullity.357 By adopting much of the lower court’s confused analysis, the court reaches the “correct” result by the wrong method of analysis, thus circumventing the fine distinctions made by § 10(j) at the expense of a clear and final interpretation of the provision.358 What remains is a statutory provision stripped of distinction, a mere shell of the former § 10(j), and reduced to an inquiry into whether the public employee acted or failed to act.359
Implicitly incorporating nearly all of the appeals court’s reasoning, the Brum II court disagreed with the appeals court only as to its [*PG441]characterization of original cause.360 According to the Supreme Judicial Court, the “killers’ acts” (presumably, the invasion and attack on Robinson) and “the death of their victim” were the harmful consequences.361 The condition or situation that led to the harm was a state of “physical insecurity” within the school, or arguably, was left undefined because it was irrelevant to the court’s analysis.362 Finally, although the court never stated what the original cause was, the court explained that the original cause was not the “neglect of duty” by the school’s employees who both failed to adopt and implement security measures in the first instance as required under § 37H, and subsequently failed to implement interventionist security measures to prevent the attackers from entering the school.363 The court reasoned that to hold the school liable for neglect of a duty or a failure to act, as did the appeals court, would undermine the very basis of § 10(j), which was to immunize public employers for failures to act.364
Despite the correctness of its holding, Brum II’s analysis wholly contradicted the three-step analysis both explicitly and implicitly articulated in prior cases decided under the common law public duty rule and § 10(j), and thus churned the already muddied waters surrounding the plain language of § 10(j).365 Without a clear precedent for deciding subsequent public duty cases involving public schools, lower courts have been left to sink or swim.366 Given Brum II’s inconsistent holding and uncertain implications, several important questions emerge, neatly coinciding with the specific junctures posited by § 10(j)’s three-step analysis.367
Question one: What were Brum II’s “harmful consequences”? Despite the court’s conclusions to the contrary, the harmful consequence in Brum II was Robinson’s fatal injury—not the killers’ acts.368 To answer this question, however, it is first necessary to inquire into the court’s interpretation of “harmful consequences.”369 As mentioned above, harmful consequences refer to the final injury or inju[*PG442]ries that flow from an immediately prior condition or situation (risk of harm that came to pass).370 They do not implicate the risk of harm itself.371 In Brum II, the harmful consequences that the school failed to prevent or diminish were the injury and death of Robinson—not the fatal injury and the invasion and attack that gave rise to that fatal injury.372 This point bears repeating: the invasion and attack on Robinson were not harmful consequences—they were the risk of harm that came to pass.373
Rather than separating the harmful consequence (the fatal injury) from the risk of harm that gave rise to that consequence (the invasion and attack), the court, echoing the arguments of the plaintiff, merges these two distinct junctures into one, stating that “either the killers’ acts or the death of their victim were ‘the harmful consequences.’”374 But does not the third-party invasion and attack constitute a condition or situation giving rise to harm, rather than a consequence itself?375 The court did not think so and supported this proposition by determining that § 10(j)’s clause—”including the violent or tortious conduct of a third person”—could not grammatically or logically modify the words “condition or situation,” and therefore must necessarily modify the words “harmful consequences.”376 Thus, because the youths’ invasion and attack on Robinson was without question violent or tortious conduct, the court summarily concluded that such conduct was a harmful consequence.377
By combining the killers’ acts with the fruition of those acts, the fatal injury of Robinson becomes encapsulated within the killers’ attack itself, as both are rendered indivisible parts of the whole consequence.378 And herein lies the rub: the harmful consequences of a risk of harm that comes to pass must be kept separate from that risk of harm, for the two junctures involve two entirely different inquiries.379 To say that the invasion and the resultant death were but one [*PG443]consequence ignores the fact that one followed “logically or naturally” from the other, and renders the distinction between the two meaningless.380 This distinction is a crucial one, for by collapsing “risk of harm” into “harmful consequence,” the court relegates the true and final harmful consequence (the fatal injury) to an ancillary status that is presumed to exist but only in communion with the other harmful consequence (the invasion and attack).381
