by Eric A. Haskell*
This article represents the opinions and legal conclusions of its author and not necessarily those of the Office of the Attorney General. Opinions of the Attorney General are formal documents rendered pursuant to specific statutory authority.
On February 6, 2009, eight days after then-associate Justice Ralph Gants joined the Supreme Judicial Court, the court heard argument in Commonwealth v. Silva-Santiago, an appeal from a murder conviction in which the defendant challenged the reliability of photographic arrays that had led several eyewitnesses to identify him as the killer. Although not apparent at the time, Silva-Santiago marked the first step of an effort that would transform the relationship between scientific knowledge and the law of identification evidence in Massachusetts. That effort was the work of Chief Justice Gants, and it forms a remarkable part of his legacy.
The rudiments of that effort were visible in Justice Gants’s opinion for the court in Silva-Santiago, which was released later that spring. That opinion rejected the defendant’s contention, for which there had been expert evidence at trial, that the identifications were unreliable and should not have been admitted into evidence because the photographs used in the arrays were shown to the eyewitnesses simultaneously rather than sequentially. Citing two law review articles and an article published by the American Psychological Association, Justice Gants acknowledged a “debate among scholars and practitioners [as to] whether the sequential showing of photographs leads to greater accuracy” over a simultaneous showing, and concluded that, “[w]hile that debate evolves,” identifications produced through either procedure would be admissible.
This rationale was both curious and significant. The legal issue in Silva-Santiago was whether the identifications were so “unnecessarily suggestive” as to offend due process. Why look to an academic debate to resolve that legal issue, especially when expert evidence bearing on the answer was present in the record? And why seek conclusiveness in that academic debate before declaring an answer as a matter of law? In retrospect, Justice Gants’s reasoning in Silva-Santiago hinted at his ambition to align the law with the science behind identification evidence.
Two years later, in Commonwealth v. Walker, Justice Gants wrote for the court to again reject the argument that the court had rejected in Silva-Santiago. But Justice Gants’s opinion in Walker also took the next step: characterizing identification evidence as “the greatest source of wrongful convictions but also an invaluable law enforcement tool in obtaining accurate convictions,” it announced that a study group would be charged, among other things, with considering a new model jury instruction on “evaluating eyewitness testimony.
The SJC had adopted a model instruction on identification evidence in 1979, and had periodically modified it thereafter. That instruction exhorted the jury, when evaluating whether the government had proven the defendant’s identity as the perpetrator, to take into account certain abstract and neutral considerations such as the identifying eyewitness’s opportunity to observe the perpetrator, the circumstances surrounding the identification, and the eyewitness’s overall credibility.
The study group created after Walker returned its report in the summer of 2013. The report urged the SJC to take “judicial notice” of certain “psychological principles” concerning the mechanisms of memory and recall, as well as of factors that were said to diminish the reliability of those mechanisms. It also proposed a new jury instruction that, beyond reciting abstract considerations, would instruct the jury as to many of the same scientific principles and factors of which judicial notice was urged.
It is important to appreciate the nature of the study group’s proposal. Juries, of course, deal with science all the time, in the form of expert evidence that the court has deemed likely to be helpful in determining the facts of the particular case. But what the study group proposed was qualitatively different: its proposal was, in effect, to adopt certain scientific knowledge as legal precepts to be applied in all cases. That the scientific principles urged by the study group were well-established in the literature perhaps obscured a lurking tension: while scientific knowledge is factual in nature, iterative, and falsifiable, jury instructions are legal in nature, immutable, and to be accepted by the jury as true.
Justice Gants was promoted in the summer of 2014 and, on September 2 of that year, presided over his first arguments as Chief Justice. Featured on the calendar that day were four appeals concerning aspects of eyewitness identification. Chief Justice Gants wrote the opinion of the court in each of them.
Three of those opinions invoked and relied upon the science urged by the study group. But it was the fourth opinion, in Commonwealth v. Gomes, that transformed the relationship between the science and the law of eyewitness identification evidence, for Gomes presented the issue of what jury instruction ought to be given concerning such evidence.
In Gomes, Chief Justice Gants adopted a highly modified version of the study group’s proposal. The resulting jury instruction, which was appended to the Gomes opinion, continued to exhort the jury to consider things such as the witness’s opportunity to view the perpetrator and the quality of the witness’s perception. But it additionally limned a three-stage scientific “process of remembering,” and identified situation-specific factors—such as “the visible presence of a weapon . . . if the crime is of short duration,” “high levels of stress [felt by the eyewitness], compared to low to medium levels,” and “information the [eyewitness] received between the incident and the identification, as well as after the identification”—that, juries were to be instructed, would diminish the reliability of the identification. Chief Justice Gants explained that it was appropriate to incorporate these precepts into the “judge’s instructions of law, which the jury generally must accept,” because “there is a near consensus in the relevant scientific community . . . .”
The Gomes instruction represented an unprecedented infusion of scientific principles into the judge’s instructions of law. But it could not be said to perfectly align the science with the law because, as noted, science is dynamic and is susceptible of being disproven. Chief Justice Gants was mindful of these limitations, acknowledging that “even a principle for which there is near consensus is subject to revision based on further research findings, and that no principle of eyewitness identification should be treated as if set in stone.” Anticipating the possibility that the principles embodied in the Gomes instruction might be disputed or overtaken by later research, his opinion authorized litigants to offer expert evidence to challenge, and potentially supersede, the instruction. And, acknowledging that, “as the science evolves, we may need to revise our new model instruction,” his opinion reconstituted a committee on eyewitness identification to monitor the development of the science and recommend updates.
The influence of Chief Justice Gants’s efforts to align the law with the science of identification evidence is visible in later SJC decisions that:
- Presumptively required an instruction that “people may have greater difficulty in accurately identifying someone of a different race than someone of their own race,” unless all parties agreed that no such instruction is appropriate;
- Going beyond identification evidence, deemed advances in scientific understanding of the “shaken baby syndrome” as potential grounds for granting a new trial; and
- Looked to “the latest advances in scientific research on adolescent brain development and its impact on behavior” to inform the definition of cruel and unusual punishment vis-à-vis late-teenaged offenders.
Chief Justice Gants’s efforts on this score not only changed the law, they changed the relationship between science and the law in the Commonwealth. As the influence of these changes continues to reverberate, they showcase Justice Gants’s wisdom in recognizing both the promise and the limitations of science in helping to improve justice.
Eric A. Haskell is an Assistant Attorney General whose practice encompasses both civil and criminal matters. He recalls fondly his argument before Chief Justice Gants in Boston Globe Media Partners LLC v. Chief Justice of the Trial Court, No. SJC-12681. That argument lasted approximately forty minutes, despite having been scheduled for fifteen—and it was not the longest argument presented in that case that morning!
*This article represents the opinions and legal conclusions of its author and not necessarily those of the Office of the Attorney General. Opinions of the Attorney General are formal documents rendered pursuant to specific statutory authority.
 453 Mass. 782 (2009).
 460 Mass. 590 (2011) (“[I]t is still too soon to conclude that sequential display is so plainly superior that any identification arising from a simultaneous display is unnecessarily suggestive and therefore must be suppressed.”).
 Commonwealth v. Rodriguez, 378 Mass. 296 (1979).
 Commonwealth v. Cuffie, 414 Mass. 632 (1993); Commonwealth v. Santoli, 424 Mass. 837 (1997); see also Commonwealth v. Pressley, 390 Mass. 617 (1983).
 See generally Mass. G. Evid. § 702.
 Commonwealth v. Crayton, 470 Mass. 228 (2014); Commonwealth v. Collins, 470 Mass. 255 (2014); Commonwealth v. Johnson, 470 Mass. 389 (2015).
 470 Mass. 352 (2015).
 Commonwealth v. Bastaldo, 472 Mass. 16 (2015).
 Commonwealth v. Epps, 474 Mass. 743 (2016).
 Commonwealth v. Watt, 484 Mass. 742 (2020).
by Jamie Michael Charles
In recent years, the Supreme Judicial Court (the “SJC”) has dramatically changed the legal landscape governing the supervision upon release of people charged or convicted of criminal offenses by expanding privacy rights under the Fourth Amendment and Article 14 of the Massachusetts Declaration of Rights. Most recently, in Commonwealth v. Norman, 484 Mass. 330 (2020), the SJC limited a trial court’s ability to impose a requirement that a defendant wear a global positioning system (“GPS”) device as a condition of pre-trial release.
The rubric by which trial courts must assess the propriety of GPS monitoring as a condition of release going forward, and the peripheral ramifications of the SJC’s decision for the various parties to the criminal justice system, have broad implications for client and Commonwealth advocacy. For instance, the decision appears to mark a retreat from prior rulings tacitly approving a broader, safety-based rationale for pre-trial release conditions, and may vitiate other release conditions historically imposed pursuant to the bail statute. In the absence of legislative action, justices of the trial courts must now strike a new balance in their efforts to ensure a defendant’s appearance and protect the various parties to a criminal case. Additionally, both prosecutors and police must revisit their use of surveillance technology, particularly in the absence of legal process, to avoid constitutional violations carrying severe consequences for prosecutions that emanate from such surveillance.
In 2006, the Massachusetts Legislature amended the bail statute, M.G.L. c. 276, § 58, to permit the imposition of various “restrictions on personal associations or conduct” as conditions of release in addition to (or in lieu of) cash bail.[i] In the decade that followed, law enforcement officials, probation officers, and members of the bar operated under the assumption that, where appropriate, the trial court could require a defendant or probationer to wear a GPS ankle bracelet as a condition of release without implicating that individual’s constitutional rights.
This understanding began to unravel in Commonwealth v. Johnson, 481 Mass. 710 (2019), and Commonwealth v. Feliz, 481 Mass. 689 (2019), where the SJC concluded that GPS monitoring of post-conviction probationers amounts to a constitutional search under the Fourth Amendment and Article 14. These decisions were followed in short order by Norman, which extended their holdings to defendants subjected to pre-trial release on conditions. Norman revisited the underlying purposes for which trial courts can impose pre-trial conditions pursuant to the bail statute. In doing so, the SJC left open for question the legality of pre-trial conditions that were routinely imposed on defendants across the Commonwealth before Norman. Furthermore, viewed in combination with Johnson, as well as subsequent decisions in Commonwealth v. McCarthy, 484 Mass. 493 (2020) and Commonwealth v. Mora, 485 Mass. 360 (2020), Norman reflects the SJC’s broader concern with law enforcement’s warrantless use of ever-advancing surveillance technology.
The Norman Decision
In Norman, Medford police were investigating an armed home invasion and armed robbery. They requested a targeted search from the Massachusetts Probation Service’s Electronic Monitoring (“ELMO”) Program, which oversees compliance with GPS release conditions, to determine whether any individuals under GPS supervision were present at the time and location of the robbery. That search yielded a positive result for the defendant, who was subject to GPS monitoring as a condition of release imposed in conjunction with an unrelated drug crime. The ELMO data both placed the defendant at the scene of the robbery and led authorities to a residence where the defendant was shortly before and after the robbery. A search warrant executed at that residence yielded additional inculpatory information. A Superior Court judge allowed the defendant’s motion to suppress, agreeing that the initial procurement of ELMO data absent a warrant violated the defendant’s Fourth Amendment and Article 14 rights. The Commonwealth appealed.
Relying primarily on Massachusetts and United States Supreme Court precedent recognizing an expectation of privacy in the whole of an individual’s physical movements, the SJC found that the initial imposition of GPS monitoring as a condition of pretrial release constituted a search under Article 14.[ii] It held that, going forward, the Commonwealth must demonstrate on the particular facts of each individual case that warrantless monitoring is “reasonable” to justify the attachment of a GPS ankle bracelet. Prior to imposing GPS monitoring, a judge must balance the intrusiveness of the monitoring against any legitimate governmental interests. In light of the “severe intrusion” occasioned by around-the-clock GPS monitoring[iii] – which the justices analogized to a modern-day scarlet letter – the SJC cautioned that even monitoring that serves legitimate government interests could be deemed unreasonable.
The SJC also defined narrowly the legitimate interests served by conditions of release imposed pursuant to the bail statute. It concluded that the only permissible goals of pretrial conditions of release in the defendant’s case were “ensuring the defendant’s return to court and his presence at trial, and safeguarding the integrity of the judicial process by protecting witnesses from intimidation and other forms of influence.”[iv] In doing so, the SJC rejected the Commonwealth’s argument that, unlike cash bail, conditions of release imposed under section 58 may serve the goals of generally deterring criminal conduct or addressing dangerousness concerns (the latter being an interest served under section 58A governing dangerousness determinations). As to the latter goal, the SJC appeared to limit the scope of its earlier ruling in Brangan v. Commonwealth, 477 Mass. 691, 706 (2017), that “a judge may not consider a defendant’s alleged dangerousness in setting the amount of bail, although a defendant’s dangerousness may be considered as a factor in setting other conditions of release.”[v] As a result, prosecutors must now be cognizant that general criminal deterrence and dangerousness are not valid interests in imposing GPS monitoring pursuant to the bail statute in most cases.
Broader Implications of Norman
The reasoning of the SJC’s Norman decision may also limit other types of conditions routinely imposed by the Commonwealth’s trial courts. For example, does ordering a defendant charged with operating under the influence to abstain from drugs or alcohol, or submit to random screens designed to detect those substances, ensure that defendant’s return to court?[vi] Will ordering a defendant repeatedly charged with drug distribution to stay away from the geographical area in which the charges arose protect potential witnesses? There certainly is an argument that these conditions more readily serve the goals of deterrence or community safety, and therefore are inappropriate conditions post-Norman. In fact, the SJC appeared to explicitly disapprove of an exclusionary zone in drug distribution cases absent evidence that the condition was “intended to insulate any particular victims or civilian witnesses[.]”[vii] Without legislative action, judicial officers[viii] may need to reassess the enduring utility of these and other commonly imposed conditions, and explore alternative avenues to cabin a defendant’s pre-trial conduct.
In addition to creating legitimate questions about the legality of certain release conditions, Norman also is one of a string of recent decisions creating a ‘new normal’ for law enforcement use of data generated by electronic surveillance. In Commonwealth v. McCarthy, 484 Mass. 493 (2020), the SJC held that law enforcement review of automated license plate reader (“ALPR”) data maintained by the Executive Office of Public Safety and Security could infringe upon a reasonable expectation of privacy if the data painted a detailed enough picture (or “mosaic”) of the targeted individual’s movements. Likewise, in Commonwealth v. Mora, 485 Mass. 360 (2020), the SJC applied a virtually identical analysis to protracted pole camera surveillance, concluding that a warrant was required where such monitoring enabled investigators to uncover the defendant’s private behaviors, patterns, and associations.
The theme in these cases, as in Norman and Johnson, is the SJC’s recognition that individuals have an expectation of privacy in the whole of their physical movements. And, that regardless of its form, prolonged electronic surveillance can provide “access to a category of information otherwise unknowable.”[ix] Yet while the technologies and data at issue in Norman, McCarthy, and Mora have historically been available to law enforcement in Massachusetts without a warrant, the SJC has thus far declined to set a bright line rule as to when the aggregation of such digital surveillance data crosses the threshold into an Article 14 search.[x] Members of law enforcement must therefore exercise caution when utilizing digital surveillance data that has historically served as a building block for criminal investigations. And prosecutors should seriously consider advising their law enforcement partners to secure a warrant supported by probable cause or another governing legal standard before requesting such data.[xi] Officers who fail to seek legal process run the risk of an ex post facto determination that the aggregation of GPS surveillance data infringed upon a reasonable expectation of privacy, resulting in that data being suppressed “even if law enforcement could have met the applicable [legal] standard.”[xii]
Lessons for Law Enforcement
This new “proceed at your peril” paradigm, aptly demonstrated by the historical GPS data ultimately suppressed in Norman, provides several practical lessons for law enforcement. First, warrantless requests for ELMO data may become less common, as law enforcement must determine whether the quantum of data requested from ELMO will infringe on a potential suspect’s reasonable expectation of privacy such that a warrant is required. Although there was no need for this analysis in Norman (because the ankle bracelet’s initial imposition was itself deemed unreasonable), the SJC recognized in Johnson, 481 Mass. at 727, that, at least as to defendants on probation, there is a difference between “a targeted review of GPS data directed at times and locations of suspected criminal activity” and “mapping out and reviewing all of the defendant’s movements . . . or rummaging through the defendant’s historical GPS location data indiscriminately.” The former may, depending on the specific facts of a particular case, fall short of a constitutional search necessitating probable cause.