In other words, where the invasion and attack is deemed the harmful consequence of a preceding risk of harm, the resultant injury or death of the victim is understood to be only a part of that consequence, as opposed to a separate and independent harmful consequence of the invasion and attack.382 It is the distinction between a motorist’s involvement in a car collision, without determining what harm, if any, was suffered by the motorist; the distinction between a person’s lighting a fire without inquiring into the damages caused by the fire; or the distinction between a parolee’s violation of the terms of his parole by committing abuse, without inquiring into who was injured and the extent of the injury. Harmful consequences encompass the damages sustained, not the risk that gives rise to those damages.383 Thus, in answer to the first question, the harmful consequences contemplated by § 10(j) were indeed the injury and death of Robinson.384
Question two: What was Brum II’s “condition or situation”? Although the court implicitly determined that a condition of “physical insecurity” led to Robinson’s death, the harmful condition or situation in Brum II was in fact the attackers’ violent conduct.385 Indeed, after lumping the invasion and attack together with the fatal injury that resulted from the attack, the court arguably makes a second conceptual leap in deciding that the condition or situation that gave rise to the harmful consequences was that of “physical insecurity” against [*PG444]interlopers at the school.386 Although the court never explicitly defined the particular condition or situation that gave rise to the harmful consequences, it did cite the lower court’s mention of “the initial injury-causing condition of physical insecurity,” and thus implicitly adopted that definition in its own analysis.387 Moreover, this implicit adoption of school insecurity as the harmful condition or situation is more plausible because the court explicitly ruled out the killers’ invasion and attack as a possible condition or situation.388
Presuming that the court intended to adopt the appeals court’s definition of condition as “physical insecurity,” the court’s reasoning falls apart.389 Indeed, under the three-step analysis, it is difficult to see how insecurity could be considered the risk of harm that came to pass.390 The school’s insecurity did not itself lead to harmful consequences of injury or death, but merely preceded in time the risk of harmful conduct (invasion and attack) that subsequently followed.391 Section 10(j)’s statutory language supports this view by modifying the words “condition or situation” with the phrase “including the violent [*PG445]or tortious conduct of a third person.”392 Thus, § 10(j) contemplates that the condition or situation that gives rise to harmful consequences will likely consist of some sort of conduct (i.e., an employee’s affirmative conduct, a third party’s intervening conduct, an injured party’s contributory conduct, or even a fire’s destructive “conduct”) that comes to pass and not from peripheral and attenuated risks (i.e., insecurity) that preceded the harmful conduct.393
Contrary to the Supreme Judicial Court’s analysis, because the tortious conduct of a third party is most reasonably thought of as constituting a risk of harm with potential and independent harmful consequences, as opposed to being itself a mere harmful consequence of some precedent risk, the “tortious conduct” phrase most “logically” constitutes a “condition or situation.”394 In addition, because the “tortious conduct” phrase follows immediately after the words “condition or situation” and not after the word “consequences,” the phrase also “grammatically” constitutes a “condition or situation.”395 Furthermore, if conduct were indeed a consequence, as the court maintains, then what exactly would be the harm arising out of that consequence?396 Section 10(j) says nothing about subsidiary consequences. Therefore, because the “tortious conduct” phrase most reasonably modifies the words “condition or situation” and because the phrase contemplates a risk of harmful conduct, the court’s condition of “physical insecurity” fails for lack of any basis of conduct.397
Thus, rather than identifying the interlopers’ actual conduct of invasion and attack as the risk of harm that came to pass, the court instead traced all preceding risks in connect-the-dot fashion back to the proverbial “first” risk and deemed that risk to be the risk of harm that came to pass—regardless of its attenuation from the actual injury [*PG446]and death.398 Accordingly, the court implicitly determined that “physical insecurity” was the risk of harm that came to pass and that presumably led to the interlopers invading the school wielding weapons and the attack on Robinson and the fatal injuries sustained by Robinson.399 Such an interpretation substitutes attenuated risks for the risk of harm that actually came to pass and that gave rise to the harmful consequences.400 Indeed, under this analysis, one could hypothetically say that the risk of harm that led to the death of Robinson was Robinson’s own involvement in fights with the alleged interlopers that very morning, the previous evening or during the prior week; the interlopers’ acquisition of weapons; or the school employees’ security training.401 Under the court’s analysis, any preceding risk in the chain of events leading up to the harmful consequences is free game for pinning the tail of liability.402 Had the court correctly applied § 10(j), they would have concluded that the risk of harm that came to pass in Brum II was clearly the interlopers’ actual invasion and attack on Robinson.403 Thus, in answer to the second question, it is evident that the condition of such violent conduct—not the condition of insecurity—constituted the “condition or situation” that led to harmful consequences.404