Second, while application of the “mosaic” theory may be somewhat more straightforward in the context of ALPRs, pole cameras, and other surveillance technologies, any warrantless request will inevitably be subject to a retroactive assessment of its constitutionality. Such an analysis will consider both the volume of data requested or acquired and what that data reveals about a suspect’s movements, day-to-day routine, political and religious beliefs, and other private affairs. Absent legal process, the pyramid of evidence emanating from such data may collapse upon judicial review and cripple a prosecution.
Third, as to historical ELMO data, even a lawful, warrantless request may be subject to suppression based on a judicial determination that the suspect (whose identity was likely unknown to law enforcement at the time of the request) was improperly subjected to GPS monitoring. Given the virtually unknowable nature of this analysis at the time of an officer’s initial request, law enforcement officials may need to think twice about using ELMO data to find the proverbial needle in a haystack.
Norman and similar decisions addressing the use of digital surveillance – whether during the pre-charge investigatory phase or in conjunction with a criminal prosecution – have fundamentally altered the manner in which police and prosecutors use these technologies. Prosecutors and judges must re-evaluate how to address and constrain the behavior of defendants before and after trial. Conditions of release routinely imposed pre-trial may no longer serve the legitimate government interests vindicated by the bail statute. In particular, GPS monitoring as a condition of release and the use of associated location data by law enforcement have been significantly constrained. Police likewise must reassess technologies that previously served as building blocks of criminal investigations and weigh the utility of their warrantless use against the risk of a court suppressing the resulting evidence. One thing remains certain: as technology inevitably evolves, our appellate jurisprudence will continue to redefine the balance between the tools available to law enforcement and the courts and an individual’s right to privacy.
 See St. 2006, c. 48, § 8 (Mar. 30, 2006).
 While acknowledging that consent can justify a warrantless search, the Court largely dismissed its import in this context given the inherent coercion involved where a defendant agrees to GPS monitoring in lieu of pre-trial incarceration. Norman, 484 Mass. at 335.
 The Court highlighted how faulty alerts and charging issues associated with monitoring devices can compromise an individual’s employment and subject that person to the indignities and dangers associated with an arrest. Norman, 484 Mass. at 339.
 Norman, 484 Mass. at 338. The Court acknowledged that a separate provision of the bail statute permitted conditions of release to be imposed in certain crimes involving domestic abuse “in order to ensure . . . the safety of the alleged victim, any other individual or the community.” See G.L. c. 276, § 58, third par.
 In a footnote, the Court specifically referenced the provision of section 58 authorizing specific restrictions on personal associations or conduct. See Brangan, 477 Mass. at 706 n.18.
 Like GPS monitoring, random drug and alcohol testing constitutes a search and seizure for constitutional purposes under Article 14 of the Massachusetts Declaration of Rights. See Commonwealth v. Gomes, 73 Mass. App. Ct. 857, 859 (2009). Such conditions have been deemed appropriate in the probationary context, where a defendant’s expectation of privacy is reduced, so long as “reasonably related” to the goals of sentencing and probation, in light of the defendant’s underlying crime and his or her particular circumstances. See Commonwealth v. Lapointe, 435 Mass. 455, 459 (2001).
 Norman, 484 Mass. at 338.
 Mass. Gen. Laws. c. 276, § 58 authorizes numerous parties, including justices, clerks and bail commissioners, to set conditions of pretrial release.
 McCarthy, 484 Mass. at 500, quoting Carpenter v. United States, 138 S. Ct. 2206, 2217 (2018); Norman, 484 Mass. at 334, quoting Commonwealth v. Johnson, 481 Mass. 710, 717 (2019). The Court first recognized this privacy expectation in Commonwealth v. Augustine, 467 Mass. 230, 245-49 (2014). Acknowledging that cellular phones had become “an indispensable part of modern American life”, the Court found that “[cell-site location information] raises even greater privacy concerns than a GPS tracking device” as a “cellular telephone is carried on the person of its user, [and therefore] tracks the user’s location far beyond the limitations of where a car can travel.” Id. at 245, 249. The United States Supreme Court reached a similar conclusion in Carpenter v. United States, 138 S. Ct. 2206 (2018).
 McCarthy acknowledged that the absence of a bright-line rule would create “some interim confusion[,]” but expressed confidence that the “constitutional line” would “gradually and appropriately . . . come into focus.” 484 Mass. at 509. The Court cautioned that it “risk[ed] error by elaborating too fully on the Fourth Amendment [or art. 14] implications of emerging technology before its role in society has become clear.” Id., quoting Ontario v. Quon, 560 U.S. 746, 759 (2010).
 Massachusetts case law already recognizes that law enforcement may obtain certain forms of historical location data on a lesser showing than probable cause. For example, in Commonwealth v. Estabrook, 472 Mass. 852, 855 n.4 (2015), the Court concluded that Article 14 permits requests for less than six hours of historical cell-site location information (“CSLI”) on a showing of “specific and articulable facts” evidencing “reasonable grounds to believe” that the records “are relevant and material to an ongoing criminal investigation[.]” Likewise, Chief Justice Gants’ concurrence in McCarthy, 484 Mass. at 514, proposed a “lesser” locational mosaic threshold that would require a showing of “‘specific and articulable facts’ demonstrating reasonable suspicion that the targeted individual has committed, is committing, or will commit a crime . . . and that there are reasonable grounds to believe that the data obtained from the query are relevant and material to an investigation of the crime.”
 McCarthy, 484 Mass. at 514 (Gants, C.J., concurring).
Jamie Michael Charles is an Assistant District Attorney in the Appeals and Training Bureau of the Middlesex District Attorney’s Office.
by Jessica Conklin
In April 2018, the Massachusetts legislature passed the Criminal Justice Reform Act (the “Act”). In addition to enacting sweeping changes in the areas of bail, juvenile justice, diversion from prosecution, and reentry services, the Act established a statutory right to medical parole for all eligible inmates. The medical parole statute, codified at G.L. c. 127, §119A, provides terminally ill and permanently incapacitated prisoners who do not pose a public safety risk the right to be released from custody, regardless of the crime of conviction or the time remaining on their sentence. Until the Act was passed, Massachusetts was one of only a handful of states without this remedy.
Medical Parole Basics
Under the Act, petitioning prisoners who meet the qualifying criteria “shall be released on medical parole.” G.L. c. 127 §119(e) (emphasis added). All inmates, including those sentenced to life without the possibility of parole, have a right to medical parole if they qualify. To be eligible for release, a prisoner must meet three conditions: (i) the prisoner must be terminally ill or permanently incapacitated; (ii) the prisoner must be able to live and remain at liberty without violating the law; and (iii) the prisoner’s release must not be incompatible with the welfare of society.
As to the first condition, the Act defines “terminal illness” and “permanent incapacity,” but does not list specific qualifying illnesses or incapacities. Terminal illness is defined as “a condition that appears incurable, as determined by a licensed physician, that will likely cause the death of the prisoner in not more than 18 months and that is so debilitating that the prisoner does not pose a public safety risk.” Permanent incapacitation is defined as “a physical or cognitive incapacitation that appears irreversible, as determined by a licensed physician, and that is so debilitating that the prisoner does not pose a public safety risk.” The Act does not provide guidance on the second and third conditions; that is, it does not list factors to evaluate an inmate’s ability to live at liberty without violating the law or circumstances that might render an inmate’s release incompatible with the welfare of society.
Procedurally, the prisoner, an attorney, the prisoner’s relative, a medical provider of a correctional facility, or a Department of Correction (“DOC”) staff member may petition the superintendent or sheriff of the facility where the inmate is being held for medical parole on behalf of the inmate. The Act does not prescribe a particular form for the petition. The Act requires expeditious review and a timely decision of medical parole petitions by setting specific deadlines after receipt of the petition. Within 21 days of a superintendent’s or sheriff’s receipt of a medical parole petition, the superintendent or sheriff must provide to the DOC Commissioner (“the Commissioner”): (i) a recommendation regarding release; (ii) a medical parole plan; (iii) a written diagnosis by a licensed physician; and (iv) an assessment of the risk for violence that the prisoner poses to society. G.L. 127, §119A(c)(1). The Commissioner, who is the administrative decision maker, then has 45 days to issue a written decision granting or denying medical parole. If a prisoner’s petition is denied, there is no internal DOC appeals process. The Act allows for judicial review through a petition for certiorari under G.L. c. 249, §4. G.L. c. 127, §119A(g).
2019 Medical Parole Regulations
The Act tasks the secretary of the Executive Office of Public Safety and Security (“EOPSS”) with promulgating rules for administering the medical parole process. G.L. c. 127, §119A(h). EOPSS promulgated regulations in July 2019 (the “2019 Regulations”).
EOPSS took a restrictive view of the scope of the right to medical parole. The 2019 Regulations required the petitioner to develop a medical parole plan and authorized the superintendent or sheriff to reject petitions as incomplete. Under the 2019 Regulations, a complete petition included: (1) an adequate medical parole plan; (2) a written diagnosis by a licensed physician; (3) a release allowing disclosure of the petition and all supporting documents to other criminal justice agencies, the appropriate district attorney, and registered victims or victims’ family members; and (4) a release allowing DOC and the parole board to assess the inmate’s medical parole plan. 501 CMR §§ 17.03(3), 17.03(5). Incomplete petitions required no further action by the superintendent or sheriff. These initial regulations, however, did not stay on the books for long.
The Supreme Judicial Court Weighs In
In early 2020, the SJC invalidated several of the 2019 Regulations as contrary to the plain language of the Act and the legislative intent. Buckman v. Comm’r of Correction, 484 Mass. 14 (2020).
In January 2019, inmates Peter Cruz and Joseph Buckman each submitted a petition for medical parole which was rejected as incomplete by their respective superintendents. Both Cruz and Buckman challenged the decision, arguing that the superintendent must consider a petition regardless of his or her view of completeness or adequacy. When Cruz died in custody during the pendency of the appeal, the case continued with Buckman as the sole plaintiff.
Buckman’s appeal raised three important questions: (1) whether a superintendent must consider a petition for medical parole regardless of the superintendent’s view of the completeness or adequacy of the petition; (2) which party bears the burden of preparing a medical parole plan, obtaining a written diagnosis by a licensed physician, and preparing an assessment regarding the risk for violence the prisoner poses to society; and (3) whether the Commissioner must provide the prisoner with notice of the superintendent’s recommendation, a copy of the recommendation, and any supporting or related materials. Buckman, 484 Mass. at 15-16.
In answering the first question, the court held that a superintendent or sheriff must consider a petition for medical parole regardless of the petition’s completeness. The court noted that the medical parole plan, the written diagnosis by a licensed physician, and medical record releases are documents separate from the petition. As such, those documents are not required to initiate the petition process and trigger the statutory deadlines imposed on the superintendent and the Commissioner. Id. at 25 n.23. The separate nature of these documents is evidenced by the requirement that the superintendent or sheriff – not the petitioner – is required to transmit the medical parole plan, diagnosis, and the risk assessment to the Commissioner with the petition. Id. at 24; G.L. c. 127, §119A (c)(1) and (d)(1). To trigger the Act’s deadlines, the petitioner need not do more than submit a “written” petition. Id. at 26.
On the second question, the SJC ruled that the superintendent or sheriff bears the burden of creating a medical parole plan and obtaining a written diagnosis from a licensed physician. The court reasoned that the Legislature could not have intended to place the burden of expeditiously producing documents on a terminally ill or incapacitated prisoner because the Act only requires the submission of the written petition to trigger the 21-day countdown. Furthermore, because the Act placed the burden of creating the risk assessment on the superintendent, one could infer that the Legislature intended to place the concomitant burden on the superintendent to create the medical parole plan and obtain a diagnosis from a licensed physician. Id. at 25-29.
Finally, the court held that the prisoner must receive all supporting documents submitted by the superintendent except the superintendent’s recommendation to the Commissioner. While nothing in the Act prohibits restricting a petitioner’s right to a superintendent’s recommendation, the court found it fundamentally unfair to prohibit the petitioner from receiving documents that the district attorney could access upon request. In fact, the 2019 Regulations themselves anticipated that the petitioner would have access to the medical parole plan and medical diagnosis because the burden of producing these documents was placed (albeit erroneously) on the petitioner. Id. at 30-32.
Proposed 2020 Regulations
After Buckman, EOPSS began the process of amending its medical parole regulations. The proposed regulations, accessible here, blend new provisions with surviving sections of the 2019 Regulations. At a public hearing on September 16, 2020, lawmakers and advocates criticized the proposed regulations for ignoring the court’s guidance in Buckman, narrowing the population eligible for medical parole, and placing unnecessary roadblocks that delay and frustrate the purposes of the medical parole law.
Some issues flagged by advocates include defining the term “prisoner” to exclude pretrial detainees and individuals who have been civilly committed, construing “permanent disability” to require a higher level of disability than the Act requires, and requiring, as part of the petition, two signed releases on specific DOC issued forms. Although the proposed regulations have not been formally adopted, cases relevant to the proposed regulations are currently before the SJC.
Recent Appellate Litigation
On October 5, 2020, the SJC heard argument in three cases related to the medical parole statute: Racine v. Comm’r of Dep’t of Correction (“Racine”), SJC-12895; Harmon v. Comm’r of Dep’t of Correction (“Harmon”), SJC-12876; and Malloy et. al. v. Dep’t of Correction (“Malloy”), SJC-12961. These cases may answer a number of issues related to the 2019 Regulations, the proposed regulations, and the practical difficulties litigating medical parole cases.
In Harmon and Racine, which were argued jointly, the parties addressed: (1) whether a prisoner’s death renders moot a certiorari action for review of denial of medical parole; (2) whether the EOPSS regulation giving a prisoner the right to reconsideration upon a material decline in health precludes a prisoner from submitting a new petition for medical parole; (3) whether the Act applies only to committed offenders or includes pre-trial detainees; and (4) whether a reviewing court has authority to grant medical parole. The court requested amicus briefing on the first three issues.
On the first question, DOC took the position that death generally renders a case moot. Petitioners argued an inmate’s death (or release) should not moot a case when the issues in the plaintiff’s case are capable of repetition and will otherwise evade review. Petitioners emphasized that the lengthy process to litigate a certiorari action after denial of a medical parole petition will frequently result in plaintiffs dying before their day in court.
On the second question, both the 2019 Regulations and the proposed regulations contain a provision stating that “[n]o subsequent petitions may be submitted following the Commissioner’s denial of medical parole, unless the prisoner experiences a significant and material decline in medical condition.” 501 CMR § 17.14. Petitioners’ counsel took the position that the Act requires the superintendent to review every petition and does not restrict an inmate’s right to file a subsequent petition. DOC argued the Act does not address subsequent medical parole petitions and that EOPSS has the authority to regulate the matter.
On the third question, petitioners argued that the regulations’ exclusion of pre-trial detainees impermissibly narrows the scope of the Act. DOC contends that extending medical parole to pretrial detainees violates the separation of powers.
Finally, in deciding these cases, the SJC also may address the question of whether, on certiorari review, a reviewing court has authority to order medical parole. On this issue, DOC argued judges are limited to remanding a case to the Commissioner for further consideration, while the petitioners argued, among other things, that the certiorari remedy necessarily includes the power to order the medical parole the Commissioner improperly denied.