Question three: What was Brum II‘s “original cause”? Although the court completely failed to specify what originally caused the risk of harm, Brum II‘s original cause was the murderous intent of the attackers.405 The court’s analysis of original cause is misguided not only because of what it states, but more importantly, because of what it does not state.406 The court reasoned in the negative that the school’s failure to implement a school security policy and to apprehend the perpetrators on the day of the murder did not originally cause the [*PG447]harm, but never discussed what exactly was the original cause of harm.407 Rather, the court summarily concluded that the “principal thrust” of § 10(j) was to immunize public employers for failing to prevent harm.408 Thus, by characterizing the school officials’ conduct as a “neglect of duty”—defined as “a failure to act to prevent or diminish [harm]”—liability did not apply.409
Although the court correctly identified the legislative intent behind § 10(j), which was to provide “some substantial measure of immunity from tort liability to government employers,” the court went too far in protecting public employers.410 Absent from the court’s analysis is any significant discussion of the creation of the risk, that is, whether or not the school officials’ conduct in fact created the condition or situation of “physical insecurity” that led to the invasion, attack and death.411 All that is stated is that the school neglected a duty and thus failed to prevent Robinson’s death.412 By premising original causation solely upon an action/inaction distinction, however, without regard for the creation of the risk of harm, the court shrank the three-step analysis articulated by § 10(j) into a one-step analysis: whether or not the public employee acted.413 Indeed, the court “remedied” the ambiguity in § 10(j)’s plain language by oversimplifying the provision.414 It may be argued that because a person’s failure to act necessarily precludes that person from originally causing harm, thus eliminating the need for further inquiry into risk of harm and its consequences, only one step is needed or required by § 10(j). This assumption, however, fails on two separate grounds.
The first and most obvious reason that original causation analysis must involve more than simply an action/inaction inquiry is that where a public employee is found to have acted, that does not necessarily mean that § 10(j) immunity should not apply. Indeed, the court must scrutinize whether the public employee’s conduct materially contributed to the harm that came to pass, or whether the employee’s acts were so attenuated as to render the conduct a mere failure to prevent the risk of harm.415 Second, although it is true that a com[*PG448]plete failure to act would necessarily preclude a person from originally causing harm and would thus constitute a failure to prevent the risk of harm, a court cannot be certain that a public employee failed to act unless it first looks to the other steps in the analysis—to the risk of harm and its consequences.416 Without any inquiry into the risk of harm that came to pass, the characterization of an employee’s conduct as either “action” or “inaction” is relative, and is thus left entirely up to the discretion and whim of the court.417 In the words of Chief Justice Liacos, who cautioned against an action/inaction distinction in public duty cases: “[a] standard so flexible cannot provide the certainty and predictability essential to the law.”418
Despite these warnings, the Brum II court, in its attempt to narrowly define the circumstances in which a public employer would be held liable, enabled public employees’ future actions to be recast as harmless inaction by use of the action/inaction distinction.419 Although the court derived the correct result in Brum II because the school officials did not create the risk of harm, the danger inherent in the court’s action/inaction distinction is clearly evident in a host of other scenarios, such as: (1) where a school institutes a policy providing that a school’s floors will be cleaned while students are in class and neglects to put out a “Caution: Wet Floor” sign; (2) where a town changes a two-way street into a one-way street and forgets to post a “One-way” sign; or (3) where a public agency employs a security guard at a housing complex and neglects to check the guard’s criminal record. Despite the action taken in the above examples, the danger is that under the court’s reasoning in Brum II, the court may recharacterize any one of these circumstances as inaction—a “neglect of duty”—and thus, as a failure to prevent or diminish a risk of harm.420 [*PG449]Fearing that a broad interpretation of original cause would allow every failure to prevent harm to be recast as originally causing a harmful condition, “encompass[ing] the remotest causation and preclud[ing] immunity in nearly all circumstances,” the court thus sacrificed a clear interpretation of § 10(j) and its intricacies for an interpretation based solely upon action and inaction.421 In the wake of Brum II, what is left of § 10(j) is a streamlined, albeit more straightforward, shell of the former rule: if public employers do not act, they are not liable.422