In Malloy, the court was asked to consider whether a prisoner may continue to be held in custody after the Commissioner has granted medical parole. In that case, two inmates were each granted medical parole without a medical parole plan in place and continued to be detained for weeks while DOC attempted to find a suitable placement. Both the 2019 Regulations and the proposed regulations give the Commissioner the authority to set conditions that must be met prior to the prisoner’s release, a process which may create delay. 501 CMR § 17.11.
Petitioner’s counsel took the position that continuing to hold an inmate in custody after he has been granted medical parole is improper; that the superintendent or sheriff is required to create a comprehensive medical parole plan, including contingency options, within the 66-day window afforded by the Act; and where suitable placement has not been found prisoners should be released to a Department of Public Health facility rather than remaining incarcerated. In contrast, DOC argued the Act does not require immediate release and does not limit the period during which DOC may hold an inmate after medical parole has been granted. Regardless of the outcome, the Court’s decision in Malloy is likely to clarify the timing of an inmate’s right to release under the Act once the Commissioner has decided to grant medical parole.
With inmates facing increased vulnerability during the COVID-19 pandemic, the medical parole statute is particularly important, yet release under the Act has been rare. At the time DOC filed its brief in Malloy, 337 inmates had submitted petitions for medical parole, of which 34 had been granted. Of the 34 inmates granted medical parole, 30 had been released from custody. Three inmates were still awaiting release and one had died after being granted medical parole, but before being released from custody.
Medical parole in Massachusetts is still in its infancy. Its scope, and the procedural mechanisms that govern review of medical parole petitions, will continue to be tested and refined over the coming year.
 The Act defines a medical parole plan as: “a comprehensive written medical and psychosocial care plan specific to a prisoner and including, but not limited to: (i) the proposed course of treatment; (ii) the proposed site for treatment and post-treatment care; (iii) documentation that medical providers qualified to provide the medical services identified in the medical parole plan are prepared to provide such services; and (iv) the financial program in place to cover the cost of the plan for the duration of the medical parole, which shall include eligibility for enrollment in commercial insurance, Medicare or Medicaid or access to other adequate financial resources for the duration of the medical parole.” G.L. c. 127 §119A(a).
 Because affected prisoners are frequently infirm, subject to quick health changes, and usually nearing the end of life, expediting certiorari review is often important. See, e.g., G.L. c. 249, §4 (certiorari petitions must be filed within 60 days); Superior Court Standing Order 1-96(2) (administrative record must be filed within 90 days), Superior Court Standing Order 1-96(4) (certiorari action must be resolved through a motion for judgment on the pleadings served within 30 days of the filing of the administrative record).
 As stated in Buckman, the Legislature enacted the medical parole statute to save money on expensive end of life medical care and for reasons of compassion. 484 Mass. at 21-22.
 Superior Court judges have also tackled issues related to the Act. See Adrey v. Dep’t of Correction, Suffolk Superior Civil No. 19-3786-H, 2020 WL 4347617 (Mass. Super. June 19, 2020); Mahdi v. Dep’t of Correction, Norfolk Superior Civil No. 19-1064, Memorandum and Order (Mar. 31, 2020).
 G.L. c. 127, §119A(g) states: “A decision by the court affirming or reversing the commissioner’s grant or denial of medical parole shall not affect a prisoner’s eligibility for any other form of release permitted by law.” (Emphasis added).
 Brief of Respondent-Appellee Department of Correction, Malloy et. al. v. Dep’t of Correction, SJC-12961.
Jessica Conklin concentrates her practice in white collar criminal defense, government investigations, and school disciplinary hearings. Jessica works with students and their families who attend several local secondary schools, colleges and universities in connection with disciplinary proceedings and title IX investigations. Jessica is also a member of the board of editors for the Boston Bar Journal.
by Jonathan Klavens, Courtney Feeley Karp, and Elizabeth Mason
 As it becomes more challenging to develop large-scale solar projects in Massachusetts, it is worth taking a closer look at “dual use” or “agrivoltaic” projects – solar projects designed with specially elevated and spaced solar panels to allow for continued agricultural use of the land beneath. Some view solar development on “greenfield” sites (open space, forested land, farmland) as less desirable than installing solar on rooftops, parking lots, brownfields, and other previously developed sites. Agrivoltaic projects present an important opportunity to install additional clean energy generation in Massachusetts without the trade-offs often associated with greenfield development. Any solar project where farmland is converted to exclusive solar use gives the landowner the opportunity to supplement farm income by renting out a portion of the land to a solar developer. An agrivoltaic project can provide supplemental income without loss of farmland; it can even lead to the creation of new farmland or more active use of existing farmland, such as upgrading a hayfield to a vegetable farm. An agrivoltaic project can also play a dual role in the fight against climate change: increasing the share of energy generated from carbon-free sources while also promoting regenerative agriculture, the cultivation of plants and healthy soil that can help reduce the atmosphere’s existing carbon load. This article looks at three different regulatory frameworks that impact the development of agrivoltaic projects in Massachusetts: zoning; the Commonwealth’s solar incentive program; and taxation of agricultural land.
Local Permitting of Solar Projects
Like other commercial solar projects in Massachusetts, agrivoltaic projects face an array of permitting requirements. We will focus on the zoning landscape with special attention to several trends and dilemmas.
While state law limits the application of local zoning to solar facilities and both the Land Court and Superior Court have had occasion to interpret that law in recent years, there remains a good deal of confusion about the permissible scope of local zoning authority over solar projects. Zoning regulation of solar projects is limited by M.G.L. c. 40A, § 3 (“Section 3”), which provides that “[n]o zoning ordinance or by-law shall prohibit or unreasonably regulate the installation of solar energy systems or the building of structures that facilitate the collection of solar energy, except where necessary to protect the public health, safety or welfare.” Section 3 evidences the legislature’s intent to protect solar facilities from certain local zoning restrictions but when and to what extent?
Many zoning bylaws do not mention solar energy use (or any broader use that would include solar energy use). Given that zoning bylaws almost universally prohibit uses that are not expressly permitted, this means that in the first instance solar would be prohibited under a bylaw that is silent as to solar use. In turn, however, one Land Court decision held that Section 3 would ordinarily preempt that prohibition, effectively rendering solar use a use allowed by right.
Notwithstanding Section 3, more and more municipalities are adopting solar bylaws that regulate solar projects in one or more ways. Some provide that solar projects are allowed by right in certain zoning districts, with or without a requirement for site plan approval (a mechanism for imposing reasonable conditions on as-of-right uses). Others provide that solar projects are allowed by special permit in certain districts. Still other solar bylaws purport to prohibit solar use in certain districts. Where solar facilities are allowed, a solar bylaw often lays out special dimensional and other requirements, such as requirements for vegetative screening or for posting of financial assurance to cover the costs of removing the facility at the end of its useful life. Larger scale ground mounted solar projects are often the only subject of solar bylaws or are subject to more extensive requirements than other types of solar facilities.
With the proliferation of solar bylaws, questions have arisen about the extent to which they are enforceable in light of Section 3. In its role as reviewer of the legality of new bylaws, the Office of the Attorney General has admonished municipalities to consult with counsel to ensure they do not run afoul of Section 3, but has not rejected any solar bylaw as facially inconsistent with Section 3. In addition, courts have offered some guidance, providing several prospective “rules of thumb” to local zoning boards and solar developers. For example, although a special permit granting authority (“SPGA”) can ordinarily exercise broad discretion to deny a special permit, it likely cannot do so outside the bounds of Section 3. Moreover, certain bylaw requirements (or permit conditions) may be inconsistent with Section 3 on an as-applied basis because they effectively prohibit a project or are not “necessary to protect public health, safety or welfare.” For example, given the benign nature of a typical ground mount solar facility it might be difficult to justify a 200-foot setback requirement as necessary to protect public health, safety or welfare.
It is also unclear under Section 3 under what circumstances a municipality may allow solar energy use in certain districts while prohibiting it in others. There are two keys ways of viewing this issue through the lens of Section 3. One view is that if a municipality allows solar energy use in at least some locations, it cannot be deemed to have “prohibit[ed]” solar use within the meaning of Section 3. The alternative view is that Section 3 bars a municipality from prohibiting solar energy projects even in just a single district “except where necessary to protect public health, safety or welfare.” Id. While an initial Land Court decision seemed to provide some support for the first view, two subsequent clarifying Land Court decisions have endorsed the second.
In Briggs v. Zoning Board of Appeals of Marion, the Marion Zoning Board of Appeals argued that, as long as commercial solar energy use was allowed in some zoning districts, it could still be prohibited in a residential district consistent with Section 3. The court appeared to accept this reasoning, finding that it is “rational” and “reasonable” to prohibit commercial solar energy systems in residential districts, even though Section 3 expressly bars any prohibition of solar energy systems – not just irrational or unreasonable prohibition of solar energy systems – “except where necessary to protect the public health, safety or welfare.” Id. The court noted that the board made no findings on the impact of the proposed project on public health, safety or welfare, id. at *2, nor did the court in its de novo review make any such findings, see id. at *4-5.
In Duseau v. Szawloski Realty, Inc., issued nearly a year later, the court reached a similar conclusion, but only because it determined that the defendant solar developer had the burden of proving that the prohibition of solar energy use in the town’s rural residential district was not necessary to protect public health, safety and welfare and the developer never even argued the issue.
More recently, in PLH LLC v. Town of Ware, the court ruled that a municipality could not require, and then could not deny or condition, a special permit for a solar project in a particular district “except where necessary to protect the health, safety or welfare.” Notably, it appears that no court has yet concluded that a prohibition of solar energy use, or a denial of a permit for solar energy use, has been necessary to protect public health, safety or welfare under Section 3. Given that many larger solar projects are now operating (including many in residential districts) across Massachusetts, and that many municipalities that have hosted such projects are supportive of additional solar development, it seems likely that the parties to future litigation on this point will have a good deal of experience from which to draw.
In short, developers of larger solar projects must navigate local land use regulation and differing interpretations of Section 3 as to which aspects of local regulation are actually enforceable. Meeting the Commonwealth’s clean energy goals will likely require more balanced regulation and more certainty about how municipalities can lawfully regulate clean energy projects.
SMART Program Incentives for Agrivoltaic Projects
The Solar Massachusetts Renewable Target Program (the “SMART Program”) implemented by the Massachusetts Department of Energy Resources (“DOER”) provides a base financial incentive for production of each unit of solar energy from eligible projects in Massachusetts. The SMART Program regulations also offer extra incentives known as “adders” to promote certain types of projects, such as solar carports, solar on landfills, and community solar facilities. Agrivoltaic facilities, referred to as Agricultural Solar Tariff Generation Units (“ASGTUs”) in the regulations, are the target of one such adder. In addition to providing adders for preferred project types, the SMART Program also has what are called “greenfield subtractors” which reduce the incentive payments for solar facilities located on greenfield sites. ASTGUs are not subject to the subtractor given that the land on which they are located will continue to be farmed. ASTGUs are also exempt from strict new rules adopted in July 2020 that generally bar solar facilities from participating in the SMART Program if they are located on land designated as priority habitat, core habitat, or critical natural landscape as identified by the Massachusetts Division of Fisheries and Wildlife BioMap2 framework with the Commonwealth’s Natural Heritage and Endangered Species Program.
At the same time, the process and standards for qualification of a SMART ASTGU are quite rigorous under state regulations and guidelines. For example, the reduction of direct sunlight relating to an ASTGU cannot exceed 50% – measured on every square foot of the project site. While a SMART facility can generally be up to 5 MW AC in size, under the current ASTGU Guideline an ASTGU would typically be capped at just 2 MW AC. Id. at 3. The current regulations also contain a number of other requirements including continuous growth, growing plans, and productivity reports. 225 C.M.R. 20.06(1) (d)(3), (5); ASTGU Guideline at 3. While it is important to ensure that there are not significant detrimental effects on agriculture from an ASTGU, there could be many appropriate reasons for reduced productivity, such as a drought year or appropriate crop rotation. The approval process thus far has raised questions about the appropriate baseline for measuring impacts, determining which impacts to attribute to the solar facility or to other causes, what type or magnitude of impact would result in disqualification of an ASTGU or removal of its adder.
There may well be many more types of symbiotic solar and agricultural uses that do not fit within the current requirements for ASTGUs. For example, mushroom cultivation, beekeeping and animal husbandry are all farming activities that might benefit from shade reduction greater than 50%. The state’s Department of Energy Resources (“DOER”) has a process for seeking waivers for unique and worthwhile alternatives but obtaining an exception is not easy, quick or predictable.
Based on experience gleaned from processing ASTGU applications for almost two years, DOER has recently issued a “straw proposal” to modify the guideline governing qualification of ASTGU projects. Among other things, the straw proposal raises the possibility of allowing for ASTGUs of up to 5 MW AC in certain instances and streamlining the approval process by permitting qualification by a third party organization, which should increase speed and predictability for approval of project designs. This change would provide greater certainty for the financing of these projects and allow the full range of potential climate change benefits to come to fruition.
Property Tax Incentives for Land in Agricultural/Horticultural Use
Land in active agricultural or horticultural use is entitled under M.G.L. c. 61A (“Chapter 61A”) to reduced property tax rates. Chapter 61A land that is converted from agricultural to commercial use must be removed from Chapter 61A. So what happens when Chapter 61A land serves as the site of an agrivoltaic facility?
Before land is to be removed from Chapter 61A, the landowner must deliver to the municipality a notice of intent to convert. Such notices are accompanied by plans showing the total acreage that will cease to be farmed (the “Converted Land”) and the balance of the land that will continue to be farmed (the “Remaining Land”). The Converted Land is removed from Chapter 61A and the landowner pays roll-back taxes (and, if applicable, conveyance taxes) in connection with this removal. The Remaining Land should remain eligible for reduced taxation under Chapter 61A.
There is currently some confusion about the applicability of Chapter 61A to land under agrivoltaic facilities in light of the existence of Section 2A of Chapter 61A. Section 2A was inserted by the legislature in 2016 to address situations where ground mounted solar facilities are installed on farmland, precluding use of the land under the solar panels for agricultural or horticultural use but generating power used for the operation of the farm. Section 2A allows owners of agricultural or horticultural land who install a “renewable energy generating source” on their land which meets the requirements of Section 2A to maintain all of their land as agricultural or horticultural land under Chapter 61A, even the land that is exclusively occupied by the solar array and can no longer be farmed. Section 2A is not relevant to agrivoltaic facilities because they involve installation of solar panels above land which will continue to be farmed.
Land under and around an agrivoltaic facility is instead governed by Sections 1 and 2 of Chapter 61A. Section 1 states that land shall be considered to be in agricultural use when “primarily and directly used in raising animals, including, but not limited to, dairy cattle, beef cattle, poultry, sheep, swine, horses, ponies, mules, goats, bees and fur-bearing animals, for the purpose of selling such animals or a product derived from such animals in the regular course of business.” Section 2 states that land shall be considered to be in horticultural use when “primarily and directly used in raising fruits, vegetables, berries, nuts and other foods for human consumption for the purpose of selling these products in the regular course of business.” The Remaining Land at the site of an agrivoltaic facility, which will continue to be farmed, meets these definitions.
Note that farming the land underneath and surrounding the solar arrays of an agrivoltaic facility is something that, as noted above, facility owners are required to do under the SMART Program in order for the facilities to qualify for – and stay qualified as –ASTGUs under that program. If in the future the owner ceases farming the land underneath and surrounding the solar arrays and uses it for a non-qualifying purpose, the land would then lose eligibility for classification under Chapter 61A.