Because the court’s analysis of original causation stopped short with the finding that the school officials failed to act, the court never properly addressed why the school officials’ neglect of duty was not an original cause of the harmful condition or situation.423 A proper analysis under § 10(j) must answer this question not by showing that the public employee did not act, but rather, by showing that the public employee did not create the risk.424 In its brief, the town of Dartmouth argued that the source or root of the violent and tortious attack that led to Robinson’s death was the assailants’ murderous intent.425 Indeed, it was this murderous intent of retaliation that created the risk of invasion and attack that came to pass.426 Although the school failed to secure the building properly, the school’s failure to act “was not a source from which the assailants’ murderous intent incepted.”427 Indeed, as demonstrated by the negligent enrollment of the student in Allen, in order for a public employer to be held liable, the school officials’ conduct must have materially contributed—along with the third party’s conduct—to produce the risk of harm that came to pass.428 Thus, had the principal of the school summoned the three [*PG450]youths to the school for a meeting or enrolled the youths in the school despite knowledge of their murderous intent and thereafter a violent altercation had caused injury to the plaintiff, §10(j) immunity would have been inapplicable. In that scenario, the public employees by “placing” the students in the school would have materially contributed to the risk of harm that came to pass.429
Unlike the school officials’ negligent enrollment in Allen, however, the school officials in Brum II did not “place” the third party in the position to effect the risk of harm.430 In Brum II, the youths placed themselves in the position to effect harm when they chose to invade the school, thereby creating the risk.431 More specifically, in Allen, the employees created the risk by actively bringing the third party into the school.432 Were it not for the employee’s negligent action, the third parties would not likely have been there to effect the risk of harm in the first place.433 The same cannot be said of the employees’ inaction in Brum II.434 Because the employees did nothing to invite or deter the interlopers, the third parties may, and just as easily may not, have invaded the school and attacked Robinson.435 Thus, the school’s lack of security, although not making the attack any more difficult, did not affirmatively make the attack any easier by placing the third parties in the position to cause harm.436 In short, whereas the officials’ conduct in Allen was part and parcel of the risk that came to pass, the school officials’ failure to secure the school properly in Brum II was sufficiently removed from the violent attack that took place and could hardly be said to have materially contributed to that attack.437 In fact, much like the parole officer’s failure to supervise the parolee properly in Bonnie W., the school officials’ failure to provide security (of [*PG451]which supervision is but one type) merely preceded an intervening and unconnected risk of harm that followed.438
As Justice Kass argued in his ringing dissent in Brum I, “[n]o school official stabbed Robinson to death” and thus “[i]t requires convoluted reasoning to say . . . that the school authorities originally caused the violent act of [the three youths].”439 Although the authorities might have prevented the killing, the Justice correctly added that it was the three interlopers alone that originally caused the invasion and attack.440 Therefore, under a proper three-step analysis, the court should have concluded that: (1) Robinson’s death was the harmful consequence; (2) the risk of harm that came to pass was the attack perpetrated against Robinson by the invading interlopers; and (3) the school’s failure to secure the school properly and apprehend the perpetrators on the day of the murder did not originally cause the risk.441 Thus, in answer to question three, the murderous intent of the three interlopers—not any act or failure to act on the part of the school officials—originally caused that risk.442
Because of the confusion still surrounding § 10(j) in the wake of Brum II, it is instructive to consider several hypothetical cases within the school context and their likely outcomes under § 10(j). The suggested resolutions of these hypotheticals illustrate the practical application of § 10(j) in harmony with its plain language and legislative intent. More importantly, these resolutions demonstrate § 10(j)’s [*PG452]practical effects on the lives of public employers, school officials, parents and students.