Chapter 61A and the publications of the Massachusetts Department of Revenue’s Division of Local Services (“DLS”) are clear that it is the use of the land that determines whether or not land is eligible for classification under Chapter 61A. Section 20 of a set of FAQs published by DLS states that, in the case of solar facilities that (like the agrivoltaic projects discussed here) don’t meet the requirements of Section 2A, only land “necessary for the operation of” the solar facility or “impacted by its operation” is ineligible for continued classification under Chapter 61A. The Converted Land at the site of an agrivoltaic facility meets this definition and is the portion of the land no longer eligible for taxation under Chapter 61A. The Remaining Land is not “necessary for the operation of” the solar facilities. It will continue to be farmed and should remain eligible for classification under Chapter 61A.
Whether land under and around an agrivoltaic facility can remain in Chapter 61A can have a significant impact on the economic viability of an agrivoltaic project. If land under an agrivoltaic project is not allowed to remain in Chapter 61A, that may not just mean that the project would have to be able to support higher property taxes (potentially reducing benefits to the farmer) but could also raise questions about the project’s ability to operate as an ASTGU under the SMART Program. An agrivoltaic project can qualify as an ASTGU if it is on land currently enrolled in Chapter 61A or land that has been in Chapter 61A in the previous five years. 225 C.M.R. 20.02 (definitions of ASTGU and Land in Agricultural Use). If a project also needs a waiver under the ASTGU Guideline, however, it must demonstrate that “the primary use of the land is for agricultural or horticultural production, as defined under [Chapter 61A].” ASTGU Guideline at 4. If the land is removed from Chapter 61A because of hosting the ASTGU, the rationale for such removal would presumably be that the primary use is no longer agricultural or horticultural. This would create tension rather than synergies between laws, and would highlight the importance of interagency coordination to further the Commonwealth’s policy goals, particularly with respect to climate change. Removing any uncertainty about this issue will be important to the growth of agrivoltaic facilities and the environmental and economic benefits that flow from them.
Development of larger scale solar projects is a challenging venture, and development of agrivoltaic projects involves special challenges and special opportunities. Overcoming those challenges and realizing those opportunities requires harmonization of and certainty in land use regulation, financial incentive qualification, and property taxation. Striking the right balance would be a victory for sensible land use planning, support of local agriculture, and the transition to a clean energy future.
 The authors would like to thank Sarah Matthews, senior counsel at Klavens Law Group, P.C., and Jaidyn Jackson, law student intern at Klavens Law Group, P.C., for their valuable contributions to this article.
 See Waller v. Alqaraghuli, No. 17 MISC 000233, 2017 WL 3380387, at *4 (Mass. Land Ct. Aug. 4, 2017) (Scheier, J.). Although Section 3 does allow regulation of solar facilities as “necessary to protect the public health, safety or welfare,” in the case of a local zoning bylaw whose prohibition of solar use is preempted, a local zoning board cannot then choose to regulate solar use “by a case-by-case determination by the Board.” Id. at *5 n.7.
 See, e.g., Letter from the Office of the Attorney General Municipal Law Unit to Town of Plympton, at 2 (Apr. 3, 2020) , https://massago.onbaseonline.com/MASSAGO/1801PublicAccess/mlu.htm) (input Case Number “9750”; then click “Search”; then follow hyperlink) (advising Town that, [i]n applying [solar bylaw amendments] the Town should consult closely with Town Counsel to ensure that the Town does not run afoul of [Section 3]”).
 See PLH LLC v. Town of Ware, No. 18 MISC 000648, 2019 WL 7201712, at *3 (Mass. Land. Ct. Dec. 24, 2019) (Piper, C.J.) (holding that zoning bylaw may require a special permit for solar energy use in a particular district but special permit review “must be limited and narrowly applied in a way that is not unreasonable, is not designed or employed to prohibit the use or the operation of the protected use, and exists where necessary to protect the health, safety or welfare”); cf. Waller at n.7 (suggesting that municipal authority under Section 3 to regulate solar use as necessary to protect public health, safety and welfare can only be exercised in crafting a generally applicable bylaw, not to justify case-by-case determination with respect to particular projects).
 See, e.g., Ayotte v. Town of Cheshire Planning Board, CA No. 17-275, slip. op. at 9-13 (Mass. Sup. Ct. May 4, 2018) (Ford, J.) (refusing to uphold planning board’s denial of special permit for solar project based on concerns about solar glare and inadequate screening and remanding to the board “for the consideration and imposition of any reasonable conditions”) (emphasis in original).
 Briggs v. Zoning Board of Appeals of Marion, No. 13 MISC 477257, 2014 WL 471951 at *4 (Mass. Land Ct. Feb. 6, 2014) (Sands, J.).
 Duseau v. Szawloski Realty, Inc., Nos. 12 MISC 470612, 12 MISC 477351, 2015 WL 59500 at *8 (Mass. Land Ct. Jan. 2, 2015) (Cutler, C.J.); PLH LLC, 2019 WL 7201712 at *3.
 Briggs, 2014 WL 471951 at *4.
 Duseau, 2015 WL 59500 at *8 & n.11.
 PLH LLC, 2019 WL 7201712 at *3.
 See 225 CMR 20.07(4)(g); Mass. Dep’t of Energy Resources, Guideline Regarding Land Use, Siting, and Project Segmentation at §§ 3, 4(b) (revised Oct. 8, 2020) (the “Land Use Guideline”).
 See Land Use Guideline, §§ 5(4)-(5).
 See 225 CMR 20.02 (definition of Agricultural Solar Tariff Generation Unit) and 20.06(1)(d) (eligibility requirements); Guideline Regarding the Definition of Agricultural Solar Tariff Generation Units (the “ASTGU Guideline”).
 In addition, Sections 14(A) and (B) of the FAQs state that any roll-back and conveyance tax are to be assessed “only on that portion of the land on which the use has changed to the non-qualifying use.”
Jonathan Klavens is the principal of Klavens Law Group, P.C. He practices across the fields of corporate, land use and environmental law, with a special focus on the development, financing and purchase and sale of clean energy projects, as well as the formation, financing and ongoing support of cleantech companies.
Courtney Feeley Karp is senior counsel at Klavens Law Group, P.C. where she advises clients on development and compliance matters for clean energy projects, including those located on agricultural land. Previously she served as counsel at the Massachusetts Department of Energy Resources and the Massachusetts Senate Ways & Means Committee.
Betsy Mason is senior counsel at Klavens Law Group, P.C., where she focuses her practice on resolving real estate, land use and permitting, and environmental compliance issues arising during the development, construction and acquisition of renewable energy projects. Her past positions have included, among others, in-house counsel for real estate and business development at a leading national solar developer and Senior Assistant Regional Counsel at EPA Region 1 in Boston.
Where a company chooses to incorporate directly affects the fiduciary duties imposed upon its leadership and shareholders. Under the internal affairs doctrine, the law of the state in which a company is incorporated applies to disputes over the company’s internal workings, regardless of where the company is actually based or where the alleged conduct occurred. Consequently, the distinctions between Massachusetts and Delaware corporate law take on particular importance in the context of closely held corporations and LLCs, where each state’s case law and statutes uniquely impact the imposition of fiduciary duties and the extent to which such duties can be contractually waived. Understanding these differences is essential to making informed decisions on business formation and litigation strategy.
Fiduciary Duties In Closely Held Corporations
In both Massachusetts and Delaware, a corporate fiduciary, such as a director, generally owes a duty of care and a duty of loyalty, both of which impose a responsibility to act in the best interests of the corporation and/or its shareholders. Specifically, the duty of care requires a fiduciary to act in an informed and reasonable manner, and the duty of loyalty requires a fiduciary to act in good faith with the primary intent of furthering the best interests of the company. In the context of closely held corporations, including LLCs, Massachusetts and Delaware take divergent approaches as to whether shareholders may owe each other additional or enhanced fiduciary duties.
In Donahue v. Rodd Electrotype Co. of New England, the Massachusetts Supreme Judicial Court held that a “close corporation” is typically one in which there is “(1) a small number of stockholders; (2) no ready market for the corporate stock; and (3) substantial majority stockholder participation in the management, direction and operations of the corporation.” The Donahue Court then held that “[s]tockholders in close corporations must discharge their management and stockholder responsibilities” under a standard of utmost good faith and loyalty.
Therefore, a controlling group of shareholders in a close corporation “may not, consistent with its strict duty to the minority, utilize its control of the corporation to obtain special advantages and disproportionate benefit from its share ownership.” Even if the majority shareholders demonstrate a legitimate business purpose for their actions, minority shareholders may still maintain a claim if they can establish that the same business purpose could have been achieved in an alternative manner less harmful to the minority. Importantly, minority shareholders in Massachusetts close corporations also owe fiduciary duties and may not intentionally engage in corporate conduct to their own personal advantage that is detrimental to the corporation and its other shareholders.
In Massachusetts, where a minority shareholder has a reasonable expectation of continued employment within a closely held company, the termination of such minority shareholder’s employment may be considered a “freeze-out” and, as such, be a breach of the majority shareholders’ fiduciary duties. In this regard, the remedy for a breach of a majority shareholder against the minority shareholder is, to the extent possible, to “restore to the minority shareholder those benefits which she reasonably expected, but has not received because of the fiduciary breach.” Massachusetts courts have the ability to award monetary damages, as well as impose a wide range of equitable remedies, such as reinstating corporate officers, forcing the distribution of dividends or even amending corporate operating agreements.
In contrast, a closely held Delaware company must be specifically incorporated as a “close corporation.” Even when incorporated as such, shareholders generally will not owe each other fiduciary duties unless the articles of incorporation or an agreement among the shareholders imposes such duties.  Consequently, the “protections afforded to minority stockholders in closely-held corporations under Delaware common law are no different than those in publicly-held corporations.” 
Majority shareholders in a closely held Delaware corporation, or shareholders who otherwise exert control over the close corporation, may act in a manner that unintentionally harms the interests of minority shareholders as long as such actions do not contravene the best interests of the corporation itself. The same holds true in the context of Delaware LLCs, except that managing members still owe fiduciary duties to the company and its passive members. However, if majority/controlling shareholders are found to have engaged in a transaction in which they would uniquely benefit, such shareholders must demonstrate that the transaction was entirely fair to the corporation, in terms of both price and dealing, and conducted with the utmost good faith. In Delaware, remedies for such breaches usually are limited to either monetary damages or equitable rescission of the contested transaction.
Contractually Limiting Fiduciary Duties
Contracts, which govern shareholder conduct, may supersede common law or statutory fiduciary duties. In Massachusetts, “[a]lthough a shareholder in a close corporation always owes a fiduciary duty to fellow shareholders, good faith compliance with the terms of an agreement entered into by the shareholders satisfies that fiduciary duty.” Consequently, claims for breach of fiduciary duty with respect to things such as employment or stock purchases may only arise when a prior agreement among the parties “does not entirely govern the shareholder’s actions.” To the extent an agreement does not directly address an issue, fiduciary duties apply.
In the LLC context, Massachusetts allows for greater flexibility regarding the limitation of fiduciary duties. Massachusetts Gen. Laws c. 156C, § 63, specifically provides that “[t]o the extent that, at law or in equity, a member or manager has duties, including fiduciary duties, and liabilities relating thereto to a limited liability company or to another member or manager . . . the member’s or manager’s duties and liabilities may be expanded or restricted by provisions in the operating agreement.” Additionally, “[t]he certificate of organization or a written operating agreement may eliminate or limit the personal liability of a member or manager for breach of any duty to the limited liability company or to another member or manager.” Contractual language limiting or modifying fiduciary duties “should be strictly, not expansively, construed.”
In JFF Cecilia LLC v. Weiner Ventures, for example, the plaintiff and defendant were members of a Massachusetts LLC that was primarily formed to develop a large real estate project. The plaintiff asserted claims based on the defendant’s failure to provide notice that it was publicly announcing the canceling of the development project.  The Superior Court recognized the enforceability of a clause in the LLC agreement, which stated that its members owed no fiduciary duties to the company or each other “except to the extent such duties are expressly set forth in this Agreement.” The court, however, found that the plaintiff still had a cause of action based on another section of the agreement, which imposed a duty upon its members to “consult with one another openly, fairly and in good faith,” to “work collaboratively” and to “use their reasonable efforts to keep one another informed of all known and material information with respect to” the company. Likewise, in Butler v. Moore, the relevant LLC agreement stated that its members were not “obligated to present an investment opportunity to the Company even if it is similar to or consistent with the business of the Company. . . [and they had the] right to take for their own account or recommend to others any such investment opportunity.” Nonetheless, the court refused to interpret such provisions as allowing its members to take, “for their own personal benefit,” those opportunities that already had been presented to the company. 
Similar to Massachusetts, in Delaware where a corporate “dispute arises from obligations that are expressly addressed by contract, that dispute will be treated as a breach of contract claim . . . [and] any fiduciary claims arising out of the same facts that underlie the contract obligations would be foreclosed as superfluous.” Likewise, Delaware statutorily allows provisions in LLC agreements in which a member or director’s fiduciary duties are “expanded or restricted or eliminated.” For example, in Marubeni Spar One, LLC v. Williams Field Servs. – Gulf Coast Co., L.P., the Chancery Court dismissed a claim for breach of fiduciary duty because the LLC agreement stated that the “Operating Member shall have no liability under this Agreement or otherwise to the Company or any Member for any actions taken in its capacity as Operating Member or for any actions it fails to take unless it breaches its obligations under this Agreement as a result of its gross negligence, fraud or willful misconduct.”
Although parties to a Delaware LLC agreement are not allowed to waive “the implied contractual covenant of good faith and fair dealing,” even this limitation is narrowly applied. As the Chancery Court explained, when an LLC “agreement eliminates fiduciary duties as part of a detailed contractual governance scheme, Delaware courts should be all the more hesitant to resort to the implied covenant. . . . “[r]especting the elimination of fiduciary duties requires that courts not bend an alternative and less powerful tool into a fiduciary substitute.”
Both Massachusetts and Delaware allow closely held corporations and LLCs to indemnify their directors and managers for any defense costs, settlements or judgments, which might arise from an alleged breach of their fiduciary duties. Pursuant to Mass. Gen. Laws c. 156D, § 8.51, a corporation may indemnify a director against any liability as long as the director: (1) has acted in “good faith;” (2) “reasonably believed that his conduct was in the best interests of the corporation or that his conduct was at least not opposed to the best interests of the corporation;” and (3) “had no reasonable cause to believe [the ] conduct was unlawful.” Massachusetts law also mandates that a corporation indemnify a director for their defenses costs if such director “was wholly successful, on the merits or otherwise, in the defense of any proceeding to which he was a party.”
In the context of LLCs, Massachusetts law provides even wider latitude in determining the scope of indemnification. Indeed, the burden of denying indemnification may be placed on the company. LLCs can indemnify their members and managers “against any and all claims and demands whatsoever,” without having to establish that the member actually acted in good faith or with any particular intention or knowledge. An LLC may not provide any form of indemnification if the member or manager is specifically found in the course of a proceeding “not to have acted in good faith in the reasonable belief that his action was in the best interest of the limited liability company.”
In Delaware, the circumstance under which a corporation may indemnify a director is similar to Massachusetts in that the person must have acted in good faith with the reasonable belief that he, she or they were acting in the best interest of the corporation, and with no reason to believe his, her or their conduct was unlawful. Delaware corporations are required to provide indemnity for defense costs incurred by qualified individuals who succeed on the merits of their case. Delaware provides additional statutory guidance to determine when a person qualifies for indemnification. In this regard, the termination of an action by judgment, order or settlement (except where such judgment is based on a guilty plea) “shall not, of itself, create a presumption that the person did not act in” a manner entitling him, her or them to indemnification. A company cannot provide any indemnification if a person has “been adjudged to be liable to the corporation unless and only to the extent that the” relevant court makes a separate determination that despite such judgment the person is still entitled to indemnification.
In Delaware, “[n]o criteria are established by statute to govern the indemnification that limited liability companies may offer.” Pursuant to 6 Del. C. § 18-108, “a limited liability company may, and shall have the power to, indemnify and hold harmless any member or manager or other person from and against any and all claims and demands whatsoever.” The statute “defers completely to the contracting parties to create and delimit rights and obligations with respect to indemnification and advancement,” and does not in itself create any right or limits to indemnification.
In the context of closely held corporations, Massachusetts and Delaware take somewhat divergent approaches to the imposition and waiver of fiduciary duties among shareholders. Shareholders in closely held Massachusetts companies owe each other heightened fiduciary duties, and such duties can only be eliminated by contractual agreements that specifically address the duties or situations under which such duties may arise. Consequently, founders of, or investors in, closely held companies who seek heightened protections against the potential misconduct of their fellow shareholders, may favor Massachusetts incorporation. In contrast, shareholders in closely held Delaware companies generally do not owe each other fiduciary duties, and Delaware takes a more expansive approach with respect to allowing the elimination of fiduciary duties that could be imposed upon such shareholders. Accordingly, parties whose priority is to avoid potential liability, or the threat of such liability, from their fellow shareholders, may prefer Delaware incorporation.
In either state, potential shareholders should carefully review any contractual arrangements among the shareholders, including by-laws, purchase agreements and employment contracts. Such contracts may supersede, modify or eliminate what would otherwise be the parties’ default fiduciary duties. Awareness of these key distinctions between Massachusetts and Delaware corporate law is essential to making informed decisions about incorporation, investment and litigation.
 Harrison v. NetCentric Corp., 433 Mass. 465, 471 (2001). The majority of states adhere to the internal affairs doctrine. Notably, California and New York have particular statutory exceptions to the doctrine, which require consideration of whether the company has substantial contacts with the forum state. See Cal. Corp. Code § 2115; N.Y. Bus. Corp. L. §§ 1317–20.
 Donahue v. Rodd Electrotype Co. of New England, 367 Mass. 578, 586 (1975) (holding that the categorization of company as a “close corporation” is a factual inquiry); Allison v. Eriksson, 479 Mass. 626, 636 (2018) (in determining whether an LLC is closely held company, a court also examines the manner “in which a particular LLC is structured”).
 Id. at 593 (1975); Butler v. Moore, No. CIV. 10-10207-FDS, 2015 WL 1409676, at *61 (D. Mass. Mar. 26, 2015) (“As a matter of logic and fairness, there is no reason why the fiduciary duties of members of a closely held LLC should be materially different from those of shareholders of a closely held corporation.”); Blank v. Chelmsford Ob/Gyn, P.C., 420 Mass. 404, 408 (1995) (“They may not act out of avarice, expediency, or self-interest in derogation of their duty of loyalty to the other stockholders and to the corporation.”).
 Donahue, 367 Mass. at 598.
 Wilkes v. Springside Nursing Home, Inc., 370 Mass. 842, 851-52 (1976).
 Donahue, 367 Mass. at 593; Zimmerman v. Bogoff, 402 Mass. 650, 658 (1988)
 Selmark Assocs., Inc. v. Ehrlich, 467 Mass. 525, 536 (2014) (“Freeze-outs can occur when a minority shareholder is deprived of employment.”); Pointer v. Castellani, 455 Mass. 537, 551 (2009).
 Brodie v. Jordan, 447 Mass. 866, 870–71 (2006) (internal quotations omitted)
 Allison v. Eriksson, 479 Mass. at 638.
 Pursuant to Del. Code Ann. tit. 8, § 342, a Delaware company may incorporate as a close corporation. Such corporate form places restrictions on the number of stockholders and types of stock transfers, as well as prohibits any “public offering” of the company’s stock. However, the statute does not impose any additional fiduciary duties.
 Blaustein v. Lord Baltimore Capital Corp., No. CIV.A. 6685-VCN, 2013 WL 1810956, at *14 (Del. Ch. Apr. 30, 2013), aff’d, 84 A.3d 954 (Del. 2014) (explicitly contrasting its view with the approach of Massachusetts courts).
 Id. (citing Gilbert v. El Paso Co., 1988 WL 124325 (Del. Ch. Nov. 21, 1988))
 Feeley v. NHAOCG, LLC, 62 A.3d 649, 662 (Del. Ch. 2012) (“Managers and managing members owe default fiduciary duties; passive members do not.”)
 Nixon v. Blackwell, 626 A.2d 1366, 1376 (Del. 1993)
 Basho Techs. Holdco B, LLC v. Georgetown Basho Inv’rs, LLC, No. CV 11802-VCL, 2018 WL 3326693, at *49 (Del. Ch. July 6, 2018), aff’d sub nom. Davenport v. Basho Techs. Holdco B, LLC, 221 A.3d 100 (Del. 2019).
 Merriam v. Demoulas Super Markets, Inc., 464 Mass. 721, 727 (2013).
 Selmark Assocs., Inc., 467 Mass. at 539 (“[T]o supplant the otherwise applicable fiduciary duties of parties in a close corporation, the terms of a contract must clearly and expressly indicate a departure from those obligations.”).
 Mass. Gen. Laws Ann. ch. 156C, § 63.
 Mass. Gen. Laws Ann. ch. 156C, § 8.
 Butler v. Moore, No. CIV. 10-10207-FDS, 2015 WL 1409676, at *73 (D. Mass. Mar. 26, 2015); Selmark Assocs., 467 Mass. at 539 (2014) (“[T]o supplant the otherwise applicable fiduciary duties of parties in a close corporation, the terms of a contract must clearly and expressly indicate a departure from those obligations.”).
 JFF Cecilia LLC v. Weiner Ventures, LLC, No. 1984CV03317-BLS2, 2020 WL 4464584, at *11 (Mass. Super. July 30, 2020)
 Butler, 2015 WL 1409676, at *73.
 Id. at *74.
 Nemec v. Shrader, 991 A.2d 1120, 1129 (Del. 2010)
 6 Del. C. § 18-1101
 Marubeni Spar One, LLC, 2020 WL 64761 at *10; see In re Sols. Liquidation LLC, 608 B.R. 384, 407 (Bankr. D. Del. 2019) (finding that plaintiffs’ claims for the defendants’ alleged breach duty of loyalty and good faith were precluded by the LLC agreement, which eliminated such duties).
30] 6 Del. C. § 18-1101.
 Lonergan v. EPE Holdings, LLC, 5 A.3d 1008, 1018 (Del. Ch. 2010).
 Mass. Gen. Laws Ann. ch. 156D, § 8.51
 Mass. Gen. Laws Ann. ch. 156D, § 8.52.
 Mass. Gen. Laws Ann. ch. 156C, § 8
 Del. Code Ann. tit. 8, § 145.
Branin v. Stein Roe Inv. Counsel, LLC, No. CIV. A. 8481-VCN, 2014 WL 2961084, at *4 (Del. Ch. June 30, 2014).
 6 Del. C. § 18-108.
 Majkowski v. Am. Imaging Mgmt. Servs., LLC, 913 A.2d 572, 591 (Del. Ch. 2006).
Nicholas Nesgos is a Partner in the Complex Litigation Group at Arent Fox. He handles a wide variety of business disputes including disputes among shareholders in closely held companies.
Benjamin Greene is an Associate in the Complex Litigation Group at Arent Fox LLP. Benjamin represents a range of individual and corporate clients in complex commercial, employment and real estate litigation.
Representations and warranties insurance (“RWI”) is a common feature of private M&A transactions, aligning the interests of seller and buyer by transferring the risk of a breach of the representations given by the seller in the underlying purchase agreement to an independent, creditworthy insurer. Before “stepping into the shoes of the seller” and issuing a policy to the buyer, the RWI insurer must underwrite several risks, including the seller’s failure to disclose known matters addressed by the representations given in the underlying purchase agreement.
A central pillar of the RWI underwriting process is that parties negotiate at arms’ length, with a seller engaging in a robust disclosure process to ensure known matters are disclosed pursuant to schedules included in the transaction documents. To encourage a thorough “scheduling” process, RWI insurers have historically required sellers to remain liable for a portion of potential losses. A high proportion of transactions are now structured to eliminate the seller’s liability, with such transactions being commonly referred to as “no indemnity” or “public-style” deals. In order to incentivize robust seller disclosure, RWI insurers preserve their right to pursue sellers in the event of seller fraud.
This article explores the rights available to an insurer to mitigate the risk of inadequate disclosure, and those available to a seller to limit the scope of recourse available to a buyer and/or RWI insurer in a transaction.
Representations, Disclosure and Moral Hazard
The primary purpose of a buyer demanding representations in the underlying agreement is to elicit disclosure from a seller. The information obtained from the disclosure exercise enables a buyer to determine the appropriate purchase price. In this way the representations and disclosure act as a “pre-signing” price adjustment mechanism. The secondary purpose of representations is to act as a “post-signing” price adjustment mechanism, allowing a buyer to recoup any overpayments. When a seller is liable for a breach of the representations, there is a clear incentive to fully disclose known matters, as doing so avoids a post-closing claim against the seller. However, in an RWI-backed deal, the seller has limited or no liability which, prima facie, removes the incentive to disclose; indeed, if an insurer bears the risk of a post-closing claim, the seller is incentivized to limit disclosure in order to achieve a higher upfront price. How do insurers control for this “moral hazard”?
“Moral hazard” is the tendency to increase exposure to risk when the consequences of the risk are borne by a third party (e.g., insurer). As the policyholder, a buyer is a party to the RWI insurance contract, and the RWI insurer can control for the buyer’s moral hazard directly. Matters within its knowledge are carved out of coverage through a “no claims declaration” and corresponding exclusion. The no claims declaration operates as an anti-sandbagging provision, precluding a buyer from making a claim for matters of which it had prior actual knowledge.
Of greater importance to an RWI insurer is mitigating the moral hazard risk of a seller not scheduling known matters. As the seller is not a party to the insurance contract, an insurer has no direct means of controlling seller behavior. The insurer must therefore seek to influence the behavior of the seller indirectly – via the rights of a buyer through the doctrine of subrogation.
The principle of subrogation enables an insurer to attempt to recoup its loss once it has paid the insured under the policy. After payment, the insurer can “step into the shoes” of the insured and proceed against any third party responsible for causing loss. This can be any claim that the insured may have against a third party, including contract, tort or statutory claims. Although an insurance policy will typically contain express subrogation provisions, the rights of subrogation will generally apply even if not stipulated in the policy wording. It is important to understand that an insurer’s right of subrogation derives from the rights of the insured. In the context of RWI, this is the right of a buyer against the seller within the underlying purchase agreement.
Why “fraud” matters
An important mechanism for an RWI insurer to incentivize thorough seller disclosure is to retain the right to recoup from a seller any money paid to the buyer as a result of “fraud.”
As explained below, “fraud” has many interpretations, and it is therefore imperative for a seller to define it appropriately. Undefined or poorly drafted fraud carve-outs in the purchase agreement might expose a seller to unintended claims, e.g., fraud of the management team of which a private equity sponsor had no knowledge. “Fraud carve-out” clauses are frequently included within the “limitation provisions” of the purchase agreement, delineating the instances in which a breaching party will be unable to “shield” itself behind the carefully negotiated limitation (e.g., caps, survival periods).
Two Delaware Court of Chancery cases, ABRY Partners v. F&W Acquisition LLC (“ABRY”) and EMSI Acquisition, Inc. v. Contrarian Funds, LLC. (“EMSI”), demonstrate the importance of carefully drafting limitation provisions and associated fraud carve-outs. While ABRY demonstrates that even a well-crafted limitation provision will not shield a seller from its own intentional fraud with respect to express representations and warranties in a transaction document, EMSI highlights the perils of imprecise drafting in exposing the seller to others’ fraud.
In ABRY, the purchase agreement contained a limitation provision capping the seller’s liability at a defined amount with no fraud carve-out. Citing public policy, the court found that, notwithstanding the limitation cap in the agreement, the seller was unable to shield itself from a claim by the buyer in respect of its own intentional fraud that contradicted the express representations and warranties given by it in the agreement.
EMSI highlights the dangers of “inelegant” drafting and the potential for fraud to be imputed on all sellers as a result. In EMSI, the purchase agreement included a fraud carve-out provision that included “any action or claim based upon fraud.” At the pleading stage, the court ruled that such broad language could be interpreted to permit recovery against all sellers, even if those sellers had no knowledge of the fraud and/or were not responsible for the management of the business. Thus, the defendants’ motion to dismiss was not granted. This position highlights the need for sellers to explicitly limit the fraud carve-out as desired.
As a matter of law, the absence of a clearly defined fraud carve-out could result in an extensive scope of possible recourse against a seller, as “undefined fraud is an ‘elusive and shadowy term,’ which may not be limited to deliberate lying despite that common notion.” More specifically, fraud has many meanings, including “common law fraud” (which includes recklessness), “equitable fraud,” “promissory fraud” and “unfair dealings fraud.” Therefore, the possible interpretations of fraud by courts extend beyond “lies” of a seller.
Notably, ABRY ruled with respect to the express representations and warranties set forth in the purchase agreement that “when a seller lies — public policy will not permit a contractual provision to limit the remedy of the buyer to a capped damage claim.” Consistent with ABRY, RWI insurers are primarily concerned with sellers who knowingly make false representations. Therefore, based on ABRY, practitioners representing sellers should seek to limit the definition of “fraud” to a seller’s actual (not constructive) knowledge of the inaccurate representation expressly given in a purchase agreement, made with intent to induce the other party to rely on the misrepresentation. Defining fraud in such a way avoids future claims by buyers/insurers premised on (i) alleged “reckless” or “equitable fraud”; (ii) alleged fraud based on extra-contractual statements (e.g., statements made in meetings but not enshrined as representations in the contract); or (iii) alleged fraud committed by third parties such as management.
As noted above, the subrogation rights of an RWI insurer against a seller derive from those rights of a buyer against the seller. An understanding of an insurer’s subrogation rights therefore requires an examination of a buyer’s rights against the seller. While there are numerous “limitation provisions” in agreements that limit a buyer’s rights against a seller, the principle clauses are the “non-reliance,” “exclusive remedy” and “indemnification and limitation” clauses. Additionally, in an RWI deal, a seller will often require that the agreement contains a “subrogation waiver” clause to limit any claims the insurer, through subrogation, may have against the seller.
Examining each provision in turn:
Through a non-reliance clause, a seller disclaims liability for all representations other than those contained in the agreement; that is, a buyer is unable to make a claim for statements made in management presentations, data rooms, Q&A trackers and other deal documents. This limits a buyer’s rights to the four corners of the agreement. Given the wide scope of potential statements that may be made by the various parties on an M&A transaction (management, advisors, consultants), buyers typically accept that there should be no fraud carve-out to the non-reliance clause, regardless of whether RWI is used on the deal.
Through an exclusive remedy clause, a buyer’s claims (contract and tort) against a seller for a breach of the representations are limited solely to: (i) the indemnification clause and RWI policy on seller indemnity deals; or (ii) the RWI policy for “no indemnity” or “public-style” deals. It is common for buyers to insist on a fraud carve-out to the exclusive remedy provision. This is often accepted by sellers, but only if fraud is appropriately defined.
Through an indemnification and limitation clause, a seller will indemnify a buyer for a breach of the representations, subject to predetermined monetary caps and survival periods. On an RWI-backed deal with limited seller indemnity rights, the representations will typically survive for 12-18 months and be capped at 0.5% of the enterprise value. On a “no indemnity” or “public-style” deal, there will be no indemnification provisions in the agreement. It is common for buyers to insist on a fraud carve-out to the limitation provisions, and this is often accepted by sellers but only if fraud is appropriately defined.