In a hypothetical suggested by the town’s counsel at oral argument, a town would most likely be liable for injury to a child struck by a car on the way to school, if the school decided to start classes earlier in the day while it was still dark and the darkness contributed to the accident.443 In that case, although the child was in fact struck by a third-party driver, the town’s decision to start classes in the dark materially contributed to the risk of a child being injured in the dark; but for the town’s decision, the child would not have been injured.444 By contrast, if the town started classes early and several area youths began throwing rocks at the children attending school, the town would not be liable for injury to the children because its decision did not materially contribute to the risk of students being injured by stones. In that [*PG453]scenario, the youths—not the town—created the risk of harm that came to pass.
In another hypothetical, the town’s counsel suggested at oral argument that if a school principal created a policy stipulating that all school doors must be locked and then forgot to lock them, and if an interloper subsequently made his or her way into the school and attacked and injured a student, the town would be liable for the student’s injury under § 10(j).445 The language and intent of § 10(j) do not support such a conclusion. Indeed, under § 10(j), in order for liability to apply, the public employee must originally cause a condition or situation that leads to harmful consequences.446 Although the principal failed to lock the doors in contravention of school policy, this failure alone did not create the condition that led to the harm any more than if the principal had failed to lock the doors without ever having made such a policy in the first place.447 In both scenarios, the principal’s failure to lock the doors merely coincided with the murderous intent of the third party, and although it may have made the invasion easier, it did not create the risk of injury.448 Thus, in this case, an inquiry into the creation of the risk of harm leads not to the initial failure of the principal to lock the door, but rather, to the third party interloper’s murderous intent.449 Barring collusion on the part of the school or some other affirmative action taken with regard to the invasion and attack on the student, insecurity does not materially contribute to murder, but violent attacks do.450
Professor Glannon offers an additional hypothetical, involving a school official’s negligent supervision of a playground during school hours and the subsequent injury of one student by another student.451 He argues that in this case, although “[t]he direct source of the harm is the other child . . . it cannot be said that this injury is ‘attributable to the employee only in the sense that the employee . . . failed to prevent or mitigate it.’”452 Glannon maintains that because “the child is required to attend the school, and school officials have placed the child in a situation which involves a risk of injury without taking ade[*PG454]quate steps to protect him,” § 10(j) should not apply because the public employee’s negligent conduct materially contributed to producing the injury.453
At the time of Glannon’s writing, the Supreme Judicial Court had not yet decided Bonnie W., and thus, Glannon did not have the benefit of the Supreme Judicial Court’s interpretation of § 10(j) regarding claims of negligent supervision.454 Bonnie W. establishes that a claim of negligent supervision alone is generally insufficient to warrant a finding of liability against public employers—especially where the harm is caused by an intervening third party.455 In this hypothetical, the school official’s failure to monitor the school children adequately did not create the risk of harm that came to pass—that risk of harm was created by the other child’s tortious conduct.456 Unless the school official actively participated in creating the risk of harm to the child, as where an official negligently seats a child in a swing or organizes a dangerous game that results in injury to a child, the negligent supervision constitutes a mere failure to prevent harm and does not materially contribute to the risk of injury.457
In a final hypothetical, plaintiff’s counsel argued at oral argument that the school’s failure to implement security measures and to deter the known and imminent threat of three armed interlopers was comparable to a school official seeing a grenade in a schoolyard sandbox and failing to remove it.458 According to plaintiff’s counsel, liability must be imposed in either case under § 10(j).459 In contrast, the town counsel’s argued that under § 10(j), the town would not be liable “even if [the school official] saw the grenade, failed to remove it, and it exploded,” so long as the school official did not actually [*PG455]place the grenade in the sandbox in the first place.460 Both the plain language and legislative intent of § 10(j) favor the town’s interpretation of the hypothetical because the school official did not cause the grenade to be placed in the sandbox in the first place and thus did not create the risk that presumably led to harm upon explosion of the grenade.461 Rather, the school official failed to prevent the risk of harm from coming to pass, which is specifically excluded from liability under § 10(j).462 As dissatisfying as this result may be, the Legislature did not wish to make public employers liable for the malicious acts of third parties, for to do so would impose too great a burden on the employers whose officers and employees are constantly in contact with private individuals.463