Through a subrogation waiver, a buyer: (i) acknowledges the seller has limited or no liability for a breach of the representations given in the agreement; and (ii) covenants that the RWI insurer will waive any subrogation rights against the seller, save in the event of fraud. Certain sellers will desire that this waiver be given without the fraud carve-out, but this is typically unacceptable to RWI insurers.
Considerations for buyers and sellers
First, sellers must insist that “fraud” is appropriately defined so that it is limited to the seller’s intentional misrepresentation of the express representations in the agreement with intent to deceive.
Second, the parties must assess whether it is reasonable for the “non-reliance,” “exclusive remedy,” “indemnification & limitation” and “subrogation waiver” provisions to contain a fraud carve-out, taking into account RWI insurer requirements.
As previously noted, the “non-reliance clause” will typically not contain a fraud carve-out. An RWI insurer will never require a fraud carve-out, given the RWI policy only covers a breach of the representations given within the four corners of the underlying agreement. The insurer will never be liable for extra contractual representations, so it would be unreasonable and unnecessary for an insurer to request a fraud carve-out to the non-reliance clause.
In light of ABRY, there is a strong argument that RWI insurers should not require a fraud carve-out for “exclusive remedy” and “indemnification and limitation” provisions. This is because, as a matter of law, the seller is unable to shield itself from the type of fraud of which RWI insurers are primarily concerned, so an RWI insurer’s subrogation rights will be unhindered for circumstances in which it will pursue subrogation. Certain insurers (particularly if the agreement is governed by Delaware law) can accept this, while others require a fraud carve-out to the “exclusive remedy” and “indemnification and limitation” provisions. For agreements governed by the laws of other jurisdictions, particularly New York where the case law is less certain, there are still reasonable arguments for RWI insurers to accept no fraud carve-out to the “exclusive remedy” and “indemnification and limitation” provisions, but the arguments are less compelling.
With very rare exceptions, RWI insurers require the “subrogation waiver” provision to include a fraud carve-out. However, as emphasized above, a seller should insist this fraud carve-out is limited to “actual fraud” with “intent to deceive.”
Given the increasing prevalence of “no indemnity” deals, RWI insurers’ requirement to maintain subrogation rights in the event of seller fraud has never been more important. However, it is imperative that “fraud” is appropriately defined to preserve the delicate balance between an RWI insurer’s need to ensure robust disclosure and a seller’s need to avoid post-closing disputes. Lawyers representing sellers should seek to limit an RWI insurer’s rights of subrogation against a seller to instances of “fraud” that law and public policy do not permit to be limited by contract. Consistent with the ruling in ABRY, this means that the definition of fraud should be limited to a seller’s actual knowledge of an inaccurate misrepresentation given in an agreement with intent to induce a buyer to rely on such misrepresentation.
 Atlantic Global Risk, Atlantic Global Risk: M&A Insurance Market – 2019 Insights 7 (2020)
 Special thanks to Glenn D. West, Partner, Weil, Gotshal & Manges LLP, for his wonderful insights and input; and special thanks to Virginia Wong, Senior Analyst, Atlantic Global Risk, for her hard work and contributions to this article.
 See Sean J. Griffith, Deal Insurance: Representation and Warranty Insurance in Mergers and Acquisitions, 104 U. Minn. L. Rev. 4 (Forthcoming) (2020)
 See Sean J. Griffith, Deal Insurance: Representation and Warranty Insurance in Mergers and Acquisitions, 104 U. Minn. L. Rev. 4-5 (Forthcoming) (2020)
 Sean J. Griffith, Deal Insurance: Representation and Warranty Insurance in Mergers and Acquisitions, 104 U. Minn. L. Rev. 53 (Forthcoming) (2020); C.L. Tyagi & Madhu Tyagi, Insurance Law and Practice, 146
 ABRY Partners V, L.P. v. F&W Acquisition LLC, 891 A.2d (Del. Ch. 2006); EMSI Acquisition, Inc. v. Contrarian Funds, LLC, et al., C.A. No. 12648-VCS (Del. Ch. May 3, 2017)
 Glenn D. West, That Pesky Little Thing Called Fraud: An Examination of Buyers’ Insistence Upon (and Sellers’ Too Ready Acceptance of) Undeﬁned “Fraud Carve-Outs” in Acquisition Agreements, The Business Lawyer, Vol. 69 1053 (2014)
 Glenn D. West, That Pesky Little Thing Called Fraud: An Examination of Buyers’ Insistence Upon (and Sellers’ Too Ready Acceptance of) Undeﬁned “Fraud Carve-Outs” in Acquisition Agreements, The Business Lawyer, Vol. 69 1055 (2014)
Richard is a Managing Director and Co-Founder of Atlantic Global Risk, a specialist transactional risk insurance broker. Richard is responsible for directing Atlantic’s strategic growth and direction, including identifying and developing new product lines.
Alvin is an Executive Director and the head of Atlantic’s Boston office, where he counsels clients on risk mitigation solutions for complex regulatory issues and other matters.
by Donald R. Pinto, Jr.
The Massachusetts Land Court is one of a kind. Created in 1898 to administer the then-new land registration system, the court’s jurisdiction has since expanded to encompass an extraordinarily wide range of real estate and land use disputes. The only other state with a Land Court is Hawaii, and that court’s jurisdiction remains limited to land registration matters. The Massachusetts Land Court stands alone as the nation’s only all-purpose real estate specialty court.
Among the many types of cases it now hears, the Land Court has exclusive original jurisdiction over complaints for the confirmation and registration of land, as well as (except for certain domestic relations cases), “[c]omplaints affecting title to registered land . . . .” M.G.L. c. 185, §§ 1(a) & 1(a ½). From time to time this provision prompts questions concerning the jurisdiction of other trial courts over claims involving registered land. The Appeals Court recently addressed such a question in Johnson v. Christ Apostle Church, Mt. Bethel, 99 Mass. App. Ct. 699 (2019). Before turning to Johnson, some background on the development of the Land Court’s expansive jurisdiction will provide useful context.
The Evolution of the Land Court’s Subject Matter Jurisdiction
Originally named the Court of Registration, the Land Court was created by Chapter 562 of the Acts of 1898. The court’s jurisdiction was limited to “exclusive original jurisdiction over all applications for the registration of title to land within the Commonwealth, with power to hear and determine all questions arising upon such applications, and also . . . jurisdiction over such other questions as may come before it under this act . . . .” After a brief period as the Court of Land Registration, in 1904 the court was re-named the Land Court and its exclusive original jurisdiction was expanded to include four causes previously heard by the Superior Court: writs of entry; petitions to require actions to try title; petitions to determine the validity of encumbrances; and petitions to discharge mortgages. During the next 15 years the court was given exclusive original jurisdiction over petitions to: determine the boundaries of tidal flats (another transfer from the Superior Court); determine the existence and extent of a person’s authority to transfer interests in real estate; determine the enforceability of equitable restrictions on land; foreclose tax titles; and determine county, city, town, and district boundaries.
The 1930s saw an even greater expansion of the Land Court’s jurisdiction. In 1931, the court was given original jurisdiction concurrent with the Supreme Judicial Court (“SJC”) and the Superior Court over suits in equity to quiet or establish title to land and to remove clouds from title. In 1934, one of the most significant expansions of Land Court jurisdiction occurred: the court was given original jurisdiction concurrent with the SJC and the Superior Court over “[a]ll cases and matters of equity cognizable under the general principles of equity jurisprudence where any right, title or interest in land is involved, except suits in equity for specific performance of contracts.” In 1934 and 1935, the court also was given exclusive original jurisdiction over petitions under M.G.L. c. 240, § 14A to determine the validity and extent of municipal zoning ordinances, bylaws, and regulations, and original jurisdiction concurrent with the SJC and the Superior Court over suits in equity involving: redemption of tax titles; claims between joint trustees, co-executors and co-administrators; fraudulently conveyed real estate; and conveyances of real estate to municipalities, counties, and other subdivisions of the Commonwealth for specific purposes.
The Land Court’s jurisdiction remained relatively static for the next 40 years. In 1975, the legislature enacted the Zoning Act, M.G.L. c. 40A, and the court’s existing jurisdiction under M.G.L. c. 240, § 14A, was broadened, empowering it to hear (concurrently with the Superior Court) appeals from zoning boards of appeals and special permit granting authorities. Jurisdiction over appeals from planning board decisions under the subdivision control law was added in 1982.
In 1986, in response to confusion over the scope of the Land Court’s exclusive jurisdiction over the land registration system – particularly regarding other trial courts’ ability to decide claims involving registered land – the legislature added to the Land Court’s list of exclusive jurisdictional grants, “[c]omplaints affecting title to registered land . . . .” As will be discussed below, while this language clarified the issue to a degree, it left important questions unanswered.
In 2002, the Land Court’s jurisdiction was again significantly expanded. The court was given original jurisdiction concurrent with the Probate and Family Court (the “Probate Court”) over petitions for partition, and original jurisdiction concurrent with the SJC and the Superior Court over civil actions for trespass to real estate and actions for specific performance of contracts where any right, title, or interest in land is involved. The legislation also expanded the court’s jurisdiction over land-use disputes, granting the court jurisdiction to hear certiorari and mandamus actions under M.G.L. c. 249, §§ 4 and 5 where any right, title, or interest in land is involved “or which arise under or involve the subdivision control law, the zoning act, or municipal zoning, subdivision, or land-use ordinances, by-laws or regulations.” Two notable exceptions to this latter grant of jurisdiction are appeals from decisions of conservation commissions under local wetlands protection ordinances and bylaws and appeals from decisions of boards of health under Title 5 of the state sanitary code.
The most recent expansion of the Land Court’s jurisdiction occurred in 2006, when the legislature established a special “permit session” within the court. This special session provides more intensive case management and expedited handling of cases involving larger real estate developments, defined as those comprising 25 or more dwelling units, or 25,000 or more square feet of gross floor area, or both. In cases accepted into the permit session, the Land Court’s original jurisdiction (which is concurrent with the Superior Court) is even more expansive than its regular jurisdiction, encompassing virtually every type of local, regional, and state land-use permit, approval, order, and certificate. This sweeping jurisdiction includes, for example, appeals from decisions under the Boston zoning code, local wetlands protection ordinances and bylaws, and Title 5 of the state sanitary code – actions that are outside the Land Court’s regular jurisdiction.
It should be noted that in addition to the elements of the Land Court’s jurisdiction compiled in M.G.L. c. 185, § 1, and its permit session jurisdiction set forth in M.G.L. c. 185, § 3A, other statutes confer jurisdiction on the Land Court over other categories of cases. Two notable examples are M.G.L. c. 240, §§ 10A, which gives the Land Court jurisdiction concurrent with the Superior Court over actions to determine the scope and enforceability of restrictions on land, and St. 1943, c. 57, under which the court has jurisdiction concurrent with the Superior Court over suits in equity to determine, in connection with mortgage foreclosures, whether the mortgagor is a servicemember entitled to protection under the federal Servicemembers Civil Relief Act, 50 U.S.C.A. § 3901.
St. 1986, c. 463, § 1
Before 1986, there was uncertainty over the extent to which trial courts other than the Land Court could decide cases involving registered land. For example, a damages claim for breach of a purchase and sale agreement for a parcel of registered land does not affect the title to that land, and thus can be brought in Superior Court. However, a case involving the scope of an easement over registered land presents a more difficult question. In Deacy v. Berberian, 344 Mass. 321 (1962), the plaintiff filed suit in Superior Court seeking to enjoin the defendants from interfering with her use of a right of way over registered land. Based on the language of the original Land Court decree, the defendants claimed that the plaintiff’s use of the way was limited to passage “on foot and with teams,” and that passage by automobiles was precluded. 344 Mass. at 326. On appeal from a judgment for the plaintiff, the defendants argued that the Superior Court lacked jurisdiction to decide the issue. Id. at 328. In response the SJC stated, without further comment or analysis, “[w]e are of opinion [sic] that the purposes of the Land Court Act are not violated by the Superior Court interpreting the original decree so as to give effect to a common mode of transportation.” Id. Similarly, in Cesarone v. Femino, 340 Mass. 638 (1960), the plaintiff filed suit in Superior Court seeking a declaration that he was the owner of a parcel of registered land because his signature on a deed purportedly conveying that parcel was forged. 340 Mass. at 639. On appeal from a judgment for the plaintiff, the defendants argued that because it involved ownership of registered land, the plaintiff’s claim was within the Land Court’s exclusive jurisdiction. Id. The SJC disagreed, characterizing the claim as one based on general principles of equity, concluding, “it appears that either the Land Court or the Superior Court could take jurisdiction.” Id. at 639-640.
In an effort to clarify the scope of the Land Court’s exclusive jurisdiction over registered land and, by implication, the scope of other courts’ jurisdiction over cases involving registered land, in 1986 the legislature – as noted above – amended the court’s main jurisdictional statute, M.G.L. c. 185, § 1, to provide that the court has exclusive jurisdiction over “[c]omplaints affecting title to registered land . . . .” St. 1986, c. 463, § 1; M.G.L. c. 185, §1(a ½). However, it appears this amendment has failed in its mission: while the question whether a claim “affects title” to registered land seems like a simple one, in practice it has proved difficult for our appellate courts to answer in a consistent fashion.
Johnson v. Christ Apostle Church, Mt. Bethel
Such a question was at the center of the Appeals Court’s decision in Johnson v. Christ Apostle Church, Mt. Bethel, 96 Mass. App. Ct. 699 (2019). Johnson involved a dispute between the plaintiff homeowner (“Johnson”) and an adjacent church over Johnson’s use of a driveway on the church’s property that provided access to Johnson’s property. Both properties are registered land. 96 Mass. App. Ct. at 700. After years of peaceful coexistence, in 2013, the church installed a six-foot fence on the property line, which prevented Johnson from continuing to use the driveway. Id. Johnson filed suit in Superior Court asserting claims of negligence, adverse possession, and violation of the “spite fence” statute, M.G.L. c. 49, § 21, which deems certain fences a form of private nuisance. Id. After a trial solely on the nuisance claim, the court ruled for Johnson and ordered the church to install gates in its fence to restore Johnson’s access. Id. at 700-701.
On appeal, though neither side raised the issue, the Appeals Court vacated the judgment on the ground that it effectively granted Johnson “a permanent easement to use the church’s property.” Id. at 701. Citing M.G.L. c. 185, §1(a ½), the Appeals Court held, “[t]he Superior Court does not have jurisdiction to so encumber registered land.” Id. In support of its holding the Appeals Court cited Feinzig v. Ficksman, 42 Mass. App. Ct. 113 (1997), which also involved use of a driveway on registered land. In Feinzig, the Superior Court had entered a judgment enjoining the defendant from interfering with the plaintiffs’ use of the defendant’s land. 42 Mass. App. Ct. at 115. The Appeals Court vacated that judgment, characterizing it as “a de facto encumbrance in the nature of an easement” that affected the defendant’s registered title, and therefore was within the Land Court’s exclusive jurisdiction and outside the jurisdiction of the Superior Court. Id. at 117. The Appeals Court observed, “while a Superior Court judge may order the discontinuance of a trespass on registered land, that judge may not fashion a judgment which has the effect of imposing an encumbrance on the registered title.” Id. at 115-116.