Troubled by the sweeping immunity given to schools under § 10(j), as suggested by the town’s response to the “grenade-in-the-sandbox” hypothetical example at oral argument before the Supreme Judicial Court, Justice Fried pressed the issue of when a town becomes liable for school safety.464 Modifying the hypothetical slightly, Justice Fried suggested that in addition to the grenade in the school sandbox, a sign was attached to the grenade, reading: “This grenade will go off in ten minutes.”465 The Justice added that a school official read the sign, saw the grenade and yet failed to act, at which time the grenade exploded, causing harm to students.466 What result?
The town’s counsel answered that the town would still not be liable, and according to § 10(j), the town was correct.467 Because no school official placed the grenade in the sandbox or otherwise materially contributed to its placement and subsequent explosion, the school officials’ indifference toward the grenade-with-sign constitutes only a failure to prevent an independently-created risk of harm from [*PG456]coming to pass, despite the fact that the school official was on notice of the risk of harm that came to pass.468
Lurking behind Justice Fried’s thankfully farfetched hypothetical are fundamental questions of foreseeability that go to the heart of the Brum II case: what did the school know, if anything, about the impending violence and what did the school do or fail to do in light of such knowledge?469 In short, did the school officials at Dartmouth High School “read the sign”?470 These questions are irrelevant to § 10(j) analysis, which involves only a threshold determination of whether or not a public employer originally caused the condition or situation that led to harm.471 Indeed, questions of foreseeability are dealt with, if at all, only after the court has determined that § 10(j) immunity is inapplicable.472 Thus, even if the risk of harm was in fact foreseeable, so long as the public employer did not create the risk of harm that came to pass, § 10(j) suggests that courts will never reach the issue of foreseeability.473
The Brum I court, however, suggested differently, concluding that the town’s foreseeability of the risk of harm and its failure to act in light of that foreseeability, among other things, created the condition that led to harm.474 As mentioned above, such a decision marks a firm departure from the language and spirit of § 10(j), and the legislative intent to limit public tort liability, to provide predictable results through litigation and to reach a careful balance between compensating injured plaintiffs and maintaining stable government.475 By bringing foreseeability to bear on the application of the public duty rule and by holding the town liable for a complete and utter failure to act, the Brum I court destroyed the rule.476 In light of the troubling resolu[*PG457]tion of the grenade-with-sign hypothetical under § 10(j), however, perhaps the Brum I court was on to something by holding the school liable for an arguably foreseeable risk of harm, but merely approached it the wrong way.477 What § 10(j) needs is a new animus, a new spirit that will enable liability to flow for harms resulting from such indifference; if that does not work, perhaps the Legislature will settle for something more mundane—a new exception.478
In light of Brum II’s resounding denial of liability, public duty jurisprudence and the public itself demand and deserve a new exception to § 10(j).479 Currently, where a school official is aware of a risk of harm and does absolutely nothing to prevent that risk from coming to pass, the public employer may not be held liable because the official did not create the risk of harm.480 This is a proper result under the letter and intent of the provision, but as illustrated by the grenade-with-sign hypothetical, it is not a particularly satisfying result. Despite concern over unpredictable court interpretations and crippling financial burdens on local governments that derive from an affirmative duty of schools to protect, we want school officials to protect students from harm of which they are aware. In addition, despite the importance of public tort immunity to the creative and efficient provision of public services by public servants, who can not possibly prevent every attack nor dissuade every attacker, we do not want school officials to act with indifference toward harm—we want them to act. Thus, in addition to the four exceptions enumerated under § 10(j), the Legislature and the courts should consider a new exception, § 10(j)(5), which whould stipulate that § 10(j) shall not apply to “any claim in which a public employer or any other person acting on behalf of the public employer had actual notice481 of the condition or [*PG458]situation that led to harmful consequences, and acted in deliberate indifference482 to that condition or situation.”483