The Appeals Court’s Johnson decision omits any reference to O’Donnell v. O’Donnell, 74 Mass. App. Ct. 409 (2009), a decision that is hard to square with Johnson. In O’Donnell, the defendant mother was embroiled in litigation in the Probate Court with one of her sons over the validity of deeds to two parcels of registered land. 74 Mass. App. Ct. at 411. The mother claimed that those deeds had been procured by undue influence and fraud, and in breach of the son’s fiduciary duty. Id. The son unsuccessfully moved to dismiss the Probate Court action on the ground that it fell within the Land Court’s exclusive jurisdiction over registered land. Id. In response, the son and his brothers filed a new case in the Land Court seeking a declaration that the deeds were valid. Id. The Land Court dismissed that case on the ground of the prior pending Probate Court action, concluding that the judgment the mother sought in the Probate Court “would not of its own force purport to modify the registered title,” and therefore did not intrude on the Land Court’s exclusive jurisdiction over claims “affecting title to registered land.” Id. The Appeals Court affirmed, noting that both the Land Court and the Probate Court have general equity jurisdiction and can decide claims concerning registered land, “as long as the action desired would not have the effect of altering the registered title.” Id. at 412, citing Steele v. Kelley, 46 Mass. App. Ct. 712, 725 (1999). The Appeals Court added that, if a Probate Court judge were to find the deeds valid, “they still would represent no more than ‘a contract between the parties, and . . . evidence of authority to the recorder or assistant recorder [of the Land Court] to make registration.’ A separate act of registration would remain necessary to modify the title directly.” Id., quoting Steele, supra.
It is true that under our system of land registration, with a few narrow exceptions, no matter can formally affect a registered title unless it appears in the certificate of title or is noted on that certificate’s memorandum of encumbrances. M.G.L. c. 185, § 57 crisply states, “[t]he act of registration only shall be the operative act to convey or affect the land.” This is the principle on which O’Donnell rests. But if the Probate Court can enter a judgment determining the validity of a deed to registered land because that judgment itself does not affect the title, why is the Superior Court, in the exercise of its equity jurisdiction, precluded from entering a judgment ordering a defendant to install gates in its fence so that the plaintiff can use the defendant’s registered land (Johnson), or enjoining a defendant from interfering with the plaintiffs’ use of the defendant’s registered land (Feinzig)? After all, such judgments would not of their own force purport to modify the registered title. They would stand simply as adjudications of the parties’ respective rights, and “evidence of authority to the recorder or assistant recorder to make registration.” O’Donnell, supra at 412. Under the reasoning of O’Donnell, it appears, other courts would be free to adjudicate virtually any dispute involving registered land – not only claims concerning the validity of deeds, but claims involving easements and other lesser interests in registered land.
If there is a reasoned way to harmonize the Johnson/Feinzig view of the Land Court’s exclusive jurisdiction over registered land with the O’Donnell view, it is not readily apparent. The Johnson/Feinzig view is preferable in that it comports with the legislature’s presumed intent in 1986 to curb decisions like Deacy and Cesarone, supra, and reserve most disputes involving registered land for resolution by the Land Court, which is solely responsible for administering the registration system and has over a century of expertise in handling such disputes. The distinction that the Feinzig court drew between a claim of trespass on registered land, which does not affect title (at least where the trespasser claims no rights in the land), and a claim of a right to use registered land (whether direct or de facto), which does affect title, is sound and consistent with M.G.L. c. 185, §1(a ½). The O’Donnell view, in contrast, allows for no limiting principle and could lead to a significant erosion of the Land Court’s exclusive jurisdiction over registered land. The real estate bar will be grateful if a future appellate decision resolves the contradiction between these two approaches and finally provides the clarity that the legislature sought to achieve in 1986.
See HRS § 501-1.
 St. 1904, c. 448, § 1.
 St. 1906, c. 50, § 1.
 St. 1906, c. 344, § 1.
 St. 1915, c. 112, § 1.
 St. 1915, c. 237, § 3.
 St. 1919, c. 262, § 1.
 St. 1931, c. 387, § 1.
 St. 1934, c. 67, § 1.
 St. 1934, c. 263, § 1.
 St. 1935, c. 318, §§ 1-5.
 St. 1975, c. 808, § 3.
 St. 1982, c. 533, §§ 1 & 2.
 St. 1986, c. 463, § 1.
 St. 2002, c. 393.
 St. 2006, c. 205, § 15.
 M.G.L. c. 185, § 3A.
Donald R. Pinto, Jr. is a partner of Pierce Atwood LLP based in the firm’s Boston office. He has over 30 years of experience representing clients in all aspects of real estate and land-use litigation in the trial and appellate courts.
by Daniel Lyons
Like many popular tourist destinations, Boston benefits from the sharing economy. Innovative intermediaries such as Airbnb have helped middle-class residents supplement their incomes by monetizing their greatest assets: their homes. The new short-term rental market allows homeowners to keep up with rising living costs while providing additional capacity to attract tourists who contribute to the local economy.
Also like many cities nationwide, Boston has struggled with the unintended consequences of this new marketplace. Policymakers are concerned that the new market is incentivizing owners to remove long-term rentals from the housing stock, particularly in popular and space-constrained areas like Chinatown. To mitigate this risk, a new City of Boston ordinance (City of Boston Code, Ordinances, § 9-14) requires homeowners to register short-term rental properties with the City and prohibits certain categories of properties from being offered as short-term rentals.
But it is the enforcement mechanism that has drawn the most controversy. In addition to punishing individual homeowners who run afoul of the rules, the ordinance fines intermediaries like Airbnb $300 per day for each ineligible rental booked on the site. Presumably, the fine is designed to entice these intermediaries to police their sites for violations. But while this attempt to deputize Airbnb reduces the City’s enforcement costs, it cuts against one of the fundamental tenets of Internet governance: that platforms generally are not liable for a user’s misuse of a neutral tool. This immunity, codified in Section 230 of the Communications Decency Act, 47 U.S.C. § 230, makes it possible for companies from eBay to Twitter to connect millions of users without having to monitor their every interaction for potential legal violations. In Airbnb v. City of Boston, 386 F. Supp. 3d 113 (D. Mass. 2019), the federal district court upheld the Ordinance against a Section 230 challenge, in a decision that weakens this core statutory protection and may have significant ramifications for the broader Internet economy.
Background: Section 230
Section 230 is the legal cornerstone of the modern Internet economy. Jeff Kosseff, Professor of Cybersecurity at the United States Naval Academy describes it as The Twenty-Six Words That Created the Internet. The statute provides that
No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.
Congress passed Section 230 in 1996 to address the holding of Stratton Oakmont v. Prodigy Services Co., 1995 WL 323710 (N.Y. Sup. Ct. May 24, 1995), which held that online service providers could be held liable as publishers for defamatory statements made by their users. Section 230 itself states that it was designed to “preserve the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal or State regulation,” 47 U.S.C. § 230(b)(2), by giving platforms discretion to decide when and how to police their sites. It contains exceptions for claims arising under federal criminal statutes (including, in particular, sex trafficking), intellectual property laws (which are governed by a different intermediary liability regime), or state laws that “are consistent with this section.” 47 U.S.C. § 230(e).
The following year, the seminal case Zeran v. America Online, 129 F.3d 327 (4th Cir. 1997), displayed the expansive scope of the statute in the defamation context. This case involved ads posted on America Online (AOL) selling offensive T-shirts that made light of the 1995 Oklahoma City terrorist bombing. The ads falsely listed plaintiff Ken Zeran as the vendor and included Zeran’s home telephone number, prompting irate AOL users to inundate Zeran with angry calls and death threats. Zeran sued AOL, alleging that he notified the company of the defamatory posts but it unreasonably delayed in removing them. The Fourth Circuit found that Section 230 immunized AOL from liability even for messages that the company knew were defamatory. The court justified this broad immunity by noting that with “millions of users,” interactive computer services process a “staggering” amount of information. Id.. “Faced with potential liability for each message republished by their services, interactive computer providers might choose to severely restrict the number and type of messages posted,” a threat to free speech that Congress sought to guard against. Id..
Subsequent court cases have extended Section 230 far beyond the defamation context, to immunize Craigslist against claims of facilitating housing discrimination, eBay from products liability claims, and StubHub from violations of state ticket scalping laws. It is the resulting broad immunity, protecting intermediaries from liability for most user misconduct, that has shaped much of the current Internet ecosystem. Section 230 entices online news outlets and blogs to permit comment threads without fear of what readers may say. It allows Amazon, TripAdvisor, and Yelp to aggregate and display consumer feedback about products and services. Without Section 230, social media platforms like Facebook and Twitter likely would not exist—or would not be free—because of the high cost of screening every post for potential liability.
Of course, while Section 230 shields the platform from intermediary liability, the user remains liable if the underlying post violates the relevant law. And as the Ninth Circuit explained in Fair Housing Council of San Fernando Valley v Roommate.com, 521 F.3d 1157 (9th Cir. 2008), the platform loses its immunity if it is responsible, in whole or in part, for formulating the offending message.
Section 230 and Boston’s Short-Term Rental Ordinance
Given this robust history of Section 230, it seemed an uphill battle for Boston and similar cities seeking to deputize platforms to enforce short-term rental regulations. Like eBay and StubHub listings, the content of an Airbnb listing is written by the individual homeowner. While a local ordinance could penalize individual homeowners for listing ineligible properties, Section 230 prohibits a local ordinance from forcing Airbnb to “verify” that listed properties comply with the law by punishing it for listing an illegal unit. In 2012, a court struck down a comparable attempt by the State of Washington to fine online classified ad publishers unless they verified that models featured in online prostitution ads were adults. See Backpage.com v. McKenna, 881 F. Supp. 2d 1262 (W.D. Wash. 2012).
Boston sought to circumvent Section 230 by punishing not the listing of an illegal unit, but rather providing booking services for an illegal unit. The law provides that “any Booking Agent who accepts a fee for booking a unit as a Short-Term Rental, where such unit is not an eligible Residential unit, shall be fined” $300 per violation per day. Airbnb sued to enjoin the provision, arguing that the focus on a booking fee rather than the listing was a distinction without a difference, that the effect of the ordinance was to hold intermediaries liable for their users’ misrepresentations, and that Section 230 therefore preempts the ordinance.
On preliminary injunction, the court sided with the City. The court found that the penalty provision punished Airbnb for the company’s own conduct, namely accepting a fee for booking an ineligible unit. The court explained that the fine is not tied to the content of the underlying listing, and noted that Airbnb remains free to list ineligible units without incurring liability, as long as it does not provide booking services for one. In essence, it requires the company, at the booking stage, to confirm that a listing is eligible under the statute before collecting a fee to complete the transaction. The decision mirrored, and relied upon, two recent decisions upholding similar ordinances in California: HomeAway.com, Inc. v. City of Santa Monica, 918 F.3d 676, 680 (9th Cir. 2019), and Airbnb, Inc. v. City & Cty. of San Francisco, 217 F. Supp. 3d 1066, 1071 (N.D. Cal. 2016). In the process, the court rejected Airbnb’s argument that the First Circuit has interpreted Section 230 more broadly than the Ninth Circuit.
Although Airbnb appealed the decision to the First Circuit, it ultimately settled before argument to reduce its financial exposure. Under the settlement agreement, the company agreed to require any user posting a Boston listing to provide a City-issued Registration Number. The company also agreed to send Boston a monthly report of active listings within the City. The City will then notify Airbnb of listings that it believes are ineligible, which Airbnb will deactivate within 30 days. The agreement provides that compliance with this procedure will constitute a safe harbor shielding against booking agent liability under the ordinance.
Unintended Consequences of Court Decision
One can sympathize with Boston’s desire to rein in the excesses of the short-term rental market. Tourist demand for alternatives to traditional lodging remains high, increasing the risk that short-term rentals will siphon off housing stocks in an already capacity-constrained residential market. This is especially problematic if the properties in question receive benefits (such as low-income assistance) designed to encourage residential stability, if the property poses a risk to tourists, or if increased tourist activity harms the local community.
In that sense, it is both expected and appropriate that the City would regulate Boston homeowners who seek to participate in the short-term rental market, just as it does innkeepers and landlords. Boston has authority to decide which properties can be made available and on what terms. And it is free to enforce those regulations directly against individual violators, by dedicating resources to reviewing listings, identifying properties that are out of compliance with the ordinance, and bringing appropriate enforcement action against the lawbreakers.
But the court’s approval of the City’s plan to commandeer platforms to aid enforcement reflects a potentially problematic shift in Section 230 jurisprudence. As an initial matter, the court’s distinction between listing and booking seems strained. The court posited that Airbnb remains free to list illegal units, as long as it doesn’t actually book them. But as Professor Eric Goldman of Santa Clara University notes in connection with the similar San Francisco ordinance, listing properties that the company cannot or will not book could set up Airbnb for a false advertising suit; if it wishes to adhere to its preexisting business model and avoid bait-and-switch liability, the company effectively must verify that listings are eligible before posting.
Even if, as the court suggested, Airbnb need only verify eligibility at the point of booking, the verification obligation imposes significant costs upon these intermediaries. The court minimized this obligation, stating the ordinance “simply requires Airbnb to cross-reference bookings against the City’s list of ineligible units before collecting its fees.” But this simplifies the burden that Airbnb faces. Boston’s ordinance punishes the accepting of a fee for booking an ineligible unit, a category that includes:
- Units subject to affordability covenants or housing assistance under local, state, or federal law;
- Units prohibited from leasing or subleasing under local, state, or federal law; and
- Units subject to three or more violations of any municipal ordinance or state law relating to excessive noise, improper trash disposal, or disorderly conduct within a six-month period.
While the ordinance requires the City to create an ineligible units list, it does not provide a safe harbor for booking agents that cross-reference bookings against that list. On its face, then, booking agents must independently determine whether each Boston booking violates any of the myriad eligibility requirements.
The settlement reduced Airbnb’s compliance costs, but the ordinance remains as written for other booking agents. Of course, the cost of even the settlement’s modified monitor-and-takedown procedure is not trivial—particularly if, as Professor Goldman notes, other cities follow Boston’s example. Airbnb and other intermediaries must keep abreast of nuanced ordinances in myriad cities and states nationwide and tailor their algorithms to verify eligibility. While this increased cost may not make the booking model uneconomic, it could lead some booking companies to withdraw from more heavily regulated markets.
The proliferation of ordinances like Boston’s could also entrench existing companies by raising the costs of entry for new entrepreneurs in this space. Indeed, this could be one reason why Airbnb settled the Boston case and similar litigation in Miami Beach, Florida: as the market leader, Airbnb can perhaps bear these compliance costs easier than its competitors. The settlement agreement itself suggests that Airbnb is using regulation to secure its position: a provision titled “Fairness Across Platforms” requires the City to negotiate with Airbnb’s competitors, three of which are listed by name, mandates that the City provide Airbnb a copy of any agreement it enters with another platform, and provides for Airbnb to modify its agreement if another platform receives a more favorable provision. It also requires the City to confer with Airbnb to discuss compliance efforts taken against platforms that have not entered such agreements.
Ramifications for the Broader Internet Economy
The Boston Airbnb decision shows that the erosion of Section 230 immunity is now spreading beyond the Ninth Circuit. Other cities that share Boston’s concerns about the growth of the short-term rental market now have a model to enlist platform providers as enforcers. For Airbnb and similar platforms, this likely means staffing additional compliance resources to learn and respond to a growing number of local regulations.
Entrepreneurs and those advising platform-based startups should also recognize that this erosion is not necessarily limited to the short-term housing market. The court’s approval of a verification obligation could potentially open the door to significant state and local regulation of the Internet economy. For example, Professor Goldman notes that licensing boards could require that online marketplaces verify that sellers have appropriate business licenses before completing a transaction. Cities may require ride share operators to assure that drivers meet local qualifications. States could require eBay and other clearinghouses to confirm that goods comply with local commerce and product liability laws. And payment processors further up the supply chain could find themselves saddled with similar verification requirements.
The court’s decision also shapes how future tech entrepreneurs should structure their businesses. By bifurcating Airbnb’s listing and booking functions, the decision favors certain business models over others. Airbnb faces liability for facilitating rental of an ineligible property, while online classified ad companies like Craigslist retain Section 230 immunity for the same action, based solely on how each company chooses to fund its activities. Going forward, this decision incentivizes companies to move away from collecting fees for facilitating transactions, and instead to embrace advertising-based revenue models, or models that charge a fee per listing—both of which would remain protected under Section 230.