This two-prong test for imposing liability, requiring (1) actual notice and (2) deliberate indifference on the part of school officials, is not new to the courts and has recently been applied to public employers of public schools in the Title IX context.484 The two-prong test [*PG459]is new, however, to Massachusetts public duty jurisprudence.485 Still, because the test preserves the limited liability of public employers while enabling individual compensation in narrowly circumscribed cases, the two-prong test provides a workable model for a new exception to the statutory public duty rule.486
Indeed, this exception to public tort immunity is exactly the type of legislative solution demanded by Brum II‘s majority because it is aimed at “achiev[ing] a satisfactory result” in the “special category” of school violence cases, yet narrowly tailored to meet the “wide range of circumstances” covered by the provision, including the police context.487 Moreover, this exception mirrors the duty of protection that Justice Ireland, in his Brum II concurrence, determined should be legislatively imposed upon schools.488 Rather than enabling school officials to “figuratively shrug and say ‘the problem did not originate with us, so we are not responsible’” under § 10(j), this statutory exception would obligate schools to “take reasonable measures to protect children when they have advance notice of danger,” as Justice Ireland suggested.489
Over the past twenty years, the Legislature has paid careful attention to the Supreme Judicial Court’s pronouncements regarding public tort liability: after Whitney, the Legislature responded to the court’s prospective abrogation of sovereign immunity by enacting the MTCA, and after Jean W., the Legislature responded to the court’s prospective abrogation of the public duty rule by enacting § 10(j).490 Now, in Brum II, the court has spoken again, voicing its dissatisfaction with § 10(j)’s insulation of public schools.491 In the words of Brum II’s ma[*PG460]jority, the obligation to “achieve a satisfactory result in this special category of case” is “a task for the Legislature,” and in the words of Justice Ireland, “[t]his entire matter is within the control of the Legislature.”492 The court, the commonwealth and its municipalities, parents, teachers and students alike are waiting.
As a policy matter, such an exception would not unduly burden the court’s threshold inquiry into public duty immunity because it would admittedly apply in only a small number of cases in which there are grave concerns over whether or not the school positively knew of the risk of harm and tried to do anything about it.493 In most cases, school officials will either be unaware of the risk that comes to pass until the harm is done, in which case the plaintiff will be unable to prevail under the actual notice prong, or school officials will make some sort of attempt to prevent the harm, in which case the plaintiff will be unable to show deliberate indifference.494 In addition, the exception will not bankrupt local governments in favor of injured parties because, as mentioned above, the exception will not likely apply in the majority of cases.495 Moreover, even where there are reasonable questions pertaining to school officials’ actual notice and deliberate indifference, the plaintiff bears the heavy burden of proving both prongs of the exception, which means showing not that the official should have known of the risk of harm, but rather, that the official did know of the risk and officially refused to attempt to remedy it.496 Furthermore, even if the plaintiff prevailed on both prongs, the claim may still be barred under § 10(b) immunity, and if not, the municipality stands to lose no more than $100,000—the damage cap set by the MTCA.497
In addition, the new exception will provide consistent and predictable results because it is based upon a careful and well-reasoned two-prong analysis, and renders unnecessary courts’ attempts to circumvent § 10(j) in creative and confusing ways on public policy grounds, as the appeals court did in Brum I.498 Additionally, the new exception will advance public policy by providing an incentive for school officials to err on the side of affirmative action when they be[*PG461]come aware of risks of harm, thus resulting in greater safety for school students as well as greater accountability for the failure to act in spite of knowledge of such risks.499 Finally, the new exception will help to ease a fundamental inconsistency between federal and state law: under Title IX, a student who is sexually harassed in school may recover damages from the school, while under state tort law, the family of a child murdered in school receives nothing.500 With the passage of such an exception, school officials would become liable for displaying deliberate indifference to a risk of harm of which they have notice; indeed, school officials would become liable for failing to take any action after reading the sign on the grenade and for failing to do anything to deter the known and imminent threat of three armed intruders.501
This exception embodies the hope that when school officials are aware of a risk of harm to students, they will take action.502 In light of the young man slain in his high school social studies classroom on April 12, 1993, and in light of increasing numbers of students attacked and injured in schools in recent years, this does not seem like too much to ask.503
Since 1978, Massachusetts public duty jurisprudence has been engaged in a balancing of competing social values: “the compensation of injured individuals and the protection of government from financial burdens that threaten its ability to function.” Throughout the years, this balancing has been performed by the Massachusetts courts in concert with the Legislature, who have created and limited [*PG462]the laws enabling public employers both to be sued and to be protected from suit. More importantly, however, this balancing has been informed by the issues raised and the concerns faced by the citizens of Massachusetts, whether they be a sight-impaired child permanently blinded by a defective door at school, or a women sexually assaulted in her home by a parolee.
Today, in the wake of Brum II, the Massachusetts citizenry has a new voice. It is the voice of parents and of children, of teachers and of school administrators, and of federal and state government officials, who fear the tragic balance newly struck by the court in its decision to immunize a school for the stabbing death of its student, despite the school’s notice of the risk of attack and its deliberate indifference toward the attackers. The court has spoken, but the people wait. Now it is the Legislature’s turn to work a new balance between private compensation and public immunity, between the protection of government and the protection of a sixteen-year old high school student stabbed to death in his social studies classroom. It is the Legislature’s turn to cast their eyes and ears upon the people of Massachusetts—to see an injustice that may be averted and to hear stories of violence tinged with hope for change.
There can be no doubting that Brum II was correctly decided under the law, nor can there can be any doubt that the law is wrong. Legal accountability must exist for the failure of school officials to protect students when they have actual notice of the risk of harm and yet demonstrate deliberate indifference toward that risk. There must be change. But change in the law that accounts for unforeseen circumstances, that balances compensation and public immunity in light of evolving social values, must not come in the form of tortured judicial analyses that twist and turn the law to make it fit a particular circumstance. The appeals court tried to effect such change in Brum I and succeeded only in further confusing an already ambiguous provision. Brum II, of course, did not change anything with regard to this balance. In fact, the Supreme Judicial Court’s oversimplification of the law swung the pendulum in the other direction, far away from compensation of potential victims; so long as the conduct of a public employee can be manipulated into a failure to act to prevent harm, the court will dismiss the case. Thus, the interpretive quagmire remains.
In order for change in the balancing of social values to be lasting and meaningful, that change must be tempered by the policy of predictability. This requires both an accurate and consistent reading of § 10(j) by the courts in harmony with the three-step analysis, as well as the formulation of a clearly and narrowly defined exception to § 10(j) [*PG463]by the Legislature, based upon principles of actual notice and deliberate indifference. These proposals are lofty and the stakes are high, but needed change demands carefully chosen and effective means. Indeed, Brum I and Brum II make clear that a faulty vessel in harsh seas will never reach its destination, nor will a vessel that remains tethered to its mooring. This Article’s proposals, however, chart a middle course. The Legislature’s allowance for an exception to § 10(j) will better equip the provision to deal with special categories of cases, while the court’s consistent and correct interpretation of § 10(j) will calm the seas of confusion surrounding public duty jurisprudence. Together, the Legislature and the court may smooth the passage to a change in the balance of competing social values, favoring compensation in special circumstances. This is important because in some cases, public employers should be held liable; indeed, the lives of students and the peace of their families may well depend on it.