It is too early to state with precision what effect this decision will have on the development of the sharing economy. But the court’s decision, coupled with the San Francisco and Santa Monica cases, suggest that local regulators may have a powerful new tool to address their public policy concerns. Internet-based platform providers must adapt if they wish to continue relying upon Section 230 to shield innovative new efforts to connect buyers and sellers online.
 As the court clarified, “ineligible” properties are those that categorically cannot be offered as short-term rentals. The statute does not punish booking agents for booking eligible but unregistered properties.
 Airbnb, 386 F.Supp.3d at 120.
 Id. at 120-121.
 The Court contrasted this Penalty Provision with another part of the statute, the “Enforcement Provision,” which prohibits Airbnb from operating within Boston unless it enters an agreement with the city to “actively prevent, remove, or de-list any eligible listings.” See id. at 123-124. At oral argument, the city conceded that the threat of banishment for failure to monitor and remove listings effectively imposed liability on Airbnb for publication of third-party conduct, and on the basis of that concession, the court enjoined the Enforcement Provision. Id. at 123. The court also enjoined parts of a data reporting provision on unrelated grounds. Id. at 124-125.
 Id. at 120 n.5.
 Airbnb, 386 F.Supp.3d at 121.
 See An Ordinance Allowing Short-Term Residential Rentals in the City of Boston, Section 9-14.4A.
Daniel Lyons is a Professor at Boston College Law School, where he researches and writes in the areas of telecommunications, energy, and administrative law. Professor Lyons is also a Visiting Fellow at the American Enterprise Institute, where he regularly blogs about tech policy issues.
After Goodridge: the Potential of Equal Protection Challenges Under the Massachusetts Constitution Involving Non-Economic, Personal InterestsPosted: February 19, 2020
Massachusetts courts apply an enhanced version of rational basis review where regulations infringe on non-economic, personal interests. In light of Goodridge v. Department of Public Health, 440 Mass. 309 (2003), this analysis has the potential to support constitutional challenges on equal protection grounds to two existing Massachusetts statutes related to families. They are: (1) M.G.L. c. 119, § 39D, the grandparent visitation statute; and (2) M.G.L. c. 209C, § 10, custody of children born out of wedlock.
Equal Protection and Enhanced Rational Basis Review in Massachusetts
Equal protection requires “that all persons in the same category and in the same circumstances be treated alike.” Opinion of the Justices, 332 Mass. 769, 779-80 (1955). The standard of review in Massachusetts for equal protection claims that do not involve a fundamental right or suspect class is rational basis review. See Tobin’s Case, 424 Mass. 250, 252-53 (1997).
The Massachusetts Constitution requires that a regulation be “rationally related to the furtherance of a legitimate State interest.” Mass. Fed’n of Teachers, AFT, AFL-CIO v. Bd. of Educ., 436 Mass. 763, 777 (2002) (quoting Chebacco Liquor Mart, Inc. v. Alcoholic Beverages Control Comm’n, 429 Mass. 721, 722 (1999)). However, the Supreme Judicial Court (SJC) has determined that in cases involving non-economic regulations, Massachusetts’s Constitution “may guard more jealously against the exercise of the State’s police power” than the Federal Constitution. That is, the SJC has applied an enhanced version of rational basis review where classifications affect personal, non-economic interests that are not considered fundamental. See Friedman, supra note 2.
The most prominent example of this analysis is Goodridge. There, same-sex couples alleged that the denial of their access to marriage licenses and the status of civil marriage violated the Massachusetts Constitution. The Court held that this exclusion violated equal protection. The Court concluded that “[t]he marriage ban works a deep and scarring hardship on a very real segment of the community for no rational reason,” noting that “[t]he absence of any reasonable relationship between” the exclusion of same-sex couples from marriage and the protection of the general welfare. Goodridge, 440 Mass. at 341 (emphasis added).
As exemplified by Goodridge, enhanced rational basis review under the Massachusetts Constitution requires that there be an actual, rather than merely a conceivable, connection between the government’s legitimate regulatory interest and the imposed regulation, and that the connection be reasonable. See Friedman, supra note 2 at 418. Moreover, Massachusetts’s rational basis review for equal protection claims, “requires that an impartial lawmaker could logically believe that the classification would serve a legitimate public purpose that transcends the harm to the members of the disadvantaged class.” Goodridge, 440 Mass. at 330 (emphasis added) (internal quotations omitted) (quoting English v. New England Med. Ctr., Inc., 405 Mass. 423, 429 (1989)).
Under this enhanced rational basis review, Massachusetts’s grandparent visitation statute, as-applied, arguably violates equal protection under the Massachusetts Constitution.
M.G.L. c. 119, § 39D states in relevant part:
If the parents of an unmarried minor child are divorced, married but living apart, under a temporary order or judgment of separate support, or if either or both parents are deceased, or if said unmarried minor child was born out of wedlock whose paternity has been adjudicated by a court of competent jurisdiction or whose father has signed an acknowledgement of paternity, and the parents do not reside together, the grandparents of such minor child may be granted reasonable visitation rights . . . upon a written finding that such visitation rights would be in the best interest of the said minor child . . . .
The statute, as interpreted by Blixt v. Blixt, 437 Mass. 649 (2002), essentially imposes two conditions for grandparents to seek visitation: that the parents are separated or divorced; and that the grandparents have either a significant preexisting relationship with the grandchildren, or that the grandchildren will suffer significant harm absent visitation with the grandparents. Under Blixt, the second condition amounts to a heightened pleading requirement that grandparents must meet in order to seek visitation.
As-applied, the law discriminates against grandparents of grandchildren whose parents are not divorced or separated with respect to their ability to seek visitation when it is in the grandchildren’s best interest. The statute creates two classes of grandparents: (1) those who can seek visitation because the parents are divorced or separated; and (2) those who cannot because the parents are not divorced or separated. The classes are similarly situated because both sets of grandparents would presumably seek visitation regardless of the parents’ marital or living status on the grounds that it would be in the best interests of their grandchildren.
Applying enhanced rational basis review, the statutorily-created classifications of grandparents do not rationally serve a legitimate public purpose that transcends the harm in denying the opportunity for grandparents of parents who are not separated or divorced from seeking visitation when it is in the grandchildren’s best interest. Three potential rationales for the discrimination are: preserving judicial resources; preventing infringement of parental rights; and safeguarding the welfare of children and giving deference to Blixt’s rationale “that the burden of the traumatic loss of a grandparent’s significant presence may fall most heavily on the child whose unmarried parents live apart” in using parental status to distinguish between classes of grandparents. 437 Mass. at 664; see Goodridge, 440 Mass. at 331. Each of these rationales likely fails enhanced rational basis scrutiny.
First, the requirements for visitation would continue to preserve judicial resources even if grandparents of parents who are not separated or divorced petitioned for visitation because the heightened pleading requirements would serve as a gatekeeping mechanism. Second, the heightened pleading requirements would protect the rights of parents to make child-rearing decisions because only grandparents who can demonstrate that there was a significant preexisting relationship with the grandchildren, or that the grandchildren will be significantly harmed, could seek visitation. Finally, denying certain grandparents the right to seek visitation when it is in the grandchildren’s best interests would not safeguard children’s interests, especially when Blixt acknowledges that there may be children whose parents are not divorced or separated who would be harmed without visitation with their grandparents. See 437 Mass. at 664. It is precisely these factual circumstances that would serve as the basis for an as-applied challenge to the statute, where a court can reconsider the law in the context of an actual dispute. See Wash. State Grange v. Wash. State Republican Party, 552 U.S. 442, 449-50, 457-58 (2008).
Custody of Children Born Out of Wedlock
Similarly, Massachusetts’s statute pertaining to custody of children born out of wedlock arguably violates equal protection under the Massachusetts Constitution pursuant to enhanced rational basis review analysis.
M.G.L. c. 209C, § 10(a)-(b) provides in relevant part:
(a) Upon or after an adjudication or voluntary acknowledgment of paternity, the court may award custody to the mother or the father or to them jointly . . . as may be appropriate in the best interests of the child . . . . In awarding the parents joint custody, the court shall do so only if the parents have entered into an agreement pursuant to section eleven or the court finds that the parents have successfully exercised joint responsibility for the child prior to the commencement of proceedings pursuant to this chapter and have the ability to communicate and plan with each other concerning the child’s best interests.
(b) Prior to or in the absence of an adjudication or voluntary acknowledgment of paternity, the mother shall have custody of a child born out of wedlock. In the absence of an order or judgment of a probate and family court relative to custody, the mother shall continue to have custody of a child after an adjudication of paternity or voluntary acknowledgment of parentage.
As-applied, the law denies unwed fathers not cohabitating with the mother, the right to equal protection by discriminating against these fathers as compared to divorcing fathers when joint legal custody is in the child’s best interest. Pursuant to M.G.L. c. 208, § 31, divorcing fathers are presumed to have joint legal custody of a child until otherwise ordered and need only prove that it is in the child’s best interest for it to be awarded. Whereas, pursuant to M.G.L. c. 209C, § 10(a)-(b), unwed fathers not cohabitating with the mother are not presumed to have joint legal custody and must meet additional and burdensome requirements to have it awarded. The fathers are similarly situated because both groups are men whose parentage has either been assumed due to their marital status or established by adjudication or acknowledgment, and whose children would presumably benefit from their involvement in their life.
To be awarded joint legal custody absent an agreement with the mother, in addition to satisfying the best interest standard, an unwed father not cohabitating with the mother is required to prove that he and the mother “have successfully exercised joint responsibility for the child prior to the commencement of proceedings;” and he and the mother have “the ability to communicate and plan with each other concerning the child’s best interests.” M.G.L. c. 209C, § 10(a). A divorcing father, however, enjoys “temporary shared legal custody of any minor child of the marriage” and he need only ultimately prove that an award of joint custody is in the child’s best interest. M.G.L. c. 208, § 31.
Under an enhanced rational basis review analysis, the different treatment of unwed fathers not cohabitating with the mother and divorcing fathers pursuant to M.G.L. c. 209C, § 10(a)-(b) likely does not rationally serve a legitimate public purpose that transcends the harm in requiring that unwed fathers meet more onerous requirements. Potential public purposes that may be advanced to justify the discrimination are: preserving judicial resources; and safeguarding the wellbeing of children.
First, judicial resources would not be further expended if both classes of fathers were afforded the same presumption and held to the same standard to be awarded joint legal custody. In fact, judicial resources may be conserved if all fathers were afforded the presumption and only the child’s best interest governed. Second, the wellbeing of children is not enhanced by requiring that unwed fathers not cohabitating with the mother meet more burdensome requirements or be denied a presumption of joint legal custody because an analysis of the child’s best interest should afford all children the same protection, especially where there are divorcing fathers who do not have a history of successfully exercising joint responsibility for the child and cannot communicate or cooperate with the mother. If children of unwed fathers not cohabitating with the mother are afforded greater protections than children of divorcing fathers, this likely violates “the Commonwealth’s strong public policy to abolish legal distinctions between marital and nonmarital children in providing for the support and care of minors.” Goodridge, 440 Mass. at 325 (citations omitted); see M.G.L. c. 209C, § 1 (“Children born to parents who are not married to each other shall be entitled to the same rights and protections of the law as all other children.”). Further, these additional and burdensome requirements may discourage fathers whose children would benefit from their participation in important decisions from pursuing these rights. See Cuadra, supra note 6 at 634-35.
Looking to Goodridge as a roadmap for an enhanced rational basis review analysis, some Massachusetts statutes impacting personal interests, like a grandparent’s ability to pursue visitation or an unwed father’s ability to participate in important child rearing decisions, may well be vulnerable to challenges on equal protection grounds. This kind of judicial review recognizes that there may be instances in which legislative classifications draw lines that infringe on personal interests without adequate justification. This review does not mean every statute that differentiates among significant personal interests will fail—just that the Commonwealth must be able to articulate a reason for the discrimination that has a basis in fact. On the right facts, an enhanced rational basis review analysis has the potential to push the legal landscape to better serve families by pushing the Commonwealth either to justify the discrimination or abandon distinctions that no longer make sense.
Steven E. Gurdin is a partner at Fitch Law Partners LLP, concentrating his practice in and frequently presenting on family law and probate litigation matters. He represents clients in all aspects of family law including, divorce proceedings, paternity actions, child removal actions, modification and contempt actions, grandparent visitation cases, and alimony and child support issues.
Kelly A. Schwartz is an associate at Fitch Law Partners LLP. Her practice is in family law, which includes matters involving divorce, child custody, alimony, child support, and asset division.
 Special thanks to Lawrence Friedman, Professor of Law at New England Law, for his input and support.
 See Blue Hills Cemetery, Inc. v. Bd. of Reg. in Embalming & Funeral Directing, 379 Mass. 368, 373 n.8 (1979); Coffee-Rich, Inc. v. Comm’r of Pub. Health, 348 Mass. 414, 421-22 (1965); Lawrence Friedman, Ordinary and Enhanced Rational Basis Review in the Massachusetts Supreme Judicial Court: A Preliminary Investigation, 69 Albany L. Rev. 415 (2006) (discussing history of Massachusetts cases in which courts applied less deferential rational basis review under the Massachusetts Constitution); see also Goodridge, 440 Mass. at 328 (“The Massachusetts Constitution protects matters of personal liberty against government incursion as zealously, and often more so, than does the Federal Constitution, even where both Constitutions employ essentially the same language.”), n.18 (“We have recognized that our Constitution may more extensively protect individual rights than the Federal Constitution in widely different contexts.”).
 See Friedman, supra note 2 at 419-22, 440.
 Of note, Massachusetts House Bill No. 1534, which was submitted in the current session on January 18, 2019, provides that any grandparent may file an original action for visitation rights if it is in the child’s best interest and if, among other conditions, “the child is living with biological parents, who are still married to each other, whether or not there is a broken relationship between either or both parents of the minor and the grandparent and either or both parents have used their prenatal authority to prohibit a relationship between the child and the grandparent.”
 See Dep’t of Revenue v. C.M.J., 432 Mass. 69, 77 (2000) (determining that under M.G.L. c. 209C, § 10(b), an unwed father cohabitating with the mother and providing support to his children was presumed a custodial parent even absent a court order regarding custody); Trial Court Judgment of Dismissal and Memorandum of Decision (Gibson, J. Nov. 5, 2013); see also Com. v. Gonzalez, 462 Mass. 459, 464 & n.12 (2012); 14B Mass. Prac., Summary of Basic Law, § 8:264 (5th ed. 2019) (“[I]f the father is living with the mother and the child or children born out of wedlock, the mother’s custody is not sole custody, but joint custody with the father.”) & nn.4-5.
 See Bernardo Cuadra, Family Law–Maternal and Joint Custody Presumptions for Unmarried Parents: Constitutional and Policy Considerations in Massachusetts and Beyond, 32 W. New Eng. L. Rev. 599, 620-22 (2010) (footnote notation omitted) (discussing the difference in Massachusetts between divorcing parents and unmarried parents with regard to the presumption of joint legal custody).
 Department of Revenue v. C.M.J. did not, however, address whether unwed fathers cohabitating with the mother must also meet the additional requirements outlined in M.G.L. 209C, § 10(a) to be awarded joint legal custody.
 See Cuadra, supra note 6 at 633-34 (“Joint legal custody has been shown in the context of divorce to increase a father’s involvement with his child, including parenting time and overnights. Subsequent research suggests the same outcome for unwed fathers.”).
 See Goodridge, 440 Mass. at 330; see also Cuadra, supra note 6 at 639 (suggesting that Massachusetts’s different treatment of divorcing and unwed fathers should be examined pursuant to a rational basis review where the statute would “be scrutinized with something greater than sweeping deference.”) & n.251 (citing and quoting Goodridge).
 See Goodridge, 440 Mass. at 331; Blixt, 437 Mass. at 663; Trial Court Judgment, supra note 5